Active Network Service Composition

International Journal of Engineering Research and Development
e-ISSN: 2278-067X, p-ISSN: 2278-800X, www.ijerd.com
Volume 11, Issue 04 (April 2015), PP.01-11
1
Active Network Service Composition
Israt Jahan1*
, K M Akkas Ali2
, A.K.M. Kamrul Islam3
, Morium Akter4
,
Md. Monowarul Islam5
1
Department of Computer Science &Engineering, Jahangirnagar University, Savar, Dhaka, Bangladesh.
2
Institute of Information Technology (IIT), Jahangirnagar University, Savar, Dhaka, Bangladesh.
3,4
Department of CSE, Mawlana Bhashani Science &Technology University, Tangail, Bangladesh.
5
Department of CSE, Jagannath University, Dhaka, Bangladesh.
Abstract:- Active Network is a novel approach of networking to mobile users in which the nodes are
programmed to perform custom operations on the messages that pass through the node. It provides an
architectural support for dynamically deploying new protocols in an existing network topology. The routers in
an active network can download and execute code that is contained in the packets passing through them, thus
rendering the node recognized and run totally new protocols without making any changes to the architecture of
the network. Because the network's behavior can be altered at any time, active networks could be used to
provide dynamic quality of service (QoS) or to support dynamic solutions to traffic congestion. This research
implements and tests such specialized Active Networks security service known as the firewall and the ping
service in Active Network. Active Network environment will be implemented on a small scale test scenario in
order to study the performance and characteristics of active networks
Keywords:- Active network, Firewall, Ping, Security service.
I. INTRODUCTION
ACTIVE networking started within the Defense Advanced Research Projects Agency (DARPA)
between 1994 and 1995. The idea was to find a new direction in solving some of today‟s networking problems
such as difficulty in the integration of new technologies, poor performance due to redundant operations at
certain protocol layers and difficulty in integrating new services within an existing infrastructure [1]. Active
networking is meant to reduce these problems by creating a virtual platform which is flexible, adaptive and
extensible in nature and one which will allow the rapid and distributed deployment of new services according to
user and application requirements efficiently. Conventional networks passively transfer data from source to
destination node without changing the user data. This implies that the user data is opaque to the system and is
never modified by the system. Computation within such a system is extremely limited as the header of a
protocol data unit (PDU) provides all the necessary information to ensure that the data is sent to the correct
address. The actual contents of the data are ignored and the amount of computation within the network is
restricted to header processing and signaling (for switched circuit network).
Active network is a novel approach to network architecture in which nodes may be programmed to the
user or application requirements and is therefore more dynamic and flexible in nature. This may be
accomplished by injecting small or mini programs at certain „active‟ nodes as packets are transmitted over the
network. Tennenhouse [2] explains the concept of an active network as one where nodes – the switches, routers,
hubs, bridges, gateways etc, - have the ability to allow customized computations to be performed on the user
data. This kind of network is called active because new computations can be dynamically injected into the
network, thereby allowing the network to behave differently with different data types. Packets in an active
network (also known as capsules) carry mini programs in addition to the original data. Replacing passive
packets with capsules that are active means that new services may be added to the network or existing services
may be enhanced, customized, and independent of the underlying infrastructure when necessary. Also, specific
functions according to a particular application may be deployed at certain locations within the network. Some
immediate applications of active network technology include firewalls, web proxies, mobile computing and
application services. However, the potential uses of this sophisticated technology may yet be unexplored.
II. OBJECTIVES
TO resolve the issues of efficiency and fairly managing both the processing and bandwidth resources in
programmable active networks and providing better QoS guarantees to the competing flows. This research work
has been focused on firewall and ping services implementation on active network. This research will pursue the
following objectives.

