100% found this document useful (5 votes)
21 views82 pages

(Ebook PDF) Multimedia Enabled Sensors in IoT Data Delivery and Traffic Modelling 1st edition by Fadi Al Turjman 1351166026 9781351166027 full chapters - Download the complete ebook in PDF format and read freely

The document provides access to various eBooks authored by Fadi Al Turjman, focusing on topics related to IoT, multimedia-enabled sensors, and wireless sensor networks. It includes links for instant downloads in multiple formats and highlights the content of each book, such as routing protocols and data delivery models. The document also contains bibliographic information and acknowledgments for the author's contributions.

Uploaded by

buetyramilo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100% found this document useful (5 votes)
21 views82 pages

(Ebook PDF) Multimedia Enabled Sensors in IoT Data Delivery and Traffic Modelling 1st edition by Fadi Al Turjman 1351166026 9781351166027 full chapters - Download the complete ebook in PDF format and read freely

The document provides access to various eBooks authored by Fadi Al Turjman, focusing on topics related to IoT, multimedia-enabled sensors, and wireless sensor networks. It includes links for instant downloads in multiple formats and highlights the content of each book, such as routing protocols and data delivery models. The document also contains bibliographic information and acknowledgments for the author's contributions.

Uploaded by

buetyramilo
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 82

Instant Ebook Access, One Click Away – Begin at ebookball.

com

(Ebook PDF) Multimedia Enabled Sensors in IoT Data


Delivery and Traffic Modelling 1st edition by Fadi
Al Turjman 1351166026 9781351166027 full chapters

https://ebookball.com/product/ebook-pdf-multimedia-enabled-
sensors-in-iot-data-delivery-and-traffic-modelling-1st-
edition-by-fadi-al-turjman-1351166026-9781351166027-full-
chapters-10478/

OR CLICK BUTTON

DOWLOAD EBOOK

Get Instant Ebook Downloads – Browse at https://ebookball.com


Your digital treasures (PDF, ePub, MOBI) await
Download instantly and pick your perfect format...

Read anywhere, anytime, on any device!

(Ebook PDF) Cognitive Sensors and Iot Architecture


Deployment and Data Delivery 1st edition by Fadi Al
Turjman 1351595563 9781351595568 full chapters

https://ebookball.com/product/ebook-pdf-cognitive-sensors-and-iot-
architecture-deployment-and-data-delivery-1st-edition-by-fadi-al-
turjman-1351595563-9781351595568-full-chapters-10474/

ebookball.com

(Ebook PDF) Artificial intelligence in IoT 1st edition by


Fadi AI Turjman 3030041107 9783030041106

https://ebookball.com/product/ebook-pdf-artificial-intelligence-in-
iot-1st-edition-by-fadi-ai-turjman-3030041107-9783030041106-13536/

ebookball.com

(Ebook PDF) Smart Things and Femtocells From Hype to


Reality 1st edition by Fadi Al Turjman 0429951191
9780429951190 full chapters

https://ebookball.com/product/ebook-pdf-smart-things-and-femtocells-
from-hype-to-reality-1st-edition-by-fadi-al-
turjman-0429951191-9780429951190-full-chapters-10468/

ebookball.com

(Ebook PDF) Wireless Sensor Networks Deployment Strategies


for Outdoor Monitoring 1st edition by Fadi Al Turjman
1351207016 9781351207010 full chapters

https://ebookball.com/product/ebook-pdf-wireless-sensor-networks-
deployment-strategies-for-outdoor-monitoring-1st-edition-by-fadi-al-
turjman-1351207016-9781351207010-full-chapters-14872/

ebookball.com
(Ebook PDF) Internet of Nano Things and Wireless Body Area
Networks WBAN 1st edition by Fadi Al Turjman 0429516851
9780429516856 full chapters

https://ebookball.com/product/ebook-pdf-internet-of-nano-things-and-
wireless-body-area-networks-wban-1st-edition-by-fadi-al-
turjman-0429516851-9780429516856-full-chapters-10472/

ebookball.com

(Ebook PDF) Security in Iot Enabled Spaces 1st edition by


CRC Press 042962817X 9780429628177 full chapters

https://ebookball.com/product/ebook-pdf-security-in-iot-enabled-
spaces-1st-edition-by-crc-press-042962817x-9780429628177-full-
chapters-10470/

ebookball.com

Wireless Sensor Networks Deployment Strategies for Outdoor


Monitoring 1st Edition by Fadi Al Turjman ISBN
9781351207010 1351207016

https://ebookball.com/product/wireless-sensor-networks-deployment-
strategies-for-outdoor-monitoring-1st-edition-by-fadi-al-turjman-
isbn-9781351207010-1351207016-10464/

ebookball.com

(Ebook PDF) Biomedical Signals and Sensors III Linking


Electric Biosignals and Biomedical Sensors 1st edition by
Eugenijus Kaniusas 331974917X 9783319749174 full chapters

https://ebookball.com/product/ebook-pdf-biomedical-signals-and-
sensors-iii-linking-electric-biosignals-and-biomedical-sensors-1st-
edition-by-eugenijus-kaniusas-331974917x-9783319749174-full-
chapters-22584/
ebookball.com

(Ebook PDF) Sensors and Actuators in Mechatronics Design &


Applications 1st edition by Andrzej M Pawlak 1351837524
9781351837521 full chapters

https://ebookball.com/product/ebook-pdf-sensors-and-actuators-in-
mechatronics-design-applications-1st-edition-by-andrzej-m-
pawlak-1351837524-9781351837521-full-chapters-25460/

ebookball.com
Multimedia-enabled
Sensors in IoT
Data Delivery and Traffic Modelling
Multimedia-enabled
Sensors in IoT
Data Delivery and Traffic Modelling

Fadi Al-Turjman
Antalya Bilim University, Turkey
MATLAB ® is a trademark of The MathWorks, Inc. and is used with permission. The MathWorks
does not warrant the accuracy of the text or exercises in this book. This book’s use or discussion
of MATLAB ® software or related products does not constitute endorsement or sponsorship by The
MathWorks of a particular pedagogical approach or particular use of the MATLAB ® software.

CRC Press
Taylor & Francis Group
6000 Broken Sound Parkway NW, Suite 300
Boca Raton, FL 33487-2742
© 2018 by Taylor & Francis Group, LLC
CRC Press is an imprint of Taylor & Francis Group, an Informa business
No claim to original U.S. Government works
Printed on acid-free paper
International Standard Book Number-13: 978-0-8153-8711-4 (Hardback)
This book contains information obtained from authentic and highly regarded sources. Reasonable efforts
have been made to publish reliable data and information, but the author and publisher cannot assume
responsibility for the validity of all materials or the consequences of their use. The authors and publishers
have attempted to trace the copyright holders of all material reproduced in this publication and apologize
to copyright holders if permission to publish in this form has not been obtained. If any copyright material
has not been acknowledged please write and let us know so we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced,
transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or
hereafter invented, including photocopying, microfilming, and recording, or in any information
storage or retrieval system, without written permission from the publishers.
For permission to photocopy or use material electronically from this work, please access www.­
copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC),
222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that
provides licenses and registration for a variety of users. For organizations that have been granted a
photocopy license by the CCC, a separate system of payment has been arranged.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and
are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging–in–Publication Data

Names: Al-Turjman, Fadi, author.


Title: Multimedia-enabled sensors in IoT data delivery and traffic modelling /
Fadi Al-Turjman.
Description: Boca Raton, FL : CRC Press, Taylor & Francis Group, 2017. |
Includes bibliographical references.
Identifiers: LCCN 2017048014 | ISBN 9780815387114 (hb : acid-free paper)
Subjects: LCSH: Internet of things. | Wireless sensor networks. |
Telecommunication--Traffic.
Classification: LCC TK5105.8857 .A43 2017 | DDC 006.2/5--dc23
LC record available at https://lccn.loc.gov/2017048014

Visit the Taylor & Francis Web site at


http://www.taylorandfrancis.com
and the CRC Press Web site at
http://www.crcpress.com
To whom I owe the true love,
To whom the depths of my soul are thundering, and the poem is created…
To the most important woman in my history, who thinks the same without speech
and speaks the same without thinking…

To my wonderful wife,
Fadi Al-Turjman
Contents

Author................................................................................................................xiii
1 Introduction............................................................................................1
1.1 Contributions.....................................................................................3
1.1.1 Book Outline.........................................................................3
References.....................................................................................................4
2 A Survey on Multipath Routing Protocols for QoS Assurances
in Real-Time Wireless Multimedia Sensor Networks..............................7
2.1 Introduction.......................................................................................7
2.2 Comparison with Related Survey Articles........................................11
2.3 Routing System Architecture and Design Issues...............................21
2.3.1 Hard and Soft Real-Time Operation and Best-Effort
for Resource Constraints......................................................21
2.3.1.1 Hard Real-Time (HRT)........................................21
2.3.1.2 Soft Real-Time (SRT)...........................................21
2.3.1.3 Best-Effort (BE)....................................................21
2.3.2 Energy Efficiency Model......................................................22
2.3.3 QoS Modeling Requirements...............................................23
2.3.3.1 Latencies...............................................................24
2.3.3.2 Bandwidth............................................................24
2.3.4 Sensor Network...................................................................25
2.3.5 Data Delivery Model...........................................................26
2.3.5.1 Continuous Time-Driven Delivery Model............26
2.3.5.2 Event-Driven and Query-Driven Models..............26
2.3.5.3 Hybrid Model.......................................................26
2.3.6 Dynamic Network...............................................................26
2.3.7 Reliability and Fault Tolerance............................................27
2.3.7.1 Local Communication (Node-to-Node)...............27
2.3.7.2 Point-to-Point (Node-Node).................................27
2.3.7.3 Convergence (Node-to-Sink)................................28
2.3.7.4 Divergence (Sink-to-Node)....................................28
2.3.8 Summary.............................................................................29

vii
viii ◾ Contents

2.4 Routing Techniques in WMSNs Classification................................29


2.4.1 Designing Issues for Multipath Routing..............................30
2.4.2 The Taxonomy of Multipath Routing Techniques................30
2.4.3 Summary.............................................................................32
2.5 Multipath Routing Protocols: Challenges and Issues........................33
2.5.1 Multipath Discovery........................................................... 34
2.5.1.1 Partially Disjoint Paths.........................................35
2.5.1.2 Braided Multipath.................................................35
2.5.1.3 Disjoint Paths........................................................36
2.5.2 Multipath Forwarding Models.............................................38
2.5.2.1 Path Selection.......................................................38
2.5.2.2 Distribution Traffics Splitting Pattern...................39
2.5.3 Maintenance of Paths.......................................................... 40
2.5.4 Summary.............................................................................41
2.6 Concurrent Multipath Unicast Forwarding......................................41
2.6.1 Multipath Multi-QoS Constraints for Efficient
Resource Allocation.............................................................41
2.6.1.1 Principal Protocols of Multipath Multi-QoS
Constraints.......................................................... 42
2.6.1.2 Discussion.............................................................47
2.6.2 Multipath Reliability Constraint for Reliable Data
Transmission........................................................................48
2.6.2.1 Principal Protocols for Multipath Reliability
Constraint.............................................................53
2.6.2.2 Discussion.............................................................60
2.7 Alternative Multipath Broadcast Flooding........................................61
2.7.1 Data-Centric Protocol Operation.........................................65
2.7.2 Data-Centric Protocol Problem............................................65
2.7.2.1 Principal Protocols................................................65
2.7.2.2 Discussion.............................................................67
2.7.3 On-Demand Fashion...........................................................68
2.7.3.1 DSR Principal Protocols........................................69
2.7.3.2 AODV Principal Protocols....................................70
2.7.3.3 Proactive Routing..................................................70
2.7.4 Discussion............................................................................71
2.8 Simulation Comparisons..................................................................73
2.9 Open Research Problems..................................................................76
2.9.1 Data Sensing and Delivery Model........................................76
2.9.2 Node Deployment...............................................................81
2.9.3 Node Capabilities................................................................81
2.9.4 Link Quality Estimators (LQEs)..........................................81
2.9.5 Mobility...............................................................................82
2.9.6 Scalability............................................................................82
Contents ◾ ix

2.9.7 Multimedia Content............................................................82


2.9.8 Energy Efficiency Considerations.........................................83
2.9.9 Multi-Constrained QoS Guarantee.....................................83
2.9.10 Cognitive Radio (CR)..........................................................83
2.10 Conclusion...................................................................................... 84
Acknowledgments...................................................................................... 84
References.................................................................................................. 84

3 Optimized Multi-Constrained Quality-of-Service Multipath


Routing Approach for Multimedia Sensor Networks............................95
3.1 Introduction.....................................................................................95
3.2 Related Works..................................................................................97
3.3 Partitioning Multi-Constrained Multipath Routing (PMMR)
Protocol............................................................................................97
3.3.1 Problem Formulation...........................................................98
3.3.2 Link Quality Modeling........................................................98
3.3.3 Neighboring Node-Disjointed Discovery Procedure..........103
3.3.4 Path-Disjointed Discovery Procedure.................................104
3.3.5 Path-Disjointed Selection Procedure..................................104
3.4 Multi-Constraints QoS Parameters Modeling................................105
3.4.1 Power-Consumption Modeling..........................................106
3.4.2 Delay-Constraint Modeling...............................................108
3.5 Performance Evaluation.................................................................. 110
3.5.1 Experiment 1: Effectiveness of Cut-off Determination....... 111
3.5.2 Experiment 2: Comparison against Three Routing
Algorithms......................................................................... 115
3.6 Conclusions....................................................................................123
References.................................................................................................123

4 Green Data Delivery Framework for Safety-Inspired Multimedia


in Mobile IoT.......................................................................................127
4.1 Introduction...................................................................................127
4.2 Related Work..................................................................................129
4.2.1 Multipath Unicast Forwarding..........................................130
4.2.1.1 Multipath QoS-Based Protocols..........................130
4.2.1.2 Reliability Constraints........................................131
4.2.2 Alternative Multipath Broadcasting...................................132
4.2.2.1 Data-Centric Approaches....................................132
4.2.2.2 On-Demand Approaches....................................133
4.3 System Model.................................................................................135
4.3.1 Network Architecture........................................................135
4.3.2 Lifetime and Energy Model...............................................136
4.3.3 Communication Model.....................................................137
x ◾ Contents

4.4 Multipath Disruption-Tolerant Approach (MDTA).......................138


4.5 Theoretical Analysis on Lifetime....................................................141
4.6 Performance Evaluation..................................................................143
4.6.1 Performance Metrics and Parameters.................................143
4.6.2 Simulation Setup...............................................................144
4.6.3 Simulation Results............................................................. 145
4.7 Conclusions.................................................................................... 151
References................................................................................................. 151

5 A Delay-Tolerant Framework for Integrated RSNs in IoT...................153


5.1 Introduction................................................................................... 153
5.2 Related Work..................................................................................156
5.2.1 Architectures for Integrated RSNs.....................................156
5.2.2 Node Placement in Integrated Architectures......................160
5.2.3 Data Transfer in DTNs.....................................................162
5.3 System Models................................................................................163
5.3.1 Network Model.................................................................163
5.3.2 Delay Model......................................................................165
5.3.3 Communication Model.....................................................166
5.4 Integrated RSN Framework............................................................167
5.4.1 Optimal Placement of Super Nodes...................................169
5.4.2 CN Selection.....................................................................172
5.5 Discussion and Results................................................................... 176
5.5.1 Simulation Model..............................................................177
5.5.2 Simulation Results.............................................................177
5.6 Conclusion.....................................................................................183
Acknowledgments.....................................................................................184
References.................................................................................................184

6 Multimedia-Enabled WSNs Using UAVs for Safety-Oriented


Mobile IoT...........................................................................................187
6.1 Introduction...................................................................................187
6.2 Related Work..................................................................................189
6.3 System Model.................................................................................190
6.3.1 Problem Formulation......................................................... 191
6.3.2 Energy Model.................................................................... 191
6.3.3 Delay Model......................................................................192
6.3.4 Throughput Model.............................................................193
6.4 Particle Swarm Optimization (PSO) Algorithm.............................193
6.5 Performance Evaluation..................................................................196
6.5.1 Simulation Results.............................................................196
6.6 Conclusion.....................................................................................199
References.................................................................................................199
Contents ◾ xi

