Spectrum sensing framework for infrastructure based cognitive vehicular communications / Chembe Christopher
Vehicular communication is posed to aid road users overcome challenges faced today by sharing road conditions. However, vehicular communication still faces many challenges before deployment. One such challenge is insufficient radio frequency channels. Vehicular communications has been allocated 75MH...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
2017
|
| Subjects: | |
| Online Access: | http://studentsrepo.um.edu.my/8269/2/All.pdf http://studentsrepo.um.edu.my/8269/6/chembe.pdf http://studentsrepo.um.edu.my/8269/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Vehicular communication is posed to aid road users overcome challenges faced today by sharing road conditions. However, vehicular communication still faces many challenges before deployment. One such challenge is insufficient radio frequency channels. Vehicular communications has been allocated 75MHz for dedicated short range communication (DSRC) at 5.9GHz bands. Nevertheless, the channels can get congested during peak hours or accident scenarios affecting the transmission of safety messages. To alleviate the problem of scarcity of channels in vehicular communications, dynamic spectrum access and Cognitive Radio (CR) technology is proposed. CR identifies spectrum opportunities in licensed frequency bands that can be accessed by unlicensed users through spectrum sensing. Spectrum sensing is performed by an individual vehicle or cooperation. In vehicular communications, spectrum sensing is challenging due to vehicle mobility and dynamic topological changes. Additionally, many challenges associated with spectrum sensing still exist such as shadowing, multipath fading and unknown primary user (PU) activities. This research aims at mitigating some of the problems of spectrum sensing in vehicular communication mentioned above by proposing a sensing framework. The proposed framework is divided in two parts. The first part involves sensing of PU signal by individual vehicles on the road using adaptive sensing. The adaptive sensing is based on energy detection and cyclostationary feature detection. The history of sensing results by vehicles is sent to road side unit (RSU) and used in aiding the framework to predict licensed channels likely to be free later. The results are used as reward for reinforcement learning at RSU. The second part of the framework involves RSU learning behavior of PU activity patterns using the sensing history. The framework is evaluated and validated through simulation under realistic VANET scenarios. The performance of the proposed framework is compared to history based approaches in literature. |
|---|