2
1. Implement and test such specialized Active Networks security service known as the firewall and the ping
service.
2. To enable the rapid deployment of new services onto the network with greater ease;
3. To enable user-controllable functionality to be added to the network;
4. To facilitate the enhancement/customization of the network according to user/application requirements
dynamically.
III. RELATED WORKS
THE following list summarizes some potential applications of active networking in today‟s
environment, as mentioned by Tennenhouse and Calvert [3]. Firewalls filter packets that enter a particular
domain based on certain application or user-functions. The use of active networking will automate this process
by allowing only approved user or applications to authenticate themselves at the active node (firewall). In
mobile computing active networking may be used to adapt and adjust the mode of transmission or activate
certain functionality according to the type of connection of the end system to the network. For example, if the
end system is connected by a low bandwidth link such as a phone line, then the active node may do more file
caching or link compression. In addition, certain security features may be invoked if the end system is linked by
an insecure or external connection. Active networking may be used to distribute the responsibility of multicast
retransmission from the original sender to other receivers down the multicast tree. Here, all receivers are active
nodes and are aware of the states of other nearby receivers and will be able to provide retransmission in the
event of transmission loss thereby removing the load of re-transmission from the original sender as well as
reducing delay and increasing reliability. Active networking will allow state and processing to be done within
the network. In ensuring guaranteed quality of service over networks, many approaches require that the
originating node detect the conditions of the network and adapt accordingly to prevent congestion or loss during
transmission. In such cases, the inherent delays between detection and adaptation of the sender, along the path
of transmission may degrade the performance of the network especially over long links. It is therefore desirable
to distribute the state of awareness and adaptation to other nodes within the network. Thereafter, the network
would be able to adjust to the correct mode of adaptation as when necessary. Web caching is usually done to
reduce delays and congestion when a web application is accessed. Objects are normally cached from servers to
locations close to clients using a manually configured static hierarchy.
IN dynamic conditions, active networking may provide more efficient means of caching. This includes
adaptive web caching [4] using small caches with information on neighboring nodes contents [5], and using an
active network to route cache requests to preconfigured locations [6]. The usual method of polling and checking
for abnormalities restricts the management intelligence of the network to certain management stations only. This
causes delays before the anomaly can be rectified. Using an active network approach may improve the
management capabilities of a network since each node may be programmed to be aware of and adaptive to the
state of the network. Projects in this area include Netscript at Columbia [7], and Smart Packets at BBN Systems
and Technologies [8].
IV. METHODOLOGY
THE active network architecture (which is under development) discusses the common basic
functionality of the active node programming interface – the part which is visible to the end user [9-10]. Here,
matters pertaining to addressing, end-to-end services, capsule or packet processing, and node resources are
defined generically without specific reference to any language under the DARPA active networks program.
Fig. 1 Proposed Active Network Architecture

3
AS seen in Fig. 1, the common functionality of active network architecture is separated into two
entities, namely, the execution environment (EE), which is similar to a shell program that acts as an interface
providing end-to-end services, also known as active applications, AA to users, and the node operating system
(NodeOS), through which the EEs draw the necessary resources at the node. These resources include
transmission, bandwidth, computing and storage. Implementation of active network API is handled by the EEs,
whilst the NodeOS is responsible for managing access of the EEs to these local resources, hiding their details as
well as the existence of other EEs in the node [11]. An execution environment may provide only a basic service,
controlled by parameters provided by the user, or a more extensive one, such as that of a powerful interpreter -
depending on the implementation of the EEs [12][15].
V. SIMULATION AND EXPERIMENT
A.Network Topology under Test for firewall
The testing environment for firewall service has the following components:
1. Number of nodes: 16 nodes (from 0 to 15)
2. Data rate from 9 to 6 and 2 to 6: 10KB/s
3. Type of link between nodes: Duplex Link
4. Intra-Domain Bandwidth: 10Mbits/s and 56Mbits/s figure
5. Intra-Domain prop Delay: 5ms and 35ms
Fig. 2 A small scale topology
Fig. 3 Operation of Firewall Service in NS-2