7 Evaluation of a Duty-Cycled Asynchronous X-MAC Protocol


for VSNs���������������������������������������������������������������������������������������������201
7.1 Introduction...................................................................................201
7.2 Related Works................................................................................203
7.3 Overview of the X-MAC Protocol................................................. 206
7.4 Markov Model of X-MAC..............................................................207
7.4.1 The Hidden-Problem Formulation.....................................212
7.4.2 Media Access Rules of X-MAC.......................................... 214
7.5 QoS Parameters Analysis of the X-MAC Protocol..........................216
7.6 Simulation Results......................................................................... 220
7.6.1 Varying the Cycle Length..................................................221
7.6.2 Varying the Number of Nodes...........................................224
7.6.3 Discussion......................................................................... 226
7.7 Conclusion.....................................................................................227
7.8 Competing Interests.......................................................................227
7.9 Funding..........................................................................................227
7.10 Author’s Contributions.................................................................. 228
Acknowledgments.................................................................................... 228
References................................................................................................ 228

8 Mobile Traffic Modeling for Wireless Multimedia Sensor


Networks in IoT...................................................................................231
8.1 Introduction...................................................................................231
8.2 Related Work..................................................................................233
8.3 QoS-Based Multimedia Traffic Modeling Framework....................235
8.3.1 Analysis of Retransmission Channel
Access Schemes..................................................................235
8.3.2 Modeling Duty-Cycle Node Operations............................235
8.3.3 Energy and Delay Modeling..............................................238
8.3.4 Throughput Modeling.......................................................239
8.3.5 Path Loss Model................................................................239
8.3.6 Failure Modeling...............................................................240
8.3.6.1 Coverage Factor..................................................240
8.3.6.2 Sensor Failure Rate.............................................240
8.4 Use-Case Transmission Modeling..................................................240
8.5 Performance Evaluation..................................................................242
8.5.1 Simulation Results.............................................................243
8.5.1.1 The Impact of Radio Irregularity on Energy
Consumption......................................................243
8.5.1.2 The Impact of Radio Irregularity on Average
Delay.................................................................. 244
8.6 Concluding Remarks......................................................................245
References................................................................................................ 246
xii ◾ Contents

9 Information-Centric Framework for the IoT: Traffic Modeling


and Optimization................................................................................249
9.1 Introduction...................................................................................249
9.2 Related Work..................................................................................252
9.3 System Models................................................................................253
9.3.1 Network Model.................................................................253
9.3.2 Delay and Disruption Model.............................................254
9.3.3 Traffic Representation Model.............................................256
9.3.4 Problem Statement.............................................................258
9.4 CDE-Based Framework..................................................................259
9.5 Real Scenarios and Case Studies.................................................... 264
9.6 Simulation Results and Discussions................................................267
9.7 Conclusion.....................................................................................277
References.................................................................................................278
10 Conclusions and Future Directions.....................................................281
10.1 Summary of the Book....................................................................281
10.1.1 Future Directions.............................................................. 284

Index............................................................................................................287
Author

Dr. Fadi Al-Turjman is a professor at Antalya Bilim


University, Antalya, Turkey. He earned his PhD
in computing science from Queen’s University,
Canada, in 2011. He is a leading authority in the
areas of smart/cognitive, wireless and mobile net-
works’ architectures, protocols, deployments, and
performance evaluation. His record spans more
than 150 publications in journals, conferences,
patents, books, and book chapters, in addition to
numerous keynotes and plenary talks at flagship
venues. He has received several recognitions and
best paper awards at top international conferences
and led a number of international symposia and workshops in flagship ComSoc
conferences. He is serving as the lead guest editor in several journals including
the IET Wireless Sensor Systems (WSS), MDPI Sensors and Wiley. He also served
as the general workshops chair for the IEEE International Conference on Local
Computer Networks (LCN’17). Recently, he published Cognitive Sensors & IoT:
Architecture, Deployment, and Data Delivery with Taylor & Francis, CRC, Boca
Raton, Florida. Since 2007, he has been working on international wireless sensor
networks (WSNs) projects related to remote monitoring, as well as Smart Cities
related deployments and data-delivery protocols using integrated RFID (Radio
Frequency Identification)-Sensor Networks (RSNs).

xiii
Chapter 1

Introduction

Recent developments in wireless sensor network (WSN) technology (which


enables communications ranging from a few meters to city-scale applications in
order to perform simple tasks such as sensing, actuation, and computing) have
promoted a new class of multimedia applications in multidisciplinary domains,
which significantly depend on sensing technologies [1]. The dramatic development
of sensing/communication techniques makes it the most appropriate technology
in the Internet of things (IoT) paradigm. The recent advancements of WSNs in
the IoT have been widely promoted in environmental, industrial, and biomedical
sensing and monitoring applications, which significantly depend on real-time data
[1]. Consequently, a new technology called wireless multimedia sensor networks
(WMSNs) has emerged to achieve more reductions and savings in terms of mul-
timedia gathering cost. WMSNs can gather and deliver multimedia data such as
sound streams, images, video, and scalar data (e.g., temperature/humidity read-
ings). WMSNs have not only enhanced the existing applications of traditional
WSNs but have also enabled new services. However, there are several restrictions
facing their design and implementation in reality while satisfying the desired
quality of service (QoS) [2]. In this book, we focus on the constraints related to
QoS requirements and what necessitates efficient routing protocols. Such rout-
ing protocols are needed for real-time and non–real-time multimedia applica-
tions. However, the former requires different QoS metrics and design aspects,
which are supposed to be considered at the data networking/routing protocols.
Consequently, WMSNs have evolved in the literature with a special kind of rout-
ing protocols called “multipath” routing protocols [3]. Multipath routing can be
considered as a new technique in delivering data via several routing paths. Unlike
the single-path routing, multipath routing protocols provide “more” adequate
network resources, improve the packet arrival ratio of multimedia contents, and

1
2 ◾ Multimedia-Enabled Sensors in IoT

guarantee the QoS at high data rates [3]. This kind of protocol is cost-effective for
overloaded networks and lasts for prolonged lifetime periods. Accordingly, in this
book, we investigate the importance of multipath routing strategies that assure the
QoS in the IoT paradigm while contrasting their pros and cons, mechanisms for
multipath routing, classifications in terms of the multimedia delivery exploitation,
and design issues.
Most of the literature surveys about WMSNs routing protocols care about the
multimedia application and the network architecture requirements while ignoring
significant criteria, such as the lifetime of the WMSN and key QoS requirements
in these networks. The authors in Reference 4 discussed constraints and solutions
related to each layer of the WSN stack. In Reference 5, the authors reviewed sev-
eral routing protocols and a classification based on multipath availability, query
type, and in-network negotiation was proposed. In Reference 6, a taxonomy was
developed and certain design issues such as energy efficiency, geographic location,
network scalability, and cross-layer approaches were considered. The authors in
Reference 2 introduced protocols that support real-time applications in WSNs.
Existing routing strategies were classified based on the network structure, commu-
nication model, and topology shape. Position-based routing algorithms and metrics
like energy consumption, negotiation overheads, complexity, reliability, scalability,
and multipath strategies were considered in Reference 7. In Reference 8, scalabil-
ity issues and details of various protocols, merits, and demerits were presented.
Constraints and factors that influence the multimedia delivery, cross-layer design
optimization, and performance evaluation were investigated in real-time multime-
dia WSNs.
In fact, the majority of research attempts in WMSNs are made to achieve
energy-efficient multipath selection. Energy-efficient routing strategies were intro-
duced in Reference 9, whereas the main characteristics of choosing the most
appropriate multipath routing schemes and their classifications were discussed in
Reference 10. Alternative path routing, reliable data transmission, and efficient
resource u­ tilization were the three major criteria in which multipath routing proto-
cols were classified in Reference 11, whereas a different classification based on selec-
tion techniques and traffic distribution mechanisms was presented in Reference 12.
In Reference 13, the Internet, end users, and Internet service providers (ISPs) were
used in multipath routing and provisioning.
This book covers all the aforementioned types of routing protocols while
classifying them based on the multipath selection criteria. It aims at providing
a comprehensive framework spanning the multimedia delivery aspects, traffic
modeling, and real case scenarios in the IoT era. In this framework, the end user
must be able to exchange multimedia traffic based on the required QoS attri-
butes. The WMSN itself must be able to dynamically adapt for the varying net-
work/user conditions and provide fault-tolerant multipath routes for the handled
multimedia traffic as indicated in Reference 14.
Introduction ◾ 3

1.1 Contributions
We propose the multipath routing approach as a key solution for multimedia deliv-
ery in large-scale IoT applications. We assume a WMSN within which super and
light nodes are introduced. These super nodes are able to make delivery decisions
that dynamically adapt to user requirements and network conditions. Toward this
end, our main contributions in this book can be summarized as follows:

1. We start with a comprehensive overview for the multipath routing techniques


and related works in the literature. We identify critical QoS assurance factors
and classify various techniques that can be applied to sensor networks in IoT
applications.
2. We combine the use of light cost-effective WSNs and the smart paradigms in
the IoT, which can potentially support mobile multimedia applications. This
combination supports and resolves key IoT design aspects related to mobility
and multimedia traffic in a cost-effective manner.
3. According to the proposed WMSN architecture, we investigate the data
routing problem while stressing the fault-tolerance design factor. We pro-
pose multipath routing strategies that guarantee the QoS in multimedia
delivery.
4. We propose two heuristic data delivery approaches that could be the most
appropriate for multimedia delivery in this work to cope with the next gen-
eration of IoT trends. The proposed data delivery approaches either help to
choose paths that deliver data with the least delay toward the sink or iden-
tify data delivery paths that are more energy-aware and cost-efficient. Both
approaches have been discussed in detail and have shown promising results.
5. We analyze and quantify the network traffic of the aforementioned WMSN
under realistic operational conditions. Solid mathematical models are recom-
mended and endorsed based on analytical studies.
6. Moreover, we investigate the most appropriate traffic modeling techniques
in the IoT and future Internet era. We characterize data requests based on
content demand ellipse (CDE), focusing on efficient content access and dis-
tribution as opposed to mere communication between data consumers and
publishers.

1.1.1 Book Outline


The rest of this book is organized as follows. In Chapter 2, we delve into an over-
view for the field of multipath routing in the IoT era. In Chapter 3, we provide
the details of our optimized multipath routing approach for WMSNs in the
IoT paradigm. Chapter 4 provides an energy-aware data delivery framework for
safety-inspired multimedia in mobile IoT. Chapter 5 introduces a delay-tolerant
4 ◾ Multimedia-Enabled Sensors in IoT

framework for integrated radio-frequency identification (RFID) and sensor net-


works (RSNs) in the IoT era. Chapter 6 introduces a use case in which unmanned
air vehicle (UAV)-enabled WSNs for multimedia delivery in safety-inspired mobile
IoT is proposed and discussed. In Chapter 7, we assess a common protocol in
WSNs, called the duty-cycled asynchronous X-MAC protocol, for dynamic sensor
networks such as the vehicular network. In Chapter 8, we provide a mobile traffic
model for WMSNs in the IoT. In this chapter, we provide a dynamic model for
IoT-specific paradigms to improve the end user satisfaction. Chapter 9 debates a
generic information-centric framework for traffic modeling and optimization in the
IoT. Finally, we conclude this book in Chapter 10 with potential perspectives on the
proposed work and discuss the directions of future work.

References
1. F. Al-Turjman, Cognition in information-centric sensor networks for IoT applica-
tions: An overview, Annals of Telecommunications Journal, vol. 72, no. 1, pp. 1–16,
2016.
2. M. Z. Hasan, H. Al-Rizzo, and F. Al-Turjman, A survey on multipath routing pro-
tocols for QoS assurances in real-time multimedia wireless sensor networks, IEEE
Communications Surveys and Tutorials, vol. 19, no. 3, pp. 1424–1456, 2017.
3. F. Al-Turjman, Cognitive routing protocol for disaster-inspired internet of things,
Future Generation Computer Systems, 2017. doi: 10.1016/j.future.2017.03.014.
4. N. Al-Karaki and A.E. Kamal, Routing techniques in wireless sensor networks: A
survey, IEEE Wireless Communications, vol. 11, no. 6, pp. 6–28, 2004.
5. A. Boukerche, B. Turgut, N. Aydin, M. Ahmad, L. Bölöni, and D. Turgut, Routing
protocols in ad hoc networks: A survey, Computer Networks, vol. 55, no. 13, pp.
3032–3080, 2011.
6. M. Korkalainen, M. Sallinen, N. Kärkkäinen, and P. Tukeva, Survey of wireless sen-
sor networks simulation tools for demanding applications, in Proceedings of the IEEE
International Conference on Networking and Services, Valencia, Spain, pp. 102–106,
2009.
7. Z. Jin, Y. Jian-Ping, Z. Si-Wang, L. Ya-Ping, and L. Guang, A survey on position-
based routing algorithms in wireless sensor networks, Algorithms, vol. 2, pp. 158–182,
2009.
8. Z. Hamid and F. Hussain, QoS in wireless multimedia sensor networks: A layered
and cross-layered approach, Wireless Personal Communications, vol. 75, no. 1, pp.
729–757, 2014.
9. S. Ehsan and B. Hamdaoui, A Survey on energy-efficient routing techniques with
QoS assurances for wireless multimedia sensor networks, IEEE Communications
Surveys & Tutorials, vol. 14, pp. 265–278, 2012.
10. M. Masdari and M. Tanabi, Multipath routing protocols in wireless sensor networks:
A survey and analysis, International Journal of Future Generation Communication and
Networking, vol. 6, no. 6, pp. 181–192, 2013.
11. M. Radi, B. Dezfouli, K. Abu Bakar, and M. Lee, Multipath routing in wireless sen-
sor networks: Survey and research challenges, Sensors, vol. 12, pp. 650–685, 2012.
Introduction ◾ 5

12. E. David, N. Ndih, and S. Cherkaoui, On enhancing technology coexistence in the


IoT Era: ZigBee and 802.11 case. IEEE Access, vol. 4, pp. 1835–1844, 2016.
13. F. Al-Turjman, Cognitive caching for the future fog networking, Elsevier Pervasive
and Mobile Computing, 2017. doi: 10.1016/j.pmcj.2017.06.004
14. F. Al-Turjman, H. Hassanein, and M. Ibnkahla, Efficient deployment of wire-
less sensor networks targeting environment monitoring applications, Computer
Communications, vol. 36 no. 2, pp. 135–148, 2013.
Chapter 2

A Survey on Multipath
Routing Protocols for QoS
Assurances in Real-Time
Wireless Multimedia
Sensor Networks

2.1 Introduction
The demand for a wide variety of network services and various multimedia appli-
cations has been the major driving force behind the innovation and development
of various networking technologies such as IEEE 802.11n, 4G/5G long-term evo-
lution-advanced (LTE-A), and wireless sensor networks (WSNs) [1]. Integration
of these applications in modern networking poses a new set of constraints on
quality of service (QoS) and often requires suitable routing strategies. Routing
strategies are key for meeting the different demands for network capacity pro-
visioning and QoS guarantees in such networks. Wireless multimedia sensor
networks (WMSNs) have been used in numerous applications in the era of the
Internet of things (IoT) that require ubiquitous access to both real- and non-
real-time applications as depicted in Figure 2.1 [2]. WMSNs utilize the multi-
media sensing technology to monitor changes in the surrounding environment
and route the collected information via routing protocols to remote controlling
units [3]. WMSNs are expected to be among the pillars in realizing the IoT para-
digm by fostering applications such as smart cities, smart agriculture, and smart

7
8 ◾ Multimedia-Enabled Sensors in IoT

Power system
Internet of things
Forest fire Structural health
detection Industrial
Urban internet
control
Water leakage Traffic congestion
Smart environment Smart cities
WMSN
Smart farming applications General tracking

Smart Domotic and home


agriculture automation

Security and Detection of gas


emergency Detection of people

Figure 2.1 Applications of WMSNs in different environments.

