Research On Handoff Under Imperfect Spectrum Sensing In Cognitive Radio Networks

In cognitive radio networks, spectrum handoff is required to guarantee the quality of service(Qo S) of licensed users(Primary User, PU).If PU re-occupies the licensed spectrum,the unlicensed users(Secondary User,SU) must vacate the licensed spectrum unconditionally. On the one hand, the efficiency of spectrum handoff is severely affected by the performance of spectrum sensing. This is due to the fact that the probabilities of false alarm and missed detection are unavoidable in realistic spectrum sensing scenarios, which leads to the invalid spectrum handoff. On the other hand, the imperfect spectrum sensing may change the SU's target channel sequence in spectrum handoff. The different target channel sequences have a great impact on the entire performance of spectrum handoff. It is important to conduct the research from two aspects of spectrum handoff and optimal target channel selection algorithm under imperfect spectrum sensing.Firstly, for the inaccuracy of spectrum handoff of SU under imperfect spectrum sensing, PRP(Preemptive Resume Priority) M/G/1 queuing network model is used to describe the process of spectrum handoff and a new spectrum handoff algorithm with error data retransmission scheme is proposed in this thesis. It derives the expression of spectrum handoff delay with the premise of SU's probabilities of false alarm and missed detection. Then the performance of the algorithm is analyzed through two parameters of transmission delay and extended data delivery time. Finally, the relationships between the probabilities of missed detection and extended data delivery time and the traffic loads is discussed by simulation, and the influence of missed detection on SU's extended data delivery time is analyzed quantitatively, which verify the effectiveness of the proposed algorithm.Secondly, for the problem of increasing handoff delay owing to pre-determined target channel having high failure risk in the proactive-decision spectrum handoff, an optimal target channel selection algorithm based on hybrid spectrum handoff is proposed. The proposed algorithm coordinates the advantages of both the proactive-decision and reactive-decision spectrum handoffs. With the metric of minimizing the cumulative handoff delay, the impacts of secondary user's multiple interruptions, primary user's arrival rates and imperfect sensing on secondary user's cumulative handoff delay are considered synthetically. Secondary user's cumulative handoff delay is analyzed and deduced in detail. The simulation results demonstrate that compared with random and proactive-decision and reactive-decision target channel selection schemes, the proposed algorithm can reduce the secondary user's cumulative handoff delay efficiently and its performance advantages are very obvious, especially with extreme situations of frequent channel state variations or heavy traffic loads.