4
FIG. 3 show the operation of active network firewall. Node 6 has been equipped with the firewall
capsule which is why it is represented by a blue hexagon around it. The black packets or unwanted packets are
being sent from node 9 to node 6 and the blue packets are being sent from node 2 to node 6. What we can
observe from figure 3 is that the black packets of node 9 are being dropped, whereas the blue packets of node 2
are being received by node 6, since the firewall of node 6 blocks incoming traffic from node 9.
SEVERAL Several graphs were plotted based on the results obtained by simulating topology of Fig. 2. The first
and most important parameter to analyze is the throughput of Fig. 4. Fig. 4 shows the total number of packets
generated at all nodes. We can see that a maximum of 55 packets were generated at any particular time.
Fig. 4 Throughput of generating packets
Fig. 5 shows the total number of packets generated at all nodes. We can see that around 30 packets
were dropped. This shows that only packets from node 2 were allowed and the packets of node 9 were dropped.
Fig. 5 Throughput of dropping packets
FIG. 6 show the total number of packets generated at the sending nodes together with their destination nodes.
Fig. 6 also show that node 2 generates packets for node 6 and node 9 also generates packets for node 6.

5
Fig. 6 Packets generated at nodes 2 and 9
FIG. 7 show the total number of packets received together with the source node. Fig. 7 also shows that packets
from node 2 are accepted by node 6 and packets from node 9 are rejected.
Fig. 7 Packets accepted by node 6
FCTIVE network allows special packets to be treated differently by executing special programs at
nodes. These specialized nodes then have the ability to cater the needs of the special data packets. The ability to
execute and especially cater for certain packets gives active network the ability to dynamically develop and test
new protocols.
B. Result and Discussion for Ping
The Ping stands for Packet Internet Groper (which command is used to verify that a network
connection exists between two computers). Ping command always gives the following statistics for each host:
 Number of sent/received packets for this host
 Loss percentage
 Minimum/average/maximum RTT as measured from these packets
For small topologies we choose hand built network or for large networks choose a predefined binary tree
network as well as the deployment mode.

6
Fig. 8 AN Scenario
FROM Fig. 8, we can see that there is an “Active Network Scenario” and we choose parallel Deployment and
Hand build Domains because of small network topology.
Fig. 9 Active network Simulator
IN Fig. 9, this network scenario uses two protocols namely 103 and 104 with two domains. Specifically
the number of Hosts per domain, the domain Adjacency Matrix, and the external connections specifying what
nodes interfaces connect to the external port of the domain. Other parameters like the domain network address
and network mask as well as the control node of the domain are also specified for building a domain.
Fig. 10 Domain 0

7
Fig. 11 Domain 1 or Hosts
Fig. 11 shows that protocol 103 and 104 are just simple examples of active applications.
THE protocol 103 functions in fact like a Ping, where the sending node sets a timer and waits for a
response before running out of the timer. Intermediate nodes will simply route the packets from source to
destination by executing the code associated to the protocol 103.
THE protocol 104, functions like a Ping without return, so when the packet reaches its destination a
message „Hello World‟ is sent to its output port. In order to be able to send packets of the protocol 103 from the
node domain0 or h3 to the node domain1 or h1 and packets of the protocol (prot104) from the node domain1 or
h2 to the node domain0 or h3, the control node h2 of the (domain0) should possess the protocol 104 and the
control node h3 of (domain1) should possess the protocol 103.
Fig. 12 Simulation parameter
IN Fig. 12, Simulation Parameter frames which contain Intra-Domain Bandwidth, Intra-Domain Prop Delay,
and Normal Capsules‟ Buffer Size and request Capsules‟ Buffer Size are shown.

8
Fig. 13 Output parameter
IN Fig. 13, Output parameter is created and we have to choose Deployment time, Packet Loss and Round Trip
Time (RTT plotter).
Fig. 14 TCL0
Fig. 14 shows that when traffic begins to flow, this same window can be used for monitoring in order to trace
packets.
Fig. 15 RTT plotter

9
ON the graph that packets from 20 to 31 have been lost. Their lost is represented by a RTT equal to the
Timer fixed as a parameter. These losses are due to the saturation of the normal capsules buffer. This RTT
plotter shows up in real time and traces the Round-Trip-Time as a function of the number of the transmitted
packet between a source and a destination node.
Fig. 16 Link traffic analysis
In Fig. 16, 0utput parameters window, 3 links has been chosen.
Fig. 17 Deployment time analysis
C. Ping Statistics
THIS frame displays for each packet of the protocol 103 it‟s RTT, the maximal, minimal and average RTT as
well as the percentage of lost packets. Results are available for every couple of source & destination.
Fig. 18 Ping statistics