security and emergency systems [4]. Certain applications of WMSNs may oper-
ate for several months or even years [5]. The deployment of multimedia devices
depends on the application and the environment [5]. This means that each sensor
node should be modified distinctively according to the application scenarios [6].
For example, in security and emergency applications, low-cost maintenance is
required, whereas urban applications are more apt to risk of environmental inter-
ference [5]. Therefore, the network designer may encounter several conflicting
design aspects, challenges, and factors that influence the design of WMSNs in
these environments. These characteristics, challenges, and factors depend on the
nature of real-time multimedia traffic data flow, such as the application specifics
of QoS requirements, high bandwidth, tolerable end-to-end delay, resource con-
straints, multimedia source coding techniques, cross-layer coupling of function-
ality, and multimedia in-network processing [7].
These characteristics are handled either by modifying existing protocols in
WSNs or by proposing new methodologies such as multi-radio multi-channel
systems, switching between multiple channels, multipath routing, or mixtures
of these methods [42]. In Table 2.1, we provide some popular applications of
WMSNs, proposed multipath routing protocols, and the optimal solutions pro-
posed for different application scenarios and environments. Generally, routing
protocols for real-time applications impose severe demands on different QoS met-
rics such as low delay, high throughput, and high reliability [35]. Therefore, these
characteristics, along with other research design issues such as security, connectiv-
ity, and coverage, have received considerable attention and are most likely to be
considered in the different layers of the communication network protocol stack as
shown in Figure 2.2 [43].
Energy efficiency is a common challenge for WMSNs, given the various energy-
efficient mechanisms developed at the different layers of the network protocol stack
[44]. For example, energy-aware routing protocols in which the available energy in
each node can be used to select the optimal path while satisfying the QoS require-
ments has been proposed [44].
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 9

Table 2.1 Applications of WMSNs


Applications
Applications Scenario Routing Protocol Propose Solution
Smart Power system, RTLD [8], Task Multipath multi-
environment smart farming, allocation for QoS constraints
water leakage, real-time
and forest fire applications [9], RPAR
detection [10], SONS [11]
EQSR [12], RTRR [13], Multipath reliability
ICPSOA [14]. constraints
EADD [15] Data-centric
protocol
SHRP [16] On-demand fashion
Smart cities General SAR [17], SPEED Multipath multi-
geographical [18,19], SPEED-EE [20], QoS constraint
tracking REAR [21], QuESt [22],
applications THVR [23], Disjointed
multipath routing
[24], AMPMCR [25],
IAMVD [26]
MMSPEED [27], Multipath reliability
Re-InForM [28], constraints
NC-RMR [29], EERMR
[30], DMRF [31],
QoSMR [32], MCMP
[33], ECMP [34], PMR
[35], QoSNet [36]
DD [37] Data-centric
protocol
RMDSR [38], On-demand
TinyONDMR [39], Fashion
RMRP [40]
Security and Detection of Routing of high- Multipath multi-
emergency gas, detection priority packets in QoS constraint
of people WSNs [41]

Multipath routing strategy has emerged as the technology of choice in WSNs,


which can fulfill QoS metrics and networks constraints in most real-time applica-
tions [45]. The demands of multipath routing strategy combined with various per-
formance demands dictated by different applications have led to the proposal of a
number of new routing protocols to efficiently utilize the limited available resources
10 ◾ Multimedia-Enabled Sensors in IoT

Connectivity and coverage

Hardware and testbeds

In-network processing Application layer


Resource constraints

Optimizing reliability vs. congestion control

Quality of services
Transport layer

QoS based on real-time data traffic flow Routing layer

Single-channel MAC, time division


MAC layer
multiple access frame optimization

Use of time-hopping impulse radio


Physical layer
ultra wide band

Security

Figure 2.2 Research challenges at the various WMSNs layers of the communi­
cation protocol stack. (Adapted from I. F. Akyildiz et al., Computer Networks,
vol. 51, pp. 921–960, 2007.)

in addition to the features of multimedia data [46]. Moreover, the performance of


traditional sensor nodes should be enhanced to keep with the fast-changing and
realistic events in the real world [35].
Multipath routing operation is relevant for providing adequate network resources
in various traffic conditions in order to fulfill these requirements [47]. Multipath
routing techniques have been extensively used for improving the delivery of multi-
media content, providing fault-tolerance routing and supporting QoS in networks
from multi-hop local area networks (LANs), wireless area networks (WANs) and
the Internet, and, finally, in ad hoc networks and WSNs [48]. Furthermore, the
effectiveness of multipath routing strategies is essential to achieve high-quality net-
work services and guaranteed QoS at high data rates. Consequently, the objective
is to assign more loads to under-utilized multiple paths and fewer loads to over-
committed paths so that uniform resource utilization of all available paths can be
ensured to distribute network traffic [49,50]. Multipath routing protocol is cost-
effective for heavy load balancing structures without dynamic traffic engineering
[1,51]. To exploit multipath routing strategies, it is necessary to determine the num-
ber of available paths. Clearly, the number and quality of the selected paths dictate
the performance of the multipath routing protocol. Thus, designing a multipath
routing algorithm for multimedia applications is very challenging because of sev-
eral characteristics and features of WMSNs. The breakdown of multipath routing
schemes at different taxonomic levels and their features present a complete set of
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 11

network routing protocols and highlight the key challenges of a communication


routing protocol. With this broad picture in mind, we aim at assisting the reader in
understanding what is essential and critical to QoS assurances with the current mul-
tipath strategies in WMSNs. In order to assist the readers, we provide in Table 2.2
a list of abbreviations, along with brief definitions as used throughout the chapter.
The rest of the chapter is organized as follows. Section 2.2 presents a review of
related work. In Section 2.3, we present the system architecture and design issues of
the routing algorithm, whereas in Section 2.4, we present the proposed high-level
WSN multipath-routing protocol taxonomy. In Sections 2.6 and 2.7, state-of-the-
art multipath routing techniques are presented. Section 2.8 provides a comprehen-
sive analysis of major multipath routing approaches for WMSNs. In Section 2.9,
some open issues and research directions are identified. Finally, conclusions are
provided in Section 2.10.

2.2 Comparison with Related Survey Articles


Most of the existing surveys of QoS-enabled WSN routing protocols deal with
application and architecture requirements along with their respective routing strat-
egies [42,46,47,49]. However, there are additional criteria that should be empha-
sized when designing routing protocols for WMSNs. The most important of these
are the lifetime of the multimedia sensor network and the overall QoS requirements
of the respective real-time applications. We will review existing survey articles and
highlight the gaps in those surveys as compared to the focus of this chapter.
Early surveys focused on general architectural issues and open research problems
in the field of WSNs [42,52–60]. These surveys were based on the requirements of real-
time applications and only cover certain aspects of QoS in WSNs, such as reliability,
scalability, and ability to support real-time activities, which are not directly applicable
to the more recent WMSNs. An introductory survey was presented in Reference 59
to discuss constraints and solutions related to each layer of the WSN stack, including
sensor node hardware requirements. Open research challenges were presented for each
layer. General research challenges in terms of operating systems, networking, and mid-
dleware protocol requirements to support various WSN requirements were presented
in Reference 55. In Reference 42, routing protocols for WSNs were discussed and
classified into two main categories: network structure and protocol operations. The
authors outlined the design trade-offs between energy efficiency and communication
overheads in the context of various routing paradigms. In Reference 59, the authors
surveyed 27 WSN routing protocols that were current as of 2004 and presented a
protocol classification based on multipath, query-based, negotiation-based, and QoS-
based approaches. In addition, design issues such as network flow considerations and
quality modeling were highlighted. A taxonomy for WSN routing protocol classifica-
tion was developed in Reference 52, in which the system, network, operational, and
objective models were used to classify various WSN routing protocols.
12 ◾ Multimedia-Enabled Sensors in IoT

Table 2.2 Abbreviations


Abbreviation Name
ADC Analog-to-digital converter
AE Available energy
AMPMCR Adaptive multipath multi-constraint routing
AODV Ad hoc on-demand vector
AOMDV Ad hoc multipath demand vector
ARRCH Adaptive reliable routing based on clustering hierarchy
BE Best-effort
BF Broadcasting flooding
BFS Breadth first search
CC Cell controller
CMOS Complementary metal-oxide semiconductor
CMVT Cross-layer multipath video transmission
CPU Central processing unit
CR Cognitive radio
CRSNs Cognitive radio sensor networks
CSMA/CA Channel sense multi-access/collision avoidance
DD Directed diffusion
DFM Duplication of forwarding messages
DMRF Dynamical jumping real-time fault-tolerant
DPS Dynamic packet state
DSA Dynamic spectrum access
DSR Dynamic source routing
EADD Energy-aware directed diffusion
EBMR Energy-balancing multipath routing
ECCs Error correction codes
ECMP Energy-constrained multipath
EDF Earliest deadline first
EEPBC End-to-End path battery cost
EERMR Energy-efficient reliable multipath routing
EH-WSN Energy-harvesting wireless sensor network
EQSR Energy-efficient and QoS-based routing
EXT Expected transmission count
(Continued)
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 13

Table 2.2 (Continued) Abbreviations


Abbreviation Name
EYES European Youth Environmental Sentinels
FECC Forward error correcting codes
FIFO First in, first out
GBR Gadient-based routing
GPS Global position system
HRT Hard real-time
IAMVD Interference-aware multipath routing for video
ICPSOA Immune cooperative particle swarm optimization algorithm
IFS Iterated function system
IoT Internet of things
ISP Internet service provider
LANs Local area networks
LQEs Link quality estimators
LQI Link quality indicator
LTE-A Long-term evolution-advanced
MAC Medium access control
MANETs Mobile ad hoc netwoks
MCMP Multiple constraint multipath
MDP Markov decision process
MIMO Multiple input and multiple output
MIP Mixed integer programming
MOGA Multi-objective genetic algorithm
MP-DSR Multipath dynamic source routing
NC-RMR Network coding-reliable multipath routing
NSGA-II Nondominated sorting genetic algorithm
OS Operating system
PARSEC Parallel simulation environment for a complex system
PMR Partitioning multipath routing
PPP Packet reception rate
QoS Quality of service
QoSMR Quality of services multipath routing
QoSNet Quality of service network
(Continued)
14 ◾ Multimedia-Enabled Sensors in IoT

Table 2.2 (Continued) Abbreviations


Abbreviation Name
QuESt QoS-based energy-efficient routing
REAR Real-time and energy-aware QoS routing
REER Robust and energy efficient multipath routing
Re-InForM Reliable information forwarding multiple paths
RMA Rate monotonic algorithm
RMDSR Robust multipath dynamic source routing
RMRP Resilient multipath routing protocol
RPAR Real-time power-aware routing
RRP Rumor routing protocol
RSSI Received signal strength indicator
RTLD Real-time with load disturbed routing
RTRR Real-time robust routing
SAR Sequential assignment routing
SEEMR Secure and energy-efficient multipath routing
SHRP Shortest hierarchal routing protocol
SNR Signal-to-noise ratio
soft-E2E soft-end-to-end
SOCEE Source Optimized Control with Energy Efficient
SONS Self-organizing network survivability
SPEED Stateless protocol for real-time communication in sensor networks
SPEED-EE SPEED-energy efficient
SPIN Sensor protocols for information via negotiation
SPIN-BC SPIN-broadcast
SPIN-EC SPIN-energy consumption
SPIN-PP SPIN-point-to-point
SPIN-RL SPIN-reliability
SRT Soft real-time
TCP/IP Transmission control protocol/Internet protocol
THVR Two-hop velocity-based routing
TinyONDMR Tiny optimal node-disjoint multipath routing
WAN Wireless area networks
WFQ Weighted fair queuing
(Continued)
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 15

Table 2.2 (Continued) Abbreviations


Abbreviation Name
WLAN Wireless local area network
MDC Multiple description coding
WMSN Wireless multimedia sensor network
WSN Wireless sensor network
XOR Exclusive OR

Subsequent surveys considered additional issues such as energy efficiency [42],


geographic location [61], network scalability [62], and cross-layer approaches [54].
In Reference 59, the advantages and performances of various approaches for real-
time routing protocols and algorithms for WSNs were highlighted, and cross-layer
design was introduced as a viable design approach. In Reference 54, a survey of vari-
ous MAC and routing protocols for supporting real-time QoS for WSN in terms
of reliability, data aggregation, and cross-layer protocol solutions were presented.
In addition, trade-offs among different constraints in real-time applications, such
as energy efficiency and delay performance, were highlighted. Energy-constrained
routing protocols were studied in Reference 63, in which the taxonomy presented
in Reference 42 was expanded to consider energy efficiency. Existing routing strate-
gies were classified based on four main criteria: network structure, communication
model, topology, and reliable routing approaches.
In Reference 61, position-based routing protocols were surveyed for WSNs.
Flooding-based routing, curve-based routing, grid-based routing, and behavior-
based routing were discussed. Metrics such as energy consumption, negotiation
overheads, complexity, reliability, scalability, and multipath strategies were used for
comparison. Scalability issues were also considered in Reference 64. Hierarchical
routing strategies for large-scale WSNs were classified based on control overheads
and energy consumption. Details of various protocols, as well as advantages and
disadvantages in terms of the message complexity, memory requirements, local-
ization, data aggregation, clustering algorithm, intra-cluster topology, cluster head
selection, and multipath routing strategy were presented in Reference 64.
Another emerging trend in recent surveys is the focus on real-time multimedia data
transmission in WSNs. In Reference 65, the authors surveyed issues related to sup-
porting multimedia communication over WSNs. The design constraints and factors
that influence multimedia delivery over WSNs, solutions appropriate for respective
layers of the networking stack, along with their shortcomings and other major open
research issues, were highlighted. Cross-layer designs for multimedia streaming were
investigated in Reference 66. The paper discussed mechanisms for cross-layer optimi-
zation and outlined future research directions for each layer of the network stack. In
Reference 43, architectures, algorithms, and protocols were proposed for the various
layers of the networking protocol stack, as well as cross-layer designs for WMSNs along
16 ◾ Multimedia-Enabled Sensors in IoT

with an evaluation of the performance of existing hardware and test beds. Multipath
routing techniques were introduced to address QoS requirements for WMSNs.
Energy efficiency and multipath routing approaches are the main focus of current
research efforts in WMSNs. In Reference 47, energy-efficient routing strategies were
introduced for WMSNs, including traditional schedulers (for example, rate mono-
tonic algorithm [RMA] and earliest deadline first [EDF]), along with a discussion of
the performance issues of each routing strategy, energy-efficient routing challenges
for WMSNs, and limitations of current nonmultimedia routing strategies. The ben-
efits of various multipath routing protocols for wireless sensor networks (WSNs) and
their benefits were presented in References 66 and 67. The authors addressed the main
characteristics of the multipath routing schemes and classified them according to
their attributes. In Reference 46, multipath routing protocols were classified based on
three main criteria: alternative path routing, reliable data transmission, and efficient
resource utilization. The multipath routing protocol taxonomy in Reference 50 used
path selection techniques and traffic distribution mechanisms to classify the sur-
veyed protocols. The paper also discussed the suitability of the selected multipath
routing protocols for meeting the performance requirements of various applications.
Other considerations such as swarm intelligence-based routing, geographic awareness
routing, and redundant traffic reduction based on the similarity of multimedia data
sources from nearby locations were studied in Reference 7. However, the authors in
Reference 7 classified the WMSN routing protocols according to the direction of
sensor nodes equipped with multimedia devices. The classification depends on the
routing system architecture and design issues for WMSNs.
The authors in Reference 46 classified a multipath routing taxonomy for WSNs
into three categories, with emphasis on their advantages and disadvantages. The
authors in Reference 68 surveyed multipath routing and provisioning in the Internet,
as well as from the end-user and Internet service provider (ISP) perspective. However,
our survey and the survey presented in Reference 46 have several major differences.
First, our survey is more up-to-date and covers most existing multipath routing pro-
tocols of WMSNs, some of which are not covered in the previous work. Second, our
survey focuses on the importance of particular multipath routing strategies for QoS
assurances of real-time applications. Furthermore, we exploit mechanisms for the mul-
tipath routing strategy that aim not only at circumventing single-point failure but also
to facilitate network provisioning. This investigation provides a comprehensive survey
on both traditional data and more recent real-time multipath routing protocols.
Our survey presents an overall picture of QoS assurances in WMSNs via inves-
tigating and comparing advantages and disadvantages of existing multipath routing
protocols. It should be pointed out that this chapter also provides an analysis of the
classification of multipath routing protocols in terms of the exploitation and design
issues for multipath strategies aimed not only at circumventing the multipath con-
figuration but also at the effectiveness of network provision to network services and
guarantees of QoS parameters. Table 2.3 provides a comparison between our survey
and related survey articles available in the literature. To this end, a new classification
Table 2.3 A Comparison among Related Survey Articles
Considered Performance Metrics

Routing
References Strategies Energy Reliability Scalability Complexity Topology Security Description
[52]a Used system, network, operational,
and objective model metrics to
classify various routing protocols.
[54]a ✓ ✓ ✓ ✓ ✓ Surveyed MAC and routing that
support real-time QoS for WSNs.
[55]a ✓ ✓ ✓ ✓ Surveyed general research challenges
for supporting various WSN
requirements.
[59]a ✓ ✓ ✓ ✓ ✓ ✓ Discussed node hardware
requirements at each WSN stack layer.
[59]a ✓ ✓ ✓ ✓ ✓ ✓ Discussed routing design issues such
as network flow and quality modeling.
[42]a ✓ ✓ ✓ ✓ ✓ ✓ ✓ Discussed trade-offs between energy
efficiency and communication
overhead.
[61]a ✓ ✓ ✓ ✓ ✓ ✓ Surveyed position-based routing
protocols.
[62]a ✓ ✓ ✓ ✓ Surveyed routing protocols
A Survey on Multipath Routing Protocols for QoS Assurances ◾

considering scalability issues.


