5G Implementation for

Crowdsensing

Francis Mokogwu Turar Abdol

 Introduction

▪ Connected devices:
Smartphones, laptops, tablets, wearables
etc.
The growth of mobile devices is
exponential.
100% growth expected by 2025.
Every human is expected to own an
average of six mobile connectible devices.
These devices have in-built sensors.

Fig 1. Expected number of connected devices © Murat Uysal

2
 Mobile Crowdsensing
▪ Defined as the process where a large number of connected devices collectively extract and share
information to measure, map, estimate or predict any event of common interest.

▪ Common mobile device

sensors:
Accelerometer
GPS
Gyroscope
Compass

Fig 2. System model of MCS ©Capponi et al.

3
Applications of Mobile Crowdsensing

▪ Automobile traffic control

▪ Location mapping
▪ Environmental monitoring
▪ E-health

Fig 3. MCS applications applicable in different fields

4
 MCS Paradigm
Participatory Sensing Opportunistic Sensing

Users actively engage in the data Takes random sample which is

collection application defined

Users manually determine how, what, Automatic/involuntary sample

when, where to sample collection

Can avoid phone context issues Cannot avoid phone context issues

Higher burden or costs Lower burden or costs

Data filtering

Dataset ready for processing and Fig 4. Description of Participatory and

storage Opportunistic Crowdsensing

5
 Basic structure of crowdsensing system
▪ Basic components of a traditional MCS
systems include:
Sensors (devices)
Transmission media (Wi-Fi/3G/4G)
Cloud (servers)

Fig 5. Illustration of MCS architecture

6
Challenges of Cloud based MCS Systems
▪ Availability and reliability
▪ Portability
▪ Cost
▪ Downtime
▪ Use of bandwidth
▪ Latency
▪ Privacy & Security

7
5G integration: The way forward…..
▪ Why?
Peak data rates of >10Gbps
Ultra low latency
Ultra high reliability
Enhanced mobile broadband
Low energy consumption

Fig 6. Features of 5G

8
 Proposed Edge-computing based MCS system
▪ Entails the decentralization of the existing cloud based system to enable the implementation of the
5G hierarchy.

▪ Decentralized clouds help to

achieve:
Spatially distributed edge nodes
Transmission distance reduction
Localized data processing and
storage

Fig 7. Edge computing based MCS system

9
 Results

▪ Based on data generated from simulation on EdgeCloudSim to validate our proposed model, we
predefine some performance criteria.

 Service time: networking time and processing time.

 Resource utilization: processing response of the network and device CPUs.

10
 Results……
Service Time
2.5

Tim e (s) 2

1.5

0.5

0
0 5 10 15 20 25 30 35 40 45
Edge Servers

Networking Time Processing Time

Fig 8. Networking time and Processing for deployed Edge servers

11
 Results……..
Service Time
2.5

2
T im e (s)
1.5

0.5

0
0 100 200 300 400 500 600
Users

Networking time edge-based Processing time edge-based

Fig 9. Service time analysis for Edge-based deployment

12
 Results……..
Service Time
1.8
1.6
1.4
1.2
Time (s)
1
0.8
0.6
0.4
0.2
0
0 100 200 300 400 500 600
Users

Networking time cloud-based Processing time cloud-based

Fig 10. Service time analysis for Cloud-based deployment

13
 Results……..

Resource Utilization
70
60
Utilization (%)
50
40
30
20
10
0
0 100 200 300 400 500 600
Axis Title

Network Capacity Edge-based Network Capacity Cloud-based

Fig 11. Network Capacity for Edge-based and Cloud-based deployments

14
 Results……..

Resource Utilization
80
70

Utilization (%) 60
50
40
30
20
10
0
0 100 200 300 400 500 600
Users

CPU Utilization Edge-based CPU Utilization Cloud-based

Fig 11.CPU Utilization for Edge-based and Cloud-based deployments

15
Summary

▪ The tradeoffs include:

The edge-based deployment had better time response for networking time.

The cloud-based deployment had a quicker processing time..

▪ But, to buttress our point, the edge-based deployment fulfils our objective of reduced latency, high
reliability and real-time transmission.

▪ Hence, 5G can be successfully implemented.

16
 References
[1] H. Elazhary, “Internet of Things (IoT), mobile cloud, cloudlet, mobile IoT, IoT cloud, fog, mobile edge, and edge emerging
computing paradigms: Disambiguation and research directions,” Journal of Network and Computer Applications, vol. 128, no.
June 2018, pp. 105–140, 2019, doi: 10.1016/j.jnca.2018.10.021.

[2] V. Raida, P. Svoboda, M. Lerch, and M. Rupp, “Crowdsensed performance benchmarking of mobile networks,” IEEE Access,
vol. 7, pp. 154899–154911, 2019, doi: 10.1109/ACCESS.2019.2949051.

[3] D. Belli, S. Chessa, L. Foschini, and M. Girolami, “The rhythm of the crowd: Properties of evolutionary community detection
algorithms for mobile edge selection,” Pervasive Mob. Comput., vol. 67, p. 101231, 2020, doi: 10.1016/j.pmcj.2020.101231.

[4] B. Sliwa, R. Falkenberg, T. Liebig, J. Pillmann, and C. Wietfeld, “Machine Learning Based Context-Predictive Car-to-Cloud
Communication Using Multi-Layer Connectivity Maps for Upcoming 5G Networks,” IEEE Vehicular Technology Conference, vol.
2018-Augus, pp. 1–7, 2018, doi: 10.1109/VTCFall.2018.8690856.

[5] N. Ullah, X. Kong, L. Lin, M. Alrashoud, A. Tolba, and F. Xia, “Real-time dissemination of emergency warning messages in 5G
enabled selfish vehicular social networks,” Computer Networks, vol. 182, no. May, p. 107482, 2020, doi:
10.1016/j.comnet.2020.107482.

17
Thank you for your audience

18