Research On The Technology Of Distributed Multi-channel Selection And Access For Multi-user

With the rapid development of wireless communication technology, various new types of multimedia business bring a strong visual impact and higher requirement for wireless network services is required. However, the spectrum is limited and allocated fixedly nowadays. As a solution to this problem, cognitive radio can allocate spectrum dynamically and remarkably improve the spectrum utilization. In cognitive radio network, multi-channel selection and access for multi-user is a key technique, which allows secondary users (SU) to use the spectrum not utilized by primary users (PU) at this moment and return it to primary users when needed. This access method brings a new problem to the design of access algorithms:in distributed environment, more than one SU may choose one channel at the same time, leading to a collision among SUs and causing an interference for PUs. In order to solve this problem, this paper focuses on the algorithms and protocols of multi-channel selection and access for multi-user in distributed environment.Firstly, this paper gives a brief introduction of cognitive radio and its key technology, models of dynamic spectrum access as well as three models to solve the problem of spectrum selection and access, especially the Multi-Armed Bandit (MAB) model because of its further utilization. Secondly, for multi-user distributed network scenario, two classical multi-channel selection and access algorithms based on MAB model are introduced in detail. Then an algorithm based on TDFS and grouping is proposed. By simulation, it is seen that this algorithm can not only improve the fairness between multiple cognitive users, but also increase the accumulated channel access time of system. Finally, a multi-channel protocol based on channel hopping in distributed networks is proposed, which considers the negotiation between sender and receiver. This protocol contains negotiation phase and data transmission phase. In the negotiation phase, all the cognitive users negotiate according to the given channel hopping sequence. Once some SUs negotiate successful, they will transmit data immediately and continue to transmit in the data transmission phase. And other SUs, who fail the negotiation, have to wait for the next negotiation phase. Data transmissions in both negotiation and data transmission phase adopt channel selection and access algorithm based on TDFS and grouping. By simulation we can see that this protocol can not only improve the fairness between the multiple cognitive users, but also increase the accumulated channel access time of each cognitive user and improve the spectrum efficiency.