17

(Continued)
18

Table 2.3 (Continued) A Comparison among Related Survey Articles


Considered Performance Metrics

Routing
References Strategies Energy Reliability Scalability Complexity Topology Security Description
[63]a ✓ ✓ ✓ ✓ ✓ ✓ Expanded taxonomy in Reference 42
to consider energy efficiency.
[64]a ✓ ✓ ✓ ✓ ✓ ✓ Surveyed hierarchical routing
strategies for large-scale WSNs.
[65]b ✓ ✓ Surveyed issues related to supporting
real-time multimedia streaming traffic
without providing analysis of
multipath mechanism.
[66]b ✓ ✓ Discussed mechanism for cross-layer
Multimedia-Enabled Sensors in IoT

design optimization but does not


provide analysis of multipath
mechanism in terms of exploitation
and design.
[43]b ✓ ✓ ✓ ✓ ✓ ✓ Surveyed most routing protocols in
WMSNs, along with introducing
multipath routing mechanism, but
does not provide any analysis of
multipath mechanism in term of
exploitation and design.
(Continued)
Table 2.3 (Continued) A Comparison among Related Survey Articles
Considered Performance Metrics

Routing
References Strategies Energy Reliability Scalability Complexity Topology Security Description
[47]b ✓ ✓ ✓ ✓ ✓ ✓ Introduced energy-efficient routing
strategies.
[66,67]b ✓ ✓ ✓ ✓ ✓ ✓ Discussed the main characteristics of
multipath routing schemes and
provided classification according to
their attributes.
[46]b ✓ ✓ ✓ ✓ Provided classification based on main
criteria but does not focus on the
importance of particular multipath
routing strategies.
[47]b ✓ ✓ ✓ ✓ ✓ ✓ Introduced energy-efficient routing
strategies.
[50]b ✓ ✓ ✓ ✓ ✓ Used path selection techniques and
traffic distribution mechanisms to
classify the various surveyed
protocols.
[7]b ✓ ✓ ✓ ✓ ✓ ✓ Provided classification for WMSN
routing protocols according to the
direction of sensor nodes equipped
A Survey on Multipath Routing Protocols for QoS Assurances ◾

with multimedia devices.


19

(Continued)
20

Table 2.3 (Continued) A Comparison among Related Survey Articles


Considered Performance Metrics

Routing
References Strategies Energy Reliability Scalability Complexity Topology Security Description
[68]b ✓ ✓ ✓ ✓ ✓ Surveyed multipath routing and
provisioning in the Internet, as well as
from the end-user and ISP perspective.
Our ✓ ✓ ✓ ✓ ✓ ✓ Provided an analysis of the
classification of multipath routing
◾ Multimedia-Enabled Sensors in IoT

protocols in terms of the exploitation


and design issues for multipath
strategies.
a Surveys based on the requirements of real-time applications and only cover certain aspects of QoS in WSNs. However, they can
not be directly applied to WMSNs.
b Surveys based on classification of routing protocols. However, they did not cover the importance of particular multipath rout-
ing strategies for QoS assurances of real-time applications in WMSNs.
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 21

is proposed that classifies existing protocols according to the exploitation and selec-
tion of multiple paths. Moreover, a complete description with comparisons among
different WMSN applications are presented in our survey.

2.3 Routing System Architecture and Design Issues


The extension of WMSNs has stretched the horizon of traditional WSNs. WMSNs
allow monitoring and control of real-time information with low-cost and miniatur-
ized sensor nodes. This brings different architectures, design goals, and constraints.
How can WMSNs reach to the derived reliable scalable architectural frame
model to achieve QoS parameters with lower algorithmic complexity? Particularly, a
routing strategy can be embedded into low-cost microprocessors to extend the life-
time of the sensor network without jeopardizing reliable and efficient communica-
tions. Generally, the architecture and design issues for routing protocols in WMSNs
give rise to new challenges. The most dominant and challenging factors to achieve
efficient communication in WMSNs are addressed in the following subsections.

2.3.1 Hard and Soft Real-Time Operation and


Best-Effort for Resource Constraints
Real-time WMSNs consist of three types: hard and soft real-time systems and best-
effort (BE) systems.

2.3.1.1 Hard Real-Time (HRT)


HRT applications involve a mechanism that operates on a preemptive and con-
text-switching operating system (OS) with switching tasks in and out of execution
modes based on the scheduled time slots to maintain real-time characteristics [56].
These applications might require an OS to deterministically delay different tasks.
Traditional schedulers (for example, rate monotonic algorithm (RMA) and earliest
deadline first [EDF]) might be suitable for such applications [56,57].

2.3.1.2 Soft Real-Time (SRT)


SRT applications involve a system that attempts to meet time constraints related to
tasks, operations, and applications by applying system resources, including a high
clock rate, fast processors, cache, and buses [56].

2.3.1.3 Best-Effort (BE)


BE applications are necessary to guarantee the best achievable performance; how-
ever, they may not be relatively focused on static measures such as central process-
ing unit (CPU) allocation.
22 ◾ Multimedia-Enabled Sensors in IoT

Designing routing protocols for WMSNs is influenced by many factors (such as


the cost of hardware, network topology, power consumption, etc.), which all affect
either probabilistic or deterministic end-to-end QoS metrics. When considering
real-time support in WMSNs, energy efficiency should not be ignored, as energy
consumption is the main constraint in WSNs. Therefore, there is often a trade-off
among these constraints [69,70].

2.3.2 Energy Efficiency Model


WMSNs can be used in various applications spanning over an extended physi-
cal area. Therefore, for easier installation, the integration with a deployed wireless
communication network module is a viable option. A large number of nodes are
required for adequate coverage and collection of multimedia information for a pro-
longed duration. Typically, generating higher volumes of data traffic requires not
only extensive processing but also higher transmission rates [47].
Basic components of a typical wireless multimedia node are illustrated in
Figure 2.3 [71], including the sensor module responsible for capturing sensory
information fed to a processing module. A sensor module consists of two sub-com-
ponents: sensory devices for capturing a signal and an analog-to-digital converter
(ADC). The main role of the processing module is to process the sensory data,
encapsulation, and forwarding to other nodes or to a sink via a wireless module.
Multimedia sensors provide audio and video sensing capabilities to make the sensor
node operational in the WMSN environment. These multimedia extensions come

Location finding unit Mobilizer unit

Sensor module Processing module Wireless communication module


Processing unit
Processor Application layer
Sensing

ADC

Storage Transport layer

Routing layer

MAC layer

Physical layer

Power unit

Solar cell Battery Wall outlet

Figure 2.3 Basic components of WMSNs.


A Survey on Multipath Routing Protocols for QoS Assurances ◾ 23

with depicting costs in the shaded unit in Figure 2.3. Because audio and video
units are required, higher energy consumption is required, a larger space for random
memory is expected, and a higher amount of bandwidth may be required due to the
excessive multimedia content [71].
A major component in a sensor node is the power supply unit, which has limited
power resources. Basically, the power unit consists of power managements (solar cell,
battery, wall outlet, and/or capacitor) and an energy sink [72]. A special kind of sen-
sor network, the energy harvesting-wireless sensor network (EH-WSN), in which
each node is equipped with a rechargeable power supply instead of a traditional
battery, has been reported [73–75]. Moreover, a battery is periodically recharged to
keep the node continuously working rather than minimizing energy consumption.
The node lifetime is defined by the time during which the energy level of the node
is above a certain threshold that allows it to perform operations such as transmit-
ting, receiving, or processing data [74]. Therefore, once the energy level drops below
this threshold, the node should go to sleep to charge up; otherwise, the battery is
considered dead and unable to continue the operations [74]. This means that it is
desirable to know that different nodes harvest different amount of energy. However,
this will be a tiresome task or might even be impossible since most WMSNs depend
on deployment and environmental conditions that are needed to recharge the bat-
tery in order to get as much autonomy as possible [44]. Reducing or eliminating
the problem of the limited lifetime will enable the network designers to develop the
functionality of nodes by adding extra features and components [73].
A combination of energy-efficient mechanisms at the different layers in the pro-
tocol stack of WMSNs is used to minimize energy consumption and maximize
the lifetime of the network. Thus, how could these routing protocols maximize the
workload in the energy harvesting network as well as the lifetime of the networks?
The answer to this question is to develop an energy-efficient routing protocol that
is able to balance and optimally allocate available energy (AE) among active nodes.
To improve energy-saving performance, each sensor node can be used to select the
optimal paths based on the variation of environmental conditions [75].
Single-path and multipath routing mechanisms consider the most recent energy-
efficient routing mechanisms according to the energy level of a particular node.
Single-path approaches vary among different selected paths, thereby b­ alancing the
energy depletion among the nodes and extending the network lifetime. Multipath
routing can achieve better balancing than single-path routing since multipath
approaches can achieve load balancing and prevent congestion in the network by
distributing the energy consumption among optimally selected paths [75].

2.3.3 QoS Modeling Requirements


End-to-end QoS requirements using a large number of mechanisms and algorithms
have been pursued in different network stack layers to satisfy QoS parameters
[42,59,60,70,71]. At the same time, different wireless network communications
24 ◾ Multimedia-Enabled Sensors in IoT

may impose specific constraints on supporting the QoS requirements, depending


on their particular characteristics, challenges, and design requirements.
QoS support can be met by a plethora of approaches: for example, data-centric,
protocol-centric, cross-layer optimization, cooperative, and distributed algorithms
[47,76–78]. Our survey focuses on QoS requirements imposed by the applications
on the network because of the proliferation of applications requiring some guar-
antee of services from the network and the performance of applications depends
largely on the QoS assurances in WMSNs [78].
More particularly, how can the underlying network route the QoS-constrained
data while efficiently utilizing network resources? The QoS requirements of
WMSNs may be very different from those of WSNs in terms of delay, jitter, band-
width, and energy consumption.
As a result, some WMSN applications rely on new QoS parameters that prefer
measurement of the transmission of the sensor data in an efficient and effective
manner [79]. Furthermore, network designers should be able to investigate which
system architecture or mechanism can be exploited to provide better services than
the BE services such as differentiated services (soft-QoS) and guaranteed services
(hard-QoS) for the WMSNs applications. Some QoS requirements are briefly
described in the following sub-sections.

2.3.3.1 Latencies
Ensuring stringent timeliness is essential for real-time video applications. Timeliness
can be provided either on a guaranteed or BE basis according to the tolerance level
of the multimedia application [47]. In addition to timeliness, there are several other
factors that cause latencies in WMSNs, including in-network processing, transmis-
sion, and queuing delays [2,43,47]. Thus, reducing delays is a crucial task for real-
time WMSNs applications. Generally, end-to-end latency can be classified into two
classes: soft-latency bounded systems (deterministic constraint) and hard-latency
bounded systems (predictive constraint) [47]. In hard-latency systems, the service
cannot ensure meeting its deadline. Therefore, failure to do so is considered as a
failure of the whole system, whereas the delay of a fraction of data traffic can be
probabilistically guaranteed in soft-latency systems [47].

2.3.3.2 Bandwidth
Multimedia content, especially video streams, requires a high amount of band-
width for transmission. In addition to multimedia transmission, some intermediate
sensor nodes act only as relay nodes in a dynamic topology because of low transmis-
sion range and multi-hop communication. Therefore, the use of single-path rout-
ing strategy for transmission of multimedia may exhaust the energy of the path,
resulting in network failure. Thus, a bandwidth constraint based on low-power
transmission consumption must be handled by the multipath routing strategy [47].
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 25

Generally, QoS support becomes increasingly challenging because of the


increasing desirability of connectivity and exchange of best-quality media informa-
tion at any time, at any location, and in any manner. Therefore, designing effec-
tive routing strategies in WMSNs should be performed for QoS requirements for
different classes of traffic patterns. This remains an open research area and a major
challenge in the WMSNs field [80,81].

2.3.4 Sensor Network


Modeling large-scale sensor node deployment (such as for nodes communicating
over wireless networks with lossy links and without infrastructure) is considered
another challenge in WMSNs research [2,43,47]. The models can be classified
either as deterministic or self-organizing (randomized). In the deterministic model,
the sensor node is deployed, and packets are routed through predetermined paths.
In the self-organizing model, sensor nodes are scattered randomly by creating an
infrastructure in an ad hoc topology [82]. In terms of infrastructure distribution,
the position of the sink, the cluster head, and the scope of the monitored area are
crucial in terms of energy efficiency, performance, and real-time routing protocols.
The architecture of WMSNs consists of three classes: single-layered, flat-layered,
and homogeneous, as shown in Figure 2.4 [47].

Users
IP-based network
Internet

IP-less network Legend

Multimedia
processing hub
Video sensor
Audio sensor
High end
video sensor
Scalar sensor
Wireless
gateway
Storage hub
(a) (b) (c)
Single-tier flat, Single-tier clustered Multi-tier, Sink
homogeneous sensors, heterogeneous sensors, heterogeneous sensors,
distributed processing, centralized processing, distributed processing, Wired link
centralized storage centralized storage distributed storage

Figure 2.4 Architecture of WMSNs. (Adapted from S. Ehsan and B. Hamdaoui,


IEEE Communications Surveys & Tutorials, vol. 14, pp. 265–278, 2012.)
26 ◾ Multimedia-Enabled Sensors in IoT

WMSN architectures support not only heterogeneous nodes but also indepen-
dent applications with different requirements. Thus, it is necessary to develop a
flexible and hierarchical architecture that can accommodate the requirements of
all applications in the same infrastructure [2,83]. Typically, heterogeneous sen-
sory data are reported at different rates to some multimedia applications that may
require a diverse mixture of additional QoS constraints [84].

2.3.5 Data Delivery Model


Data sensing and reporting models depend on the applications for both WSNs and
WMSNs and the time criticality of data reporting. Because various applications
have common requirements, it is sufficient to analyze each application by the classi-
fication of the data-delivery model. Such models are concerned with reporting data
toward the sink along with their corresponding requirements. Generally, data sens-
ing and reporting can be categorized into four basic traffic data delivery models or
services: time-driven (continuous), event-driven, query-driven, and hybrid [80,85].

2.3.5.1 Continuous Time-Driven Delivery Model


This model is suitable for applications that require periodic dispatches of real-time
data for monitoring, such as surveillance or reconnaissance, in which the sensor
node constantly switches between the sensors and transmitters to sense the environ-
ment and transfer data in a periodic order.

