Distributed Spectrum Access With Context-aware

With the increasing development of wireless networks and wireless users,cognitive radio spectrum resources are increasingly scarce, the fixed spectrum access can not meet the grim situation of lacking spectrum resources, the dynamic spectrum access emerged, however, compared with centralized spectrum access, distributed spectrum access has advantage about no centralized control unit, good robustness and low complexity, therefore distributed spectrum access is the popular direction of the mainstream research about cognitive radio in the future. In the scene of distributed spectrum access, the main problem is how to share the primary user-provided set of channels available between cognitive users, relating to parameters about the position of cognitive users, channel conditions, the probability of cognitive users contending for the channel and so on. But in the most existing research, the interference is mutual affects among all users, tending to ignore the impact of user location, user behavior,user demand and other communications environment(context) properties, however,these issues have great significance on mining and upgrading spectrum resources utilization, therefor distributed spectrum access game mode with context-aware is proposed to select distributed channel based on user location, user behavior and other context properties. the main work and innovations of this thesis is as follows:1、We generalize spectrum spatial reuse strategy based on cognitive user location information, considering relative space position of cognitive users may providing different opportunities of spatial reuse, in which the interference diagram of spatial position describes the interference relationship of neighboring users and measures interference tolerance, the channel state is associated with a rate constraint, then game problems between cognitive users will be modeled to maximizing throughput for finding optimal channel selection set. And the existence of Nash equilibrium is proved in the theory. The reasonable and effective about strategy is verified by simulation, the results show that spatial reuse strategy based on the user