Improving Spectrum Sensing and Multi-user Cooperation in Wireless Cognitive Radio Networks

The Hong Kong University of Science and Technology
Department of Computer Science and Engineering


PhD Thesis Defence


Title: "Improving Spectrum Sensing and Multi-user Cooperation in
Wireless Cognitive Radio Networks"

By

Mr. Chengqi Song


Abstract

Nowadays wireless spectrum resource is more and more scarce because most 
wireless spectrum bands have been allocated to licensed users, however 
measurements show that many bands are very underutilized. To solve this 
problem, Cognitive Radio (CR) is proposed to improve spectrum utility 
without interference to licensed users. CR need to keep aware of wireless 
environment by spectrum sensing and access unused bands without 
interference to licensed users. In this proposal we focus on two important 
issues in CR networks: spectrum sensing and multi-user cooperation. i) 
Spectrum sensing is the key component of CR technology. We improve 
spectrum sensing by studying spectrum usage prediction and cooperative 
spectrum sensing. Spectrum usage prediction can significantly reduce the 
sensing range basing on history information. We conduct a large scale 
empirical study basing on realworld measurement to reveal the 
predictability and ideal predictors for spectrum usage. We also propose a 
new spectrum sensing scheme assisted with spectrum prediction. Cooperative 
spectrum sensing is introduced to improve sensitivity with spatial 
diversity. We improve it by designing a Carrot-and-Stick based strategy to 
stimulate cooperation, relaxing it into asynchronous mode and introducing 
multi-channel coordination mechanism. ii) Because the wireless environment 
of CR networks is more complicated and unstable comparing with traditional 
wireless networks, the cooperation of multiple users is more important. 
Cooperative spectrum sensing is a good example and we also study another 
two important problems: stimulating data forwarding and CR network 
planning. For stimulating data forwarding, we show that the traditional 
solutions are context-based and our new proposed context-free solution is 
more lightweight and bug-free. For CR network planning, we show that our 
solution can support more clients than traditional planning method.


Date:			Thursday, 12 August 2010

Time:			2:00pm – 4:00pm

Venue:			Room 3501
 			Lifts 25/26

Chairman:		Prof. Ajay Joneja (IELM)

Committee Members:	Prof. Qian Zhang (Supervisor)
 			Prof. Lin Gu
 			Prof. Jogesh Muppala
                         Prof. Vincent Lau (ECE)
                         Prof. Xiaohua Jia (Comp. Sci., CityU)


**** ALL are Welcome ****