2.3.5.2 Event-Driven and Query-Driven Models


These models consider the most commonly used applications in WMSNs. When
a certain event occurs, the sensor nodes react immediately to the changes in the
sensed attribute, or a query is generated by the base station [86].

2.3.5.3 Hybrid Model


This model is used in some networks that use a combination of continuous, event-
driven, and query-driven data delivery based on the conditions of patterns car-
ried out by the traffic and the classification of traffic requirements (for example,
a surveillance application that periodically sends both event-triggered video and
temperature). The design of the routing protocol is highly influenced by the data
delivery model, especially when the design involves minimization of energy con-
sumption and route stability [78].

2.3.6 Dynamic Network


Component-based architecture in the WSNs consists of three main components:
the sensor node, sink, and monitored event [87]. Most researchers consider a
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 27

stationary state of sensor nodes while designing the network architecture, whereas
others consider the mobility status of sensor nodes and even the sink node to
improve network conditions. In addition to energy consumption, real-time routing
stability becomes another issue in routing the message to or from the mobile nodes
and other QoS metrics [88]. The sensing/reporting can be either for dynamic or
static events; for example, dynamic events are target detection/tracking applica-
tions, and forest monitoring for early fire prevention is a static event. Monitoring
a static event allows networks to work in a reactive mode, simply generating traffic
when reporting [59].

2.3.7 Reliability and Fault Tolerance


The concept of reliability concerns the ability to deliver multimedia data to the sink
with minimum packet loss [47]. A different reliability constraint arises and needs to
be imposed depending on the application demands that may require delivering the
packets in a reliable transmission manner to the sink with timeliness and without
loss of packets.
Similar to WSNs, the primary path in WMSNs may not often be able to sup-
port the sensed data because of broken links, among other reasons. Therefore,
establishing the reliability of single-path and multipath routing protocols is
challenging since multimedia applications generate high asymmetric traffic pat-
terns, requiring high transmission rates and extensive processing. This challenge
arises mainly because of the functional model of operational characteristics of
routing protocols for data to reach to the sink. As long as WMSNs are used in
various multimedia applications, this may generate a wide range of traffic pat-
terns. The pattern of data traffic can be either single-hop or multi-hop [47,89].
The pattern can be divided according to the node density of the network or
whether the network supports in-network processing, that is, the number of
transmissions and reception of data packets processes in sensor nodes and/or
whether any processing procedure is applied in the network, into the following
patterns as depicted in Figure 2.5 [63].

2.3.7.1 Local Communication (Node-to-Node)


This type of communication has been proposed for ad hoc networks in which a
node broadcasts its status to the nearest neighbor [80].

2.3.7.2 Point-to-Point (Node-Node)


This pattern is used to send an arbitrary node data packet to an arbitrarily directed
node. This pattern is primarily used in wireless local area network (WLAN) envi-
ronments. Point-to-point routing is not widely used for WSNs because building
28 ◾ Multimedia-Enabled Sensors in IoT

Local communication (Node-to-node) Point-to-point (Node-to-node) Convergence (Node-to-sink)

Divergence (Sink-to-node)

Figure 2.5 The pattern of data traffic in WMSNs. (Adapted from N. A. Pantazis
et al., IEEE Communications Surveys & Tutorials, vol. 15, pp. 551–91, 2013.)

a routing tree does not scale well for sensor nodes that have only small data
memory [90].

2.3.7.3 Convergence (Node-to-Sink)


This pattern needs to be supported to route responses back to the sink. Typically,
this communication is defined as many-to-one (reverse multicast), in which multi-
ple nodes or a single node respond for collection directly or indirectly to the sensed
data packet and transmit it to the sink. This pattern is commonly used for data
collection in WSNs and can also be unicast [90].

2.3.7.4 Divergence (Sink-to-Node)


This pattern needs support routing requests originated from the sink to multiple
nodes. This is a general paradigm including other group-based routing concepts
such as anycast (one-to-many), multicast, and manycast routing, implying that any
node that has the requested data can respond to the query [70]. Particularly, a diver-
gence traffic pattern is used when a sink node or base station sends control messages
or queries to sensor nodes. Several techniques have been proposed to understand
how different routing approaches respond when handling various requests in differ-
ent network environments [42,91–98].
One of these techniques adopts a multipath routing strategy for perform-
ing fault-tolerant routing. The idea is to provide alternative or resilient routes
for guaranteeing a reliable flow of transmission of data packets over multiple
hops [99]: for example, sending copies of the same packets over multipaths to
increase the probability that at least one of the copies reaches the sink with
timelines.
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 29

2.3.8 Summary
Designing a routing scheme for real-time WMSNs is influenced by several fac-
tors, such as cost, transmission medium, network topology, power consump-
tion, etc., which all affect either probabilistic or deterministic end-to-end QoS
metrics guarantees, such as delay, jitter, and throughput. Moreover, WMSNs
have specific requirements that are difficult to fulfill, such as traffic flow of large
bursts of data at a high bit rate. These characteristics need to be addressed,
taking into account hardware, bandwidth, and power limitations of the sen-
sor nodes. Therefore, designing effective routing strategies that achieve all QoS
requirements for different classes of multimedia data traffic remains a major
challenge in WMSNs.

2.4 Routing Techniques in WMSNs Classification


A routing strategy is a key building block in a network protocol stack. It is a sub-
component of the network layer in the sensor network stack layer model and is
central to the proper functioning of any multi-hop communication system [63].
Routing strategy is defined in terms of the process of discovering and selecting
paths in wireless networks from the target area that sends network traffic toward
the sink node [42]. Therefore, the prime role of a routing protocol is to establish a
path between the sources and sink node while keeping track of the path availabil-
ity and facilitating successful transmissions of data along the selected paths [42].
In the following subsections, we will discuss the main classification of routing
techniques of multipath routing protocols in the realm of WMSNs. Different cri-
teria used to classify multipath routing protocols are given. Furthermore, we pro-
vide taxonomies for multipath routing protocols in WMSNs. Generally, there are
three main phases of multipath routing: path discovery, path selection, and path
maintenance. Once the paths are discovered, a routing protocol should decide how
to select a path for sending data. Actually, the discovery, selection, and mainte-
nance of paths depend on some node-specific and/or network-wide metrics such as
QoS requirements, residual-energy budget along the selected path, and more [94].
Discovering and maintaining paths are also impacted by the data traffic, which is
a major issue in designing the architecture of sensor nodes for multimedia appli-
cations. For example, the available buffer sizes, the limitation of resources (e.g.,
energy, memory, and processing power), QoS constraints, distribution of node
density, their connectivity, and unexpected changes in node status during the duty
cycling (e.g., inefficiency or failure) give rise to frequent and unforeseen topology
alterations. Furthermore, these issues are also impacted by the fact that the com-
mon wireless channel is broadcast in nature, such as the control and corruption
of data packets at the physical layer or collision of medium access control (MAC)
protocols [100,101].
30 ◾ Multimedia-Enabled Sensors in IoT

2.4.1 Designing Issues for Multipath Routing


Traditional routing approaches have been developed for cellular networks or
wireless ad hoc networks and thus are not sufficient for WSNs, which are more
demanding than other wireless networks [63]. Many new routing strategies have
been proposed to solve routing problems in traditional WSNs [94–98]. The design
of routing protocols for WMSNs, however, is still an open research area. In addi-
tion to the major issues of designing routing protocols in WSNs, there are new
characteristics and constraints due to the nature of multimedia content that must
be handled over the network such that routing protocols for WSNs are not appli-
cable to WMSNs.
The most recent work seeks to handle these characteristics and their design
challenges to solve the routing problem of streaming real-time multimedia content
by modifying previous routing protocols in WSNs (for example, using multiple
performance metrics to meet the additional QoS requirements) [42,102]. New solu-
tions are proposed based on various new methodologies: for example, using mul-
tiradio, multichannel, or multiple-input and multiple-output (MIMO) systems,
switching between multiple channels, selecting multipath routing, or a mixture of
these methods [2,42]. Other approaches utilize optimization for cross-layer design
between multimedia source coding techniques at the application layer and the rout-
ing layer to exploit optimal multipath selection or in-network processing [42]. In
addition, the cross-layer design between the MAC layer and routing allows packet-
level service differentiation or priority-based scheduling and more power-efficient
routing mechanisms [50].

2.4.2 The Taxonomy of Multipath Routing Techniques


A plethora of research on taxonomies for mobile ad hoc networks (MANETs)
and WSNs has reported the baseline model of routing protocols [7,103]. A rout-
ing protocol can be classified into one of four main parallel schemes, and each
scheme is used to classify the routing protocol according to topology communica-
tion model, network structure, and reliable routing [59]. Some routing protocols
adhere to QoS requirements in order to handle various types of multimedia and
mixed traffic [104]. Designing multipath routing protocols for WMSNs requires
additional research that extends the protocols proposed for WSNs [7]. Indeed,
there are several ways to classify multipath routing protocols, depending on dif-
ferent parameters and factors such as reliability, latency, bandwidth, and load
balancing [105]. One of these classifications is based on the number and type
of QoS constraints of paths. The classes and sub-classes of classification are not
mutually exclusive because many protocols belong to more than one class or sub-
class [105]. Hence, the main division of taxonomy is based on equal assignment
of different functionality to the sensor nodes.
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 31

Routing protocols

Topology-based Communication model Network structure Reliable routing

Multipath-based protocols QoS-based protocols

Concurrent multipath Alternative multipath


unicast forwarding broadcast flooding

Multipath, multi-QoS Multipath reliability constraints Data-centric protocol On-demand fashion


constraints for efficiency for reliable data transmission operation

SAR SPEED RTLD MMSPEED Re-InForM ECMP MCMP

Figure 2.6 The taxonomy of multipath routing protocols.

Reliability has emerged as a factor to classify multipath routing protocols, since a


reliable monitoring of the environment is very important in a variety of multimedia
applications [106]. However, we use the reliability factor to classify the routing pro-
tocols according to the QoS requirements of multimedia applications. As depicted
in Figure 2.6, multipath routing protocols can be classified into two main catego-
ries: concurrent multipath unicast forwarding and alternative multipath broadcast
flooding. The first approach is based on sending multiple copies of the same data
packets over multipaths, which increases the accuracy of various tracking in a variety
of civilian and military applications [106]. The second approach is usually imple-
mented by collecting or distributing information to all nodes in a wireless domain
by performing a broadcast operation. In our classification, the decision to route data
over multipaths depends on the mechanism of discovering and selecting a reliable
path to meet certain objectives, such as congestion minimization, and application
constraints (throughput, delay, and bandwidth), as shown in Figure 2.7. In particu-
lar, path selection mechanisms for traffic flow to reduce the overhead and complexity
of the multipath routing scheme and to accommodate constraints on the selected
paths are emphasized [68]. Therefore, the classification of multipath routing pro-
tocols is based on the way routing paths are established during the path discovery
phase and the way routing paths are selected to distribute the traffic under specific
constraints. To successfully achieve the application’s requirements, multipath rout-
ing protocols may use the characteristics of the two methods of classification to gov-
ern the performance of WMSNs in terms of QoS parameters and energy efficiency.
32 ◾ Multimedia-Enabled Sensors in IoT

Partially disjoint
paths

Idealized braided

Multipath discovery Braided multipath Localized braided

Perfect braided Idealized algorithm

Disjointed paths Node-disjoint

Localized k-disjoint
Edge-disjoint paths multipath
Multipath forwarding

Zone-disjoint

Path selection Round-robin selector

Packet-info-based

Maintenance of paths Traffic-condition-based

Network condition
selector
Distribution traffic
splitting pattern
Packet level

Flow level

Figure 2.7 Taxonomy for the design of a multipath routing protocol.

Both taxonomies are complementary and suitable for meeting the requirements
of WMSNs. They describe the determination of the availability of multiple paths as
well as the distribution of traffic under multi-constraints and reliability. Figure 2.6
depicts the breakdown of the second level of the taxonomy of multipath routing
protocols that will be enumerated in the subsequent sections.

2.4.3 Summary
Routing governs the performance of WMSNs in terms of QoS metrics. The design
of a routing algorithm depends on several parameters that must accommodate mul-
tiple conflicting objectives and constraints, imposed by strategies on technologies
and user requirements. Multipath routing is a promising strategy for achieving QoS
metrics such as bandwidth, delay, and throughput as well as demands of the users.
Several WMSN routing protocols have been proposed based on classifications that
depend on different factors and on the users’ requirements. We use the reliability
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 33

factor to classify such routing protocols and to describe the benefits of multipath
routing protocols according to QoS requirements in multimedia applications.

2.5 Multipath Routing Protocols: Challenges and Issues


Multimedia applications encompass monitoring of real-life events, which necessitates
that efficient multipath routing mechanisms should be developed for the transmis-
sion of information while meeting QoS requirements [104]. The QoS requirements
are expressed as a combination of QoS parameters of a multipath [107]. Thus, dif-
ferent multimedia applications have different QoS requirements, and multipath
routing protocols in WSNs or WMSNs have their own unique advantages and dis-
advantages [94–96]. Consequently, it is a challenge to find or design an appropri-
ate protocol or a class of protocols that fulfill all QoS requirements of an efficient
routing protocol [35]. Furthermore, real-time multimedia applications encounter
an additional challenge for energy-efficient multimedia processing [105]. Certainly,
these challenges include optimal routing to meet the dynamic network constraints
and application-specific QoS guarantees. In response to this challenge, particular
QoS requirements require information of the current status of the network as well
as resource constraints, since routing decisions are made based on this information
[68]. For example, in certain applications, a multipath routing algorithm may select
a set of node-disjoint paths with relaxation of node disjointness. This is referred to
as a set of partially disjoint paths, or as a braided multipath that is link disjoint. For
each node in the primary path, an optimally selected alternative path may not have
any computed node in common. However, these kinds of alternative paths could
potentially have comparable latency to the primary path and could therefore expend,
more or less, the same amount of energy as the primary path [68]. Nevertheless, both
link/node-disjoint paths improve reliability and offer more aggregate bandwidth
than nondisjoint due to bottleneck link/node failure for nondisjoint paths, which
negatively impacts the performance of multipath routing [106].
Many researchers seek to discover and select reliable paths by finding node-dis-
joint as well as link-disjoint paths. These algorithms do not take into consideration
QoS parameters. However, successfully meeting certain objectives like QoS require-
ments and congestion minimization might require a global view of the network
topology as well as its resources [108]. It should be noted that they are difficult to
measure and generally can generate high overhead in networks [94], since resource
availability information is periodically exchanged among nodes [68]. Other mul-
tipath routing protocols can aid in saving batteries by distributing network traffic
uniformly among the sensor nodes; however, they increase the delay per packet
transmission along the longer selected paths [109]. Many multipath constructing
strategies have been proposed in the literature [94–96] to describe the performance
of multipath routing protocols by the number and quality of selected paths. These
constructions can be broadly classified into multipath discovery, forwarding, and
34 ◾ Multimedia-Enabled Sensors in IoT

maintenance [42]. Figure 2.7 illustrates the components and issues related to con-
structing multipath routes under the three categories discussed in this survey.

2.5.1 Multipath Discovery


As mentioned in Section 2.3.7, the pattern of data traffic in WSNs can be either
single-hop or multihop. Therefore, the main task in multi-hop communication is to
determine a set of intermediate nodes that should be selected [49].
Figure 2.8 illustrates the construction of partially disjoint paths and a braided
multipath after the route discovery mechanism to create several paths from the
source to the sink.
Obviously, designing a multipath routing protocol is more challenging than
designing a single-path routing protocol due to the difficulty in finding the number
of paths with a desired property in an effective and efficient manner. Furthermore,
nondisjoint or single paths are easy to discover due to the absence of constraints
on common nodes and links with any loop-free paths. Typically, this construction
depends on many parameters to make the right routing decisions. Among these
parameters is the extension of the unicast path to the multipath unicast path, which
in turn depends on the amount of path disjointedness and is considered the main
criterion used to utilize existing routing multipaths [46,94–96]. The performance
of multipath discovery mechanisms depends on the number and quality of dis-
covered paths, which in turn depends on the availability of network resources at
intermediate paths, characteristics of paths, and QoS requirements [68]. There are

(a) (b)

(c) (d)
B

A C D
F

Figure 2.8 Construction of partially disjoint paths and braided multipath.