10
From the Fig. 18, we can see the ping statistics. Total 50 packets are sent, 38 packets are received and 12
packets are lost.
Fig. 19 Packet loss display
Fig. 19 show that trace of normal packet loss
VI. CONCLUSION
IN this research we have demonstrated that Firewall and Ping service can be successfully implemented
a specialized Active Network security service. The test results indicate that this is a value option in Active
Network security implementation. This research implements and tests one such specialized Active Networks
security service known as the firewall. Usual firewalls run on network gateways as either software or hardware.
In this experiment, we designed a firewall such that it runs as an active network service so that each node on the
network can download this particular service for self-defense. This research motivates the need for an
architectural framework for the parallel deployment and Hand Builds Domains of active network service in a
small scale network environment.
REFERENCES
[1]. Shah S.Y., Szymanski B., Zerfos P., Bisdikian C., Gibson C., Harries D. “Autonomous configuration
of spatially aware sensor services in service oriented WSNs” IEEE International Conference on
Pervasive Computing and Communications Workshops (PERCOM Workshops), pp 312-314, March
2013.
[2]. Geyik S., Szymanski B., Zerfos P. "Robust dynamic service composition in sensor networks," Services
Computing, IEEE Transactions on, vol. PP, no. 99, p. 1, 2012.
[3]. Tychogiorgos G., Bisdikian C. "Selecting relevant sensor providers for meeting your quality
information needs," in Mobile Data Management, 2011 12th IEEE Int'l Conf. on, 2011.
[4]. Yachir A., Tari K., Amirat Y., Chibani A., Badache N. “QoS based framework for ubiquitous robotic
services composition” IEEE/RSJ International Conference on Intelligent Robots and Systems, pp 2019
– 2026, October 2009.
[5]. Tennenhouse D. L., Smith J. M., Sincoskie W. D., Wetherall D. J., and Minden G. J. "A Survey of
Active Network Research" IEEE Communications Magazine, pp 80-86, January 1997.
[6]. Tennenhouse D. L., and Wetherall D. J. "Towards An Active Network Architecure" Computer
Communications Review, 26, 2, pp 5-18, 1996.
[7]. Calvert K. L., Bhattacharjee S., Zegura E., and Sterbenz J. "Directions in Active Networks" In IEEE
Communications Magazine, pp 72-78, October 1998.
[8]. Zhang L., Michel S., Nguyen S., Rosenstein A., Floyd S., and Jacobson V. "Adaptive Web Caching:
Towards A New Global Caching Architecture" In 3rd International WWW Caching Workshop, 1998.
[9]. Bhattacharjee S., Calvert K. L., and Zegura E. W. "Self-Organising Wide-Area Network Caches" In
IEEE Infocom '98, 1998.
[10]. Legedza U., Wetherall D. J., and Guttag J. "Improving the Performance of Distributed Applications
Using Active Networks" In IEEE Infocom '98, 1998.

11
[11]. Yemini Y., and Silva S. da, "Towards Programmable Networks" In IFIP/IEEE International Workshop
on Distributed Systems: Operations & Management, L'Aquila, Italy, October 1996.
[12]. Schwartz B., Zhou W., Jackson A., Shayer W. T., Rockwell D., and Partridge C. "Smart Packets for
Active Networks" BBN Systems and Technologies, http://www.nettech.bbn.com/smtpkts/smart.ps.gz,
1998.
[13]. Legedza U., Wetherall D. J., and Guttag J. "Improving the Performance of Distributed Applications
Using Active Networks" In IEEE Infocom '98, 1998
[14]. http://www.j-sim.org/start.html
[15]. http://www.isi.edu/nsnam/ns/

Active Network Service Composition

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Active Network Service Composition (20)

More from IJERD Editor (20)

Recently uploaded (20)

Active Network Service Composition