Research On The Key Technologies For Relay Transmission In Cognitive Radio Ad Hoc Networks

Cognitive Radio AdHoc Networks(CRAHNs)is a kind of wireless communication network which integrates cognitive radio technology and AdHoc network technology.It has many advantages,such as low cost,easy networking and high spectrum utilization and so on.Further,the mature application of CRAHNs will provide strong communication guarantee for the network construction of information war.artificial intelligence and emergency rescue.As the key technology of CRAHNs,relay transmission technology plays an important role in expanding network coverage and improving network reliability.However,with the continuous expansion of the scale of CRAHNs,network state space,channel selection dimension and channel rendezvous time will increase,which makes the existing relay transmission technology difficult to effectively control the network delay and continuously guarantee the transmission performance.Therefore,from the factors of network state space,channel selection and channel rendezvous that affect the response time and throughput performance of CRAHNs.the following four key problems about relay transmission are studied in this paper.1.The optimal selection strategy of multi relay nodes is studied.For the problem of redundancy or insufficiency in the selection of multi relay nodes,a dynamic programming optimization model is constructed to minimize the sum of energy,and a state space pruning algorithm based on the inverse probability of Metropolis criterion about simulated annealing is proposed.According to the channel state transition probability and the optimal neighbor subset information.the optimal multi relay selection strategy is implemented.Experiments show that the strategy can effectively reduce the computational complexity.improve the throughput and decrease the average delay.2.The optimal channel selection strategy is studied.In order to solve the channel decision problem caused by the coexistence of multiple sensing channels in CRAHNs.a channel selection strategy with "channel as arm" and multi constraints based on the learning model multi-armed bandit(MAB)is proposed.Based on the Gilbert Elliott Markov channel model with two states,the maximum rewards of upper confidence bound(UCB)is used to optimize the channel selection,and a fast convergence method for the optimal channel based on channel quality estimation is proposed.3.The adaptive data transmission optimization strategy is studied.Aiming at the time-varying channel in dynamic network environment,a packet-size adaptive transmission strategy with "packet-size as arm" based on MAB is proposed.Under the constraint that the success or failure of transmission will be awarded or punished respectively,the cumulative rewards of UCB is used to maximize the system throughput,and the convergence speed of the system is improved under the double effects of confidence factor and balance coefficient.4.The blind channel rendezvous strategy based on channel hopping sequence is studied.In order to solve the problem about establishing communication link in blind channel,the channel rendezvous strategy based on channel hopping sequence in synchronous and asynchronous network is studied.Based on the theory of common channel hopping sequence SYN-ETCH and Quorum,PSCH construction method with periodic stacking is proposed.Thus,average time to rendezvous is decreased and the rendezvous stability is guaranteed.