(a) Fishbone structure, (b) Idealized, (c) Localized, (d) Perfect braided.
A Survey on Multipath Routing Protocols for QoS Assurances ◾ 35

various multipath routing discovery mechanisms, as illustrated in Figure 2.8 and


summarized below.

2.5.1.1 Partially Disjoint Paths


This mechanism is used to improve single-path routing protocols by providing a
group of multiple alternative paths that are not simultaneously used in a large dense
area [97,98]. Basically, the discovery mechanism takes advantage of the broadcast
nature in which overhead packets of nodes are transmitted by their neighbors. Once
the overheard packets arrive, a unicast route reply message is transmitted by each
node’s neighbor and records that neighbor as the next hop to the sink in its alternate
route table [97,98]. Therefore, the multipath routing mechanism is able to discover
a primary path with an alternate path that looks like a fishbone (see Figure 2.8a).
The pattern of data traffic selects one of these discovered multiple paths at a time,
whereas others are kept as a backup in case the path used becomes broken. However,
if all discovered multiple paths are broken, then a new multipath discovery proce-
dure is initiated. Typically, this mechanism is used to improve the reliability of data
delivery and helps to reduce the overheard communication and end-to-end delay.

2.5.1.2 Braided Multipath


Also referred to as a “meshed multipath” [108], this mechanism is slightly different
than the partially disjoint paths in that it increases resilience to node failure along
with longer alternate node-disjoint paths. Because of some attractive resilience of
partially disjoint path properties, more energy can be expended than that expended
on the primary path.
Generally, there exist many possible definitions for data dissemination in the
braided multipath mechanism. Therefore, a constructive definition for the braided
multipath mechanism follows the directed diffusion paradigm that can be divided
into three techniques [108]:

1. Idealized braided: For each node on the primary path, the gradient is com-
puted to determine the preferred neighbor to find the best path from the
source to the sink that does not contain that node. This results in a set of
braided paths that lie either on the primary path or geographically close to
the primary path as illustrated in Figure 2.8b.
2. Localized braided: The braided multipath technique relaxes the requirements
like the idealized braided algorithm for node disjointedness of the complete
multipath [108]. This technique uses two types of path reinforcement mes-
sages instead of finding a small number of alternate paths that depend on
some localized techniques to construct braids at each node along the primary
path. This technique can be briefly described as follows. The procedure ini-
tializes the sink to originate a primary path reinforcement message to the
Another Random Document on
Scribd Without Any Related Topics
Majd újra elhallgatott, leült a karosszékbe és maga elé bámult,
tétován, csaknem ostobán.
A leány odament hozzá és megcsókolta homlokát.
– Ez a pusztulás, – mondta Ibolya halkan. – Én is érzem, tudom,
első perctől fogva, hogy egymásra akadtunk.
VI.
A napok csöndesen teltek a barátságos házban, amely körül már
zsendülni kezdtek a gesztenyefák és ákácok.
Május volt, mikor egy nap búcsu-estére gyűltek össze a szokott
vendégek. Kardos Samu, Farkas bojtárja befejezte jogi tanulmányait
s ment vissza vidékre, a szülővárosába, hogy ott ügyvédi irodát
nyisson. A kis bojtárból valóságos, hites ügyvéd lett, akármilyen
furcsának is tetszett. Ezt az alkalmat a principális nem mulaszthatta
el s bár fájt neki, hogy a szorgalmas és szolgálatkész segédet
elveszítette, méltóan akarta elbúcsúztatni attól a háztól, amelyet az
négy év óta kutyahűséggel szolgált. A szegény fiút, aki csupasz
képével, rossz cipőivel, kabátjával, félszeg modorával éppen nem
árulta el a leendő jogtanácsost, a figyelem mód nélkül meghatotta.
Hiszen tudta, hogy dr. Farkas jó embere, azonban arra, hogy külön
ünnepség rendeztessék részére, amelyen mintegy az ő tiszteletére
vendégek is összegyülnek, nem volt elkészülve. Nem is tudta, mivel
érdemelte ki, hogy egy ilyen úri család ennyire megemlékezzék róla?
Jelentéktelenségét ő érezte a legjobban és eddig nem is volt senki,
aki ebben való hitét megingatta volna. Ám ha már szívesek voltak
iránta, ő is figyelmes akart lenni, virágot vett a hölgyeknek s két nap
óta jól megcirkalmozott búcsúzó-beszédet magolt, amely azonban
sehogy sem akart a fejébe menni, mert a dikció közepén az emlékek
felsorolásánál a jó fiu mindig elérzékenyedett s nem tudott tovább
jutni.
Farkas, aki mindenkinek szíve szerint tudott gondolkodni, a búcsu
emlékére egy vadonatúj fekete kabátot vett Samunak. Titokban vitte
neki a lakására az ajándékot, kezet fogott segédjével s arra kérte, ne
értse félre s ne utasítsa vissza a kis megemlékezést. S ezzel
átnyujtott egy csomagot, amelyből egy selyemhajtókás, ünnepélyes
Ferenc József-kabát bujt elő.
Kardos mód nélkül meg volt hatva és percekig nem jutott szóhoz.
Egy fekete kabát! A saját fekete kabátja! Jelképe a
komolyságnak, az ünnepélyességnek, a hivatásnak! Olyan kabát,
amilyen a főispánnak, a királyi ügyésznek, a képviselőnek van,
akikkel tehát egyenrangu, legalább is viselet dolgában! És ez a
dicsőség most már nem muló természetü, reggelre a rokkot nem kell
visszaküldeni Spitz Rafaelnek, hanem megtarthatja, felhúzhatja,
amikor akarja és főleg – nekitámaszkodhatik annak, aminek akar!
Különösen ez töltötte el lelkét önérzettel, sőt büszkeséggel. Lelkében
még visszazsongott az a sok aggodalom, amelyet a Spitz Rafael
hosszu ködmöneiben kerítéseknél, rossz ülésü omnibuszokban
kiállott, örökké félve, hogy megakad valami drótban s amikor
egyszer egy szögben tényleg lyukat szakított a gérokkján, egész éjjel
foltozni kellett a posztót, nehogy a sólyomszemü ruhakölcsönző
észrevegyen valamit… De most már nem kell félnie, megvan a saját
fekete kabátja és azt tehet vele, ami az eszébe jut, senkinek se
tartozik felelősséggel!
Annyira elérzékenyedett, hogy majdnem kezet csókolt Farkasnak.
Egy elérhetetlennek látszó álma teljesült s ez könnyet csalt a
szemébe.
– Jó, jó, – dörmögte a fiskális, búcsút véve, mert ő is megindult
egy kissé, – a fő, hogy az estéről meg ne feledkezzen! Viszontlátásra
Samukám… azaz ügyvéd úr!… Kolléga úr! – tette hozzá, ragyogva az
örömtől, hogy másnak örömet szerezhet.
Samu az első volt, aki délután a barátságos házba beállított. Az új
kabát pompásan illett rája, egy ráncot nem vetett, akár csak oda lett
volna öntve, a fiu szinte megnőtt, megkomolyodott benne. A leányok
mosolyogva fogták körül az új gavallért és össze-vissza kérdezték
tőle:
– Hát csakugyan elhágy bennünket?
– Gondol-e majd ránk, Samuka?
– De nagy úr lesz most belőle! Talán bizony sohase is keres föl
már!
– Most már aztán asszony után is kell ám nézni!
A jó fiu csak állt köztük, bárgyún, tágra nyilt szemmel és nem
tudott szólani. A szíve tele volt örömmel, hogy végre elvégezhette
tanulmányait s a diploma a zsebében van, de, ha eszébe jutott, hogy
a barátságos háztól most már meg kell válnia, elkomorodott. A szíve
fájt, elgondolva, hogy többé nem mehet Farkas helyett a
Csepelszigetre, Horányba, Szentendrére foglalni, hogy nem csókolhat
kezet az ügyvédnének, akinek karja oly illatos volt, mint az orgona-
virág, hogy nem nézheti titkos elragadtatással Hortenset és Ibolyát,
akiknek képe ott ragyogott állandóan a lelke közepében. Vajjon
hogyan fogja elbírni mindezt, hogyan nélkülözheti tintafoltos
íróasztalát, amelynek minden zúgából titkos szerelmének képei
mosolyogtak rája, hogyan lesz el a kis kert nélkül, amelynek minden
virágos bokra, rózsafája csak azért volt, hogy a leányok és az
asszony szépségét emelje, hogyan mond le a lassan feketedő,
rézcsatornás, barátságos házról, amely életében először a tiszta
örömet, a megnyugvást, a boldogságot, és a reménységet
jelentette? És amíg a leányok közt állott s azok vígan csacsogtak
körülötte becsületes szemében egy nagy könnycsepp jelent meg és
lassan gurult végig az arcán és az új Ferenc József-kabátján.
A leányok egy pillanatig elhallgattak, megérezték, hogy tilos
helyre tévedtek és a vidámságuk a lelkébe vág ennek a szegény
fiúnak, akinek ez a könnycsepp volt az egyetlen vallomása ama
hosszu idő alatt, amelyet házukban töltött. Soha komolyan nem
vették, soha a bensejére kiváncsiak nem voltak, de e percben valami
azt súgta a két leánynak, hogy félreismerték és lekicsinyelték Samut,
könnyelműen szaladtak el a lelke fölött, amelyből most feléjök
röppent egy titkolt sóhaj. Csönd támadt, ahogy elhallgat a gyermek,
amelyik vesszőjével kimélet nélkül, vígan kacagva csapdossa az erdő
lombjait, mikor egyszerre szemben találja magát egy szentképpel,
amely komolyan, fájdalmasan néz le reá a tölgyfa oldaláról.
Farkas szakította félbe a némaságot.
– Mit szóltok a pompás kabáthoz, – mondta, gyorsan kitérve egy
esetleges nyilvános hálálkodásnak, – amelyet a kolléga úr vett
magának? Ez aztán valami? hm? A tiszti főügyésznek se lehet
különb, de talán még a miniszternek sem!
A leányok kaptak a fordulaton, megcsodálták és kidicsérték az új
kabátot s ezzel másra terelték át a beszédet. Samu is érezte, hogy
nem szabad nagyképűnek és savanyúnak lennie, s azért jókedvűnek
iparkodott mutatkozni. Ő mindig és gyorsan tudta alárendelni
akaratát másnak. Elhagyta hát a búsulást és büszkén forgott, sétált
föl-alá az új rokkjában, aztán mosolyra erőltetve ajkát, válaszolt a
kérdésekre. Meg kellett igérnie, hogy minden héten legalább egy
levelet ír az ügyvédnek s karácsonyra, ha törik-szakad, fölkeresi őket
Pesten. Farkasék is hasonló fogadalmat tettek s tervbe vették, hogy
még a nyár folyamán felrándulnak Ungba, ahol Kardos ügyvédi
irodát nyit. Samu majd a vasútnál várja őket, körülkocsikáztatja a
városon, bemutatja annak nevezetességeit s végül megvendégeli
őket a saját házában!
A hangulat lassan vígabbra fordult, különösen, hogy a többi
vendég is beérkezett. Kelemen egy nagy okmány-tárcával lepte meg
Samut, akinek azonnal hóna alá kellett venni a széles portékát,
meglátni, hogyan illenek össze. A próba jól sikerült: a fiu alakja e
komoly fiskálisi tényező révén megnőtt és megtekintélyesedett.
Bükkös egy kötet Heine-verssel ajándékozta meg barátját, akit csak
azért nem szeretett valami nagyon, mert mindig egyforma utakon
járt és egy élő előjegyzési naptárhoz hasonlított, ahová minden óra
rendeltetése előre volt bejegyezve. A versek, úgy vélte, majd kissé
üdébbé és könnyelműbbé teszik. Még Szakáll is gratulált: de ő már
irigyen, neki fájt, hogy más is tanul, iparkodik s megelőzi, hacsak két
lépéssel is az életben. Az ügyvéd annál melegebb kivánságokkal
bocsájtotta útnak, sőt még a kedvenc pipáját is reá erőszakolta.
– Vidéki ügyvéd nem lehet el pipa nélkül, amice, – mondotta, –
ez tekintélyt ád, bizalmat kelt és a szunyogok ellen is megvéd!
– A vacsora! – szólt be a szép asszony a dohányzóba s azzal a
jókedvü társaság áttódult a tágas ebédlőbe.
Pompás búcsúlakoma várt Samura, akinek minden tiltakozása
dacára a főhelyet kellett elfoglalnia. Ami még soha nem történt meg
vele, az ma bekövetkezett: még bort is kellett innia, hogy a ház
tagjaival és a tiszteletére összegyülekezett vendégekkel
koccinthasson. Később beszédek is kerekedtek, az ügyvéd, Kelemen,
Bükkös egymásután keltek föl és köszöntőket mondottak az új
Justinianusra. Majd őt kivánta a társaság hallani. Mit tehetett,
fölállott s belekezdett nevezetes beszédébe, amely ott rejtőzött a
fekete kabátja szárnyaiban. Nagy figyelemmel hallgatták, a közepéig
el is jutott, de itt elfogta az érzékenység s a búcsúzást félbe kellett
szakítania, mert különben menten elpityeredik.
– Éljen! éljen! – harsogták Farkas és Bükkös, látva, hogy mindjárt
baj lesz s ezzel a beszéd point nélkül be is végződött. Senkisem
haragudott érte, a fiúk körülsereglették az ifju ügyvédet és
koccintottak vele, ő pedig elkeseredésében fenékig ürítette ki a
díszserleget. Ez nagy adag volt, pillanatra szédülni kezdett tőle, de
aztán csakhamar magához tért, jókedve kerekedett és ok nélkül
mindenen kacagott. A nevetése furcsa hangokból állott: mintha
vascsöveket kopogtattak volna meg, amelyek tompán, szaggatottan
feleltek s ez annyira mulattatta a nagy társaságot, hogy egyre azon
voltak, hogy Samu kacagjon. És Samu készségesen
engedelmeskedett, kacagott, nem szégyelte magát, örült, hogy
mulattathatja a hölgyeket és urakat és csöppet sem sértődött meg,
mikor a többiek sorra próbálták sajátos nevetését utánozni.
– Nem jó, nem jó! – kiáltott ilyenkor. – Igy csak én tudok
kacagni… ha! ha! ha! Igy van jól!
És mindnyájan tapsoltak és nevettek, hogy a könnyeik csurogni
kezdtek.
Vacsora után kiszállították az asztalokat, félretolták a székeket és
táncolni kezdtek. Samunak is táncolni kellett, bár sejtelme se volt a
mulatásnak erről a neméről. De nem akarta a tréfát elrontani, azért
vígan ugrált jobbra-balra, mint a bakkecske, közbe nagyokat botlott
s egyszerre – zsupsz! – elcsúszott, elvágódott és a zongora alá
gurúlt. Nagy nevetés közt húzták ki, ő is kacagott, mikor egyszerre
tekintete kabátjára esett és észrevette, hogy a karja végighasadt!
Első pillanatra rémület vett rajta erőt, szeme előtt fölvillant a
szigoru Spitz Rafael alakja, de a következő percben már
kiegyenesedett, büszkén nézett körül és így szólt:
– Az enyém! Most már elszakadhat, mert egyedül az enyém!
És szívét a tulajdonjog gőgös érzése először hevítette, lelke előtt
a vagyon ereje ekkor villant fel első ízben: egy kettéhasadt Ferenc
József-kabát ujjainak alakjában!
Igy telt el a búcsúzás estéje.
Hajnal felé a társaság útra kelt. Samu kezet csókolt a
kisasszonyoknak, mély köszönetet mondott az ügyvédnek s végül
Laura asszony elé járult.
– Isten megáldja! – nevetett rá a még mindig szép asszony és
aztán egyet gondolva, megcsókolta a zsidó fiu homlokát.
– Éljen! éljen! – nevetett, tapsolt a fiatalság.
Samuval forgolódni kezdett a világ. Soha őt még asszony, az
anyján kívül meg nem csókolta: a homloka égett, a szíve ki akart
ugrani helyéből, azt hitte, hogy fölemelkedik az égbe, de csak ezt
tudta mondani:
– Köszönöm, köszönöm… Kérem.
Még a hűvös kertben is ezt hajtogatta, még az utcán is és bár
sötét volt körülötte, azt hitte, hogy napfényes mezőn jár.
– Köszönöm, köszönöm – motyogta, míg a többiek egymást
lökdösve, elfojtottan kacagtak.
Kelemen, Bükkös és a többiek elkísérték az Andrássy-útig, ott
elváltak tőle és nyugodalmas jóéjszakát kívántak neki.
De Samu nem feküdt le. Szeméből kiszállt az álom, a vére
lüktetett a szokatlan mulatságtól. Úgy gondolta, hogy még jó lesz
egy kicsit a friss levegőn maradni.
Egy ideig ott állt a széles út szögletén, elnézte a sarki rendőr
mozdulatlan alakját, aztán a kis mellékúton kopott házáig baktatott,
majd visszafordult, ahonnan jött és lassan, nagyokat botolva
visszament a barátságos ház elé, hogy még egyszer titokban, némán
búcsút vegyen tőle.
Rideg és néma volt minden, nyirkos tavaszi éjszaka fogta körül a
fákat, a szél egyet-egyet lökött a zöldülő lombokon. A barátságos
házban egyik lámpa a másik után aludt el, a cselédszoba lakói is már
nyugovóra tértek. Az égbolt peremén keskeny, alig látható szürke
sáv tünt föl.
Szemben a rácsoskapuval, a túlsó oldalon kis pad állott. A
fiatalember arra ült le, kalapját maga mellé tette és meredt szemét a
sötét házra szögezte. Csak bámult, bámult maga elé, visszaidézte azt
a sok lepörgött napot, amelyet e falak közt eltöltött és egyszerre
azon vette észre magát, hogy szemeiből megindulnak a könnyek és
patakként áztatják az arcát. Nem tett ellenök semmit, átengedte
magát az érzéseinek – hiszen most már ember volt, a maga ura, aki
megengedheti, hogy érzései legyenek.
Egy óráig ült így némán, az árnyék által eltakarva; – a házban
már sehol egy gyertya sem égett. Samu feje kábult volt, a szeme
káprázott. Néha-néha pillái le is csukódtak, de aztán a szél
megzörgette a fákat, ez a zaj fölébresztette. Ilyenkor szeretettel
nézett az előtte elterülő sötétségre és elmosolyodott. Egyszerre
aztán erősebb neszre riadt föl: a barátságos ház kertajtaja kinyilt s
valaki kilépett rajta.
Samu meghökkent: nem volt bátor természet s azért első
pillanatban igen megijedt. Istenem, ha most valaki leütné, soha
senkise tudná meg, hogy ki volt a tettes… Csöndesen odalapult a
padhoz és elővette a tollkését, hogy legalább némi védelme legyen.
Vacogó fogakkal vetett kémlő tekintetet a kisurranó alakra, amely
gyors léptekkel fordult balra.
Férfi volt; amennyire a sűrü homályban látni lehetett: magas és
izmos. A padon ülőnek pillanatra úgy rémlett, hogy Bükkös volt.
– Te vagy az, Pál? – kérdezte Samu szívdobogva, alig hallhatóan.
Nem kapott választ, úgy látszik a szél elkapta szavait.
Az éjjeli vándor alakja gyorsan suhant tova.
– Pál, te vagy az?! – szólt utána a remegő Samu még egyszer.
Semmi válasz. Újra csönd lett. A rejtélyes alaknak már a lépései
sem hallatszottak. Kardos lassan magához tért és eszébe jutott, hogy
Bükkös vele együtt ment el a nyaralóból, épp úgy, mint a többi
vendég, ő tehát nem lehetett az éjjeli sétáló, hacsak, miután a
sarkon egymástól elbúcsúztak, vissza nem fordult. De miért tért
volna vissza?
– Talán a kertész volt, – gondolta el magában a bojtár. – Vagy
talán, meglehet, csak álmodtam… Sokat is ittam a nehéz borból, –
tette hozzá aztán és újra rámeresztette tekintetét a barátságos
házra, amely csöndesen, békésen, becsületesen nyugodott az öreg
platánok karjai között.
VII.
A szép tavaszi reggeleket Szakáll arra használta volt föl, hogy a
kertben föl- és alájárva tanuljon.
Ő volt az első, aki a házban fölkelt s munkához látott. Százszor és
százszor sétálta körül reggeliig a kert kavicsos utait, a jogi tételeket
raktározva be mindig fogékony fejébe. Foglalatoskodását nem
zavarta senkisem, a barátságos ház lakói, az igazat megvallva, nem
is nagyon törődtek vele, hogy Szakáll már hajnalban könyvvel
kezében jár-kel, nem is irigyelték érte, de valami nagyon meg se
csodálták. Csak egy lélek volt, akiben ez az örök szorgalom és
állandó munkálkodás tiszteletet ébresztett s hónapok óta egyre
nagyobb rokonszenvet keltett: a kis Hortense.
Miután messze állott attól, amit tudásnak és komolyságnak
neveznek, lelke kétszeresen nagyra becsülte, sőt csodálta a
fiatalembert, aki merőben más volt, mint bárki a házban. Apja
mindig arra tanította Hortenset, hogy nyilt szívvel közeledjen
mindenkihez; így cselekedett Szakállal is. Föntartás nélkül bámulta
szorgalmát, kitartását, akaraterejét és ezzel együtt okos, hideg
szemét, gondolkodó homlokát, amelyet másnak, szebbnek,
tartalmasabbnak látott, mint a ház többi lakója. Szakállt senkise
szerette igazában az épületben s ez még közelebb hozta alakját
Hortense szívéhez. A kis leány üldözött embert, az elhagyottat látta
benne, s úgy gondolta, hogy neki kell kipótolnia azt a gyöngeséget,
amellyel a többiek a fiatal jogásznak adósak maradtak. Már az első
perctől kezdve, hogy Szakáll elmagyarázta neki a tudás, az iparkodás
köznapi tételeit, hideg, de következetes pályáját, rokonszenvvel
fordult ez új eszmekört jelentő ember felé s ez a ragaszkodás az idők
folyamán egyre nőtt, mert Szakáll nemcsak beszélt, hanem
cselekedett is; vastürelemmel és kitartással építette élete egyenes
falait s egyre magasabban járt a fogalmak és tételek birodalmában.
Ha Szakáll nem volt a társaságban, a többiek rendszerint
gunyolódni szoktak volt fölötte és céltáblául vették tudás-szomját,
amelyet akarnokságnak minősítettek. Hortensenek sokszor fájtak e
támadások, de nem felelt rájuk semmit, csak két érzés fejlődött ki
benne révükön: az egyik a csöndes harag azok ellen, akik szerinte
méltatlanul bántották a szegény iparkodó jogászt, a másik, ezzel
kapcsolatban, a szeretet, amely titokban, lassan, de egyre
színesebben fejlődött ki lelkében Szakáll iránt. Ezt az érzését – amely
erősen hasonlított az intézeti lányok rajongására az első professzor
iránt, – senki sem sejtette, s ő maga is alig akarta, merte tudomásul
venni, még kevésbbé mások előtt megnyilatkoztatni. Ha Szakáll
beszélt, alig tudott ránézni, még a legegyszerübb dolgokat is
félszegen mondotta el s gyakran ok nélkül zavarba jött. Csak akkor
volt nyugodt, ha vacsora vagy ebéd után az úri szobába mentek át,
ott leültek, a leányok kézimunkát vettek elő, az asszony olvasott, a
fiatalember pedig egy másik sarokban a könyveket bújta s ő néha
reávethette tekintetét s elnézhette komoly homlokát, izgatottan
csillogó szemét, amely mohón, a kincskereső szívósságával, falta a
betüket. Ilyenkor úgy érezte, hogy csak ő tudja méltányolni ezt a
lenézett, kigunyolt, okos fiút s ez még egyszer, rövid idő alatt, felül
fog emelkedni valamennyiükön s akkor boldogok lehetnek, ha szóba
áll velök.
Szakáll a maga részéről keveset foglalkozott a leánnyal, csak
annyit, mint a többiekkel s nem alkotott magának képet,
tulajdonképpen Hortensenek mi a róla való véleménye? Élete
keretébe a leány nem tartozott szorosan s így csak annyit beszélt
vele, amennyi éppen szükséges volt. És Hortense sem iparkodott
előre lépni, szerényen megvonult a háttérben s csak a lelke
melegsége vette állandóan körül a fiatalembert. Látva, hogy Szakáll
milyen korán kelő, a leány meghozta a legnehezebb áldozatok
egyikét, s hogy mintegy megmutassa az ifjúnak, hogy ő komolyabb
és szorgalmasabb, értékesebb a többinél, ő is már hajnalban talpon
volt. Nagyon nehezére esett ez az elhatározás, mert azelőtt, ha
lehetett, délig is aludt, de azért vitézül kiállta és kakasszóra kiugrott
az ágyból.
Szakáll egy nap éppen a fák közt bolyongott és az egyházjogot
böngészte, mikor egy női alak jelent meg a terrasz lépcsőin. A jogász
eleintén oda sem ügyelt, azt hitte, valamelyik tévedésből korábban
fölkelt cseléd jár arra, de azután megpillantva, hogy a hölgynek
napernyő van a kezében és virágos kalap a fején, érdeklődve ment
közelébb.
– Ah! Hortense! – szólt egyszerre meglepve, mert alig akart
rövidlátó szemének hinni. – Ilyen korán már talpon?
– Igen, – felelt egy csöppet elvörösödve a leány, – elhatároztam,
hogy nem mulasztom el a szép tavaszi reggeleket. Kár lenne ezt az
időt átaludni.
– A reggelek bizony szépek, – felelte Szakáll, aki, jóllehet, már
hajnalban mindennap künn járt, soha még egy napfölkeltét, erdő-
ébredést meg nem figyelt. – Ilyenkor van a legnagyobb csönd és
ilyenkor legvilágosabb a fej!
Elhallgattak, Szakáll a könyv lapjait pörgette.
– Sétálni megy? – kérdezte a fiatalember.
– Igen, kigyalogolok a vasutig és onnét vissza.
A jogász biccentett fejével és így szólt:
– Az emberek nem is tudják, mit mulasztanak az alvással. Ezeket
a reggeli órákat ráadásnak kapjuk a sorstól, már mint azok, akik
meg tudjuk előzni halandó társainkat.
– Igen, – felelte Hortense, aztán egy kis csönd után a kezét
nyujtotta. – A viszontlátásra.
– A viszontlátásra, Hortense.
A leány karcsu alakja, rózsaszínü napernyője eltünt a kapu
mellett. Szakáll egy pillanatig utána nézett, aztán tovább olvasta az
egyházjogot.
A leány kisétált a fasoron át a vasutig, kíváncsian nézte el, a
reggeli gyorsvonatok hogyan robognak ki és be a parányi őrház
mellett, aztán visszafordult és hazatért. A fiatalember még ott sétált
a platánok között.
– Szép volt? – kérdezte Szakáll, aki éppen egy definiciót mormolt
maga elé.
– Szép, – mosolygott Hortense. – Ezután minden nap korán
fogok kelni.
– Igen helyes, – bólintott a fiatalember és tovább sétált, ujjain
számolgatva el a Graczián előtti gyüjteményeket.
A leány egy pillanatig habozott, aztán fölment a terraszra. A
szokatlan séta jó étvágyat keltett benne s már nagyon szeretett
volna reggelizni, de látva, hogy Szakáll még csüggedetlenül sétál
tovább a fák közt, elszégyelte magát. A jogász még korábban kelt s
a sétához tanult is, mégsem éhes – mennyivel több akaraterő és
önmegtartóztatás lakik ez ifjúban! És Hortense némán türte az
éhséget, bár kincset adott volna egy darab kenyérért, s csak akkor
látott a reggelihez, mikor a jogász betéve tudta a Graczián előtti
gyüjteményeket, ami pedig nem csekély dolog, miután maga Walter
már harminchét ilyet sorol föl.
Következő napon Hortense újra hajnalban kelt föl, ismét sétára
indult s újra nemes önmérsékléssel türte az éhséget. Harmadik,
negyedik nap szintén megmaradt ez állásponton, ötödik nap
azonban, amint felöltözködött és a kertbe akart indulni, rákezdett az
eső. Már az éjszaka fülledt volt a levegő és most a cifra rézcsatornák
patakokban öntötték a vizet.
Szakáll a csukott terraszon ült és a könyveit forgatta.
– Ma nincs séta, – szólt Hortensenek, mikor az szomoru arccal
toppant be a verandára. – Kár volt korán fölébredni.
– Már megszoktam a jókor való kelést, – füllentett a kis leány,
pedig nem úgy volt: mindennap teljes lelki erejére volt szüksége,
hogy az álmokat leküzdje. – Nevetek magamon, ha elgondolom,
hogy azelőtt néha délig is aludtam.
A fiatalember ránézett és nem felelt. Most vette először jobban
szemügyre a barátságos házban való tartózkodása alatt Hortenset s
úgy vélte, hogy nemcsak kedves, de szorgalmas kis lány is. Hizelgett
neki a tudat, hogy nyilván ő járt elől a jó példával, s azért egy
pillanatra megfeledkezett a tanulmányokról és tekintetét a leány
karcsu termetére függesztette.
Hortense kezébe vette az asztalon fekvő jogi könyvet.
– Ezt mind meg kell tanulni? – kérdezte a nehéz fóliánst lapozva.
– Ezt mind, – hagyta helyben Szakáll. – De ez csak egy tantárgy
s talán épp a legkönnyebb. A többi még terjedelmesebb és sokkal
nehezebb.
– Az „Egyházjogtan Kézikönyve“, – olvasta el Hortense a címet. –
„Különös tekintettel a magyar állam egyházi viszonyaira…“ Ez tényleg
nagyon nehéz lehet, – tette hozzá megszeppenve a minden oldalról
feléjevillanó latin nevektől és idézetektől. – Szabad egy kicsit
néznem?
– Csak tessék, kisasszony, – felelte a fiatalember pápaszemét
megtörölgetve.
Hortense lapozott, aztán találomra elkezdett olvasni:
– A honor azon megtiszteltetésben áll, melyben az egyház a
kegyurat, mint különös jótevőjét részesíti; ezen tiszteletbeli jogok
között azonban csakis a honor processionis sarkal a közönséges
egyházjogon, a többiek a patricularis jog és szokásnak folyományai
és aszerint itélendők meg, számíttatnak pedig oda 1. jus vagy honor
listrae…
Nem értett belőle egy szót sem: még azt se tudta, mi az a
kegyúr, nemhogy azt, mi a honor processionis és a honor listrae… És
ez csak pár sor volt a könyvből, hát még az egész munkát betéve
tudni és a latin idézeteket megtanulni! Nagy tisztelettel nézett föl
Szakállra.
– És maga… ezt mind tudja? – kérdezte félénken.
– Körülbelül mind, – bólintott a jogász, mintha azt mondaná,
hogy tudja, hány almafa van a kertben. – Próbálja fölnyitni a könyvet
valahol és kérdezzen ki.
Hortense nem akarta elhinni, az elolvasott pár sor oly
titokzatosan, oly érthetetlenül csengett fülében, hogy
elképzelhetetlennek látszott, amit a fiatalember mond. Fölnyitotta
egy helyen a könyvet.
– A metropoliták külön nemei és e hatalom természete
egyáltalán, – olvasta el a fejezet címét.
Szakáll azonnal elkezdte recitálni.
– A metropolitai hatóság folyománya a pápai főnökségnek,
melynek némely jogai különféle fokozatban egyes oly egyházakra
szálltak át, melyeknek püspökei többi püspök-társaik között külön
egységi gyúpontot képezvén, lassanként azok fölé kezdtek
emelkedni…
A leány szédülve nézett a fiatalemberre… Milyen más, mennyivel
különb ember ez a többinél, akik folyton csak zongoráznak,
beszélgetnek, pipáznak és semmi egyébbel nem gondolnak!
– Ez… ez nagy dolog, – mondotta végre büszkén.
Szakáll mosolygott.
– Csak szorgalom, – jegyezte meg, de azért jól esett neki az
elismerés.
Az eső elállott, az égbolton már csak bárányfelhők maradtak.
– Most megyek sétálni, – szólt Hortense, még mindig
megilletődve és búcsúra nyújtotta kezét.
Szakáll egy pillanatig gondolkodott, aztán a könyv után nyúlt.
– Viszontlátásra, Hortense, – mondotta nyugodtan. – Remélem,
máskor is kihallgat?
Ezután, amikor elindult hazulról s ha visszatért, a leány pár
percre megállt a kertben s kikérdezte Szakállt. Hortense mindig
elcsodálkozott ennyi tudás és szorgalom láttára. Össze-vissza
kérdezte a fiatalembert, a könyv elejéről, végéről, de Szakállt nem
lehetett megfogni, amit egyszer a fejébe raktározott, azt a safe
deposit-nél szívósabban őrizte s nem volt az a hatalom, amely tőle
elrabolta volna.
Egy szép tavaszi reggelen Szakáll így szólt a sétára induló
leányhoz:
– Kedvem volna ma egy kissé sétálni. Megengedi, hogy
elkísérjem?
– Szívesen, – felelt Hortense kifeszítve ernyőjét. – Kimehetünk a
Rákosig.
– Helyes, veszem a kalapomat és megyünk.
Elindultak végig a virágos köröndön, ki a jegenyefasoron át a
vasuti töltésnek, ahol a fővárosi élet már befejeződött s csak
veteményes kertek és apró nyaralók következtek. Ember csak kevés
járt erre, azonban a kertekben ingujjas napszámosok dolgoztak,
kalapáltak, javítgattak, a kerítéseket mázolták. A vasuti sorompónál
pár percig várakozniok kellett, most száguldott át a délről jövő
gyorsvonat.
– Még nem is jártam erre, – szólt Szakáll, szorgosan véve
szemügyre mindent. – Érdekes, milyen közel van ide a város és
mégis milyen más, mennyivel természetesebb erre minden. Még
virágokat is lehet látni.
– Várjon, majd künn a Rákosnál. Oh! Mennyi nefelejts van a
partján!
– Az messze van még?
– Csak jőjjön és ne féljen.
Szakáll nevetett, hosszu idők óta talán először.
– Nem félek, – szólt, – eleget gyakorlom magam a járásban.
Kíváncsian nézegetett körül a salátás földek közt és így folytatta:
– Ezeknek a telkeknek rövid idő alatt nagy értékük lesz. A főváros
közvetlen közelében fekszenek, meghatározott útvonal mellett;
biztos, hogy a város erre fog terjeszkedni. Ha az embernek pénze
volna, szépen kereshetne mellette.
– De miért terjeszkedne erre a város? – kérdezte Hortense.
– Mert máshova nincs útja. A víz mellett, mint évszázados
tapasztalat mutatja, egy város se akar elhuzódni. Ez olyan
atavisztikus félelem, mint a hogy a ló még ma is menekülni próbál,
ha valami zaj vagy ismeretlen tárgy megriasztja; mint az őserdőben,
ahol a futással kikerülte a párducot vagy a kigyót; hm. Az emberek
azt hiszik, hogy a víz elöntheti őket, ma, a legtökéletesebb vízművi
védekezések idejében… A bika még ma is rátámad a piros színre,
mert évezredek előtt az erdőben ez a szín a tigrist jelentette neki.
Nos. Mi sem tudjuk magunkat egészen emancipálni az őskori érzések
és félelmek alól és ezért a városok a folyókkal ellenkező irányban
terjeszkednek, vagyis errefelé, ahol most járunk. Pest a Dunapart
mellett ma is még csak annyi területet foglal el, mint száz év előtt.
– A maga figyelme mindenre kiterjed, Péter, – szólt a fiatal leány.
– Én azt hiszem, magából egyszer nagyon híres ember lesz.
– Oh, hol vagyok én attól? Én csak a kötelességemet akarom
teljesíteni. Azért, hogy tanulok, még nincs jogom várni, hogy
kiváltságos sors jut osztályrészemül. Mások is iparkodnak és
tökéletesitik magukat, ha ezek mind híres emberekké válnának, szűk
lenne Magyarország a nagy emberek számára.
Hortense megrázta a fejét.
– Nem, higyje el, maga más, maga különb, mint a többi ember:
okosabb, erélyesebb, egyenesebb… és bizonyosan boldogulni is fog.
Egy ideig hallgattak, aztán a leány így szólt:
– És akkor biztosan elfelejt majd bennünket.
Szakáll tiltakozott.
– Hogy tételezheti föl? És miért? Az emlékezetem nem olyan
rossz.
– Az sokszor nem az emlékezőtehetségtől függ, – felelt a kis
leány elhatározással. – Sok ember büszkévé lesz, gőgössé és nem
akarja, hogy eszébe jussanak azok, akikre nincs szüksége, akik
kicsinyek, akik… akik…
– Nos, akik?
A leány nem fejezte be a mondatot.
Lassan kiértek a Rákoshoz. Itt már faluvá vedlett a főváros: a
homokbuckákba csak egy-egy kunyhó, ócska csárda ékelődött, a
patak medre fölött ütött-kopott híd húzódott el. A part mellett
egypár szomoru fűz álldogált és csöndesen nézte, hogyan szaladnak
el a habok a lecsüngő ágak közt. Egy helyen vízáradásos terület volt,
abban száz és száz béka kuruttyolt.
– Mennyi béka! – szólt Szakáll meglepődve.
– Nézze a nefelejtseket, – mutatott Hortense egy fák közé zárt
helyre, – csupa kékség végig a part!
– Csakugyan, – hagyta helyben a fiatalember. – Úgylátszik, nem
kellenek senkinek.
Hortense leszaladt a víz mellé és teleszedte a napernyőjét
mosolygó nefelejtsekkel, aztán kettőt-hármat összekötött egy
fűszállal és Szakáll gomblyukába tűzte.
– Legalább nem sétált hiába, – mondta, és ami még levél,
elmorzsolódott virág a kezében maradt, azt a vízbe dobta és
gyerekes örömmel nézte, hogy úszik lefelé.
Reggeli ideje volt, mikor visszaértek a barátságos házba.
Ezek a séták később gyakrabban megismétlődtek. A ligetet és
annak minden zúgát Hortense jól ösmerte, s mintegy csöndes
megállapodásként sorra mutatta be barátjának. Egyszer kisétáltak a
zöld lombok közé ékelt lövöldébe, ahonnét vasárnaponkint nagy
puskadurrogás hallatszott, aztán a kavicsbányákhoz, amelyeknek
apró köveit hosszu, fekete vonatok cipelték be a városba, majd a
Páskál-malmot csodálták meg, ezt a várszerü ócska szerkezetet,
amely boldog nembánomsággal élte világát, mintha alig
puskalövésnyire tőle nem gőzmalmok hengerei, küllői alkotnák meg
a lisztet, – más napokon kivonultak a vasúti töltésre, s onnét nézték,
hogyan gyakorlatoznak a katonák, sőt egyszer Palotáig is
elmerészkedtek és útközben kóbor oláhokkal találkoztak, akik
tiszteletükre megtáncoltatták a medvéjüket.
Szakáll lassanként megszokta, hogy a délelőttjét Hortense-zal
töltse el s szinte megkívánta társaságát. A fiatalembernek tetszett az
a tisztelet, mellyel a leány reá és akaraterejére fölnézett; az első
kapocs, amely benne kifejlődött, mint sok más, hasonló esetben: a
hiúságé volt. Később a megszokás fejlesztette benne e viszonyt, s
hozzájárult az a mindenkit megérintő tudat, ha van valakink, aki
életünk iránt érdeklődik és sorsunkat a szívén viseli. Mikor reggel
lement a kertbe, a leány már várt reá, ha este a dohányzóban ült és
éjfélig, gyakran azon túl is, a könyveket bújta, Hortense hűségesen
kitartott mellette és bár a kisleány szeme majd leragadt és alig
beszéltek egymással tíz szót, mégis addig ébren maradt, míg Szakáll
be nem fejezte tanulmányait. A fiatalember nem hozott hasonló
áldozatokat, ő inkább csak engedte, hogy szeressék, hogy gondját
viseljék s megelégedett vele, hogy ez kellemes érzést ébreszt benne.
Önző természet volt, aki mindent a maga és a célszerűség
szempontjából fogott föl és világért túl nem lépte volna e határokat.
Hortense volt az első élő lény, akivel, hogy úgy mondjuk,
fényüzésből beszélt, aki bizonyos szórakozást jelentett neki, aki nem
volt számára kenyér és víz, levegő, vagy az élet fönntartásához
szükséges más cikk, hanem szines virág, amelyet csak azért teszünk
az íróasztalunkra, mert üde és illatos: egyéb hasznunk nem lehet
belőle, minthogy a szép és a kellemes hangulatát ébreszti bennünk.
Ez az új tényező egy ideig föl tudta kelteni érdeklődését és lekötötte
a gondolatait. Annyi sok száraz betü után jól esett egy kis életet, egy
csöpp színt látnia, éreznie, hogy tudásán kívül is lakik benne valami,
ami értékessé, színessé, kellemessé teszi az életét.
Anélkül, hogy mélyebbreható gondolatokkal foglalkozott volna,
átengedte magát ennek a jóleső hangulatnak és megfürdött abban a
melegben, melyet a leány szeretete feléje sugárzott. Ezek a kellemes
percek nem kivántak lelki áldozatot, sem meg nem rázkódtatták,
sem örvénnyel nem fenyegették; kár lett volna vakon elmenni
mellettük.
Hortense gyakran maga is érezte, hogy Szakáll lelkében nem tud
maradandó emléket állítani s ez elszomorította. Ilyenkor csöndesen
félrevonult, de nem lázadt föl a sors ellen, úgy érezte, hogy ennek
így kell lennie: némely ember osztályrésze a szenvedés, s mentől
jobban tiltakozik ellene, annyival inkább rabjává válik. A bánat
országába Szakáll nem követte el, s az egész házban csak egy ember
volt, aki meg tudta érteni, hogy szenved: ez Kelemen volt. Ő látta,
hogy a kis leány szíve gyakran vergődik kétségek között, remeg s ez
mindig elszomorította. A szerelem, amely első perctől fogva lelkében
a leány iránt fölébredt, nem veszett ki belőle s ezt a megismétlődő
bánatos képek, a szenvedés rajzai még csak fokozták.
Egy napon Kelemen a kert egyik padján találta Hortenset, aki a
lombok közt, mozdulatlanul, lehunyt szempillákkal ült. A leánynak ez
volt a kedves helye, ahol elrejtve, meg nem zavarva engedhette át
magát csöndes szerelmes gondolatainak, amelyeknek még fájdalma
is édes volt. A fiatalember odalépett melléje és halkan reászólt:
– Hortense.
A leány megrezzent és rávetette ijedt tekintetét, majd ideges pir
szaladt végig még az imént sápadt ábrázatán. Érezte, hogy Kelemen
a szívébe látott s ez keserüséggel és haraggal töltötte el, mert lelke
érzéseit a saját szentélyének tartotta, ahová idegennek nem volt
joga belépni.
– Mit kíván? – kérdezte tekintetét elfordítva a fiatalembertől.
Kelemen szavakat keresett.
– Hortense, – szólt tétován, – én érzem, látom, hogy ön…
szenved. Az arcára van írva, minden mozdulata elárulja. Az ön lelke…
A leány fölállott, szemében a harag könnycseppje csillant meg.
– Ne bántsa az én lelkemet, – szólt hidegen. – Én sem bántom a
másét.
– Ne magyarázzon félre, – folytatta a fiatalember szégyenkezve,
– nem akarok kíméletlen lenni, nem akarok a belső életébe
avatkozni, én, én együtt érzek önnel és…
Hortense sóhajtott és kezet nyújtott Kelemennek.
– Ha velem érez, – szólt megbánva ridegségét, – akkor ne
bántson… és maradjon jóbarátom. Isten vele, – tette hozzá
csöndesen.
A szegény fiu megalázva, összetörve ott állt még egy percig, nem
tudott mit mondani. Hortense lassan fölment a házba és leült az
ablak mellé, ahová a közeledő nyár minden pompája: szinek, illatok,
hangulatok verődtek be.
Kelemen nem tudta, mit csináljon, a szíve sajgott és a fejében
kalapácsok dolgoztak, amelyek összetörték önérzetét és apró,
nevetséges, ügyefogyott embert alkottak belőle, aki fölött egy
szárnylegyintéssel tért napirendre a sors. Ott állott a maga
tehetetlenségében, lealázottságában, végtelen szegénységében,
amely senkit nem érdekelt e földtekén. Az a láng, amelyről azt hitte,
hogy életét meg fogja világítani és föl fogja melegíteni, eltünt,
kihunyt s ami reá várt, az a rideg, fanyar éjszaka volt, telve
némasággal és ürességgel. És mindez égzengés, földindulás nélkül
intéződött el, mint a legapróbb macskakölyök pusztulása, mint egy
magától értetődő hétköznapi dolog, amely egy hangyának az útját
sem változtatja meg.
Ott állt, tétován, még egy félóráig a vidám sárgarigók, a buján
illatozó virágok közt és olyan kábult volt, hogy csak nagysokára
találta meg a kalapját és a sétabotját.
– Hova lett Kelemen? – kérdezte ebédnél Farkas.
Senkise tudott választ adni, Kelemen eltünt, elveszett és többé a
barátságos házba nem vetődött el.
VIII.
Egy napon Szakáll kisétált Hortense-zal a rákosi rétekre.
Napnyugta volt, szép, langyos befejezése egy csöndes nyári napnak,
amely még a kopott kavicsbányákat és a vakondturásos legelőket is
föl tudta frissíteni és kiszínezni. El voltak fáradva a hosszu
gyaloglástól és leültek egy ákác alá, amelynek lassan mozgó lombjai
közt a nap sajátságos árnyéktáncot járt. Némán ültek egymás
mellett, aztán a leány egyszerre megszólalt.
– Péter, – mondta mintegy megfeledkezve mindenről s nem tudva
ellentállni a benne küzdő érzéseknek, – ne hajítson el magától, ne
tiporjon el egy szegény kis leányt.
Megijedt, amint e szavakat kimondta, az elsóhajtott mondat
akarata ellenére tódult ajakára.
Szakáll is meg volt döbbenve és zavartan nézett reája.
– Mért gondolja ezt rólam? – szólt nyugtalanul.
– Nem, nem, bocsássa meg, – dadogta Hortense rémülten, –
nem azt akartam mondani… Ne értsen félre… Milyen együgyü is
vagyok!
Nem bírta tovább, egyszerre sírni kezdett, hangosan,
fájdalmasan, mint a beteg gyermek.
Szakáll tétován állt előtte, érezte, hogy mondania kell valamit, ez
a mélységes fájdalom megrémítette hideg lelkét.
– Hortense, – szólalt meg végre, – kedves Hortense…
A leány felszökött helyéről.
Welcome to Our Bookstore - The Ultimate Destination for Book Lovers
Are you passionate about books and eager to explore new worlds of
knowledge? At our website, we offer a vast collection of books that
cater to every interest and age group. From classic literature to
specialized publications, self-help books, and children’s stories, we
have it all! Each book is a gateway to new adventures, helping you
expand your knowledge and nourish your soul
Experience Convenient and Enjoyable Book Shopping Our website is more
than just an online bookstore—it’s a bridge connecting readers to the
timeless values of culture and wisdom. With a sleek and user-friendly
interface and a smart search system, you can find your favorite books
quickly and easily. Enjoy special promotions, fast home delivery, and
a seamless shopping experience that saves you time and enhances your
love for reading.
Let us accompany you on the journey of exploring knowledge and
personal growth!

ebookball.com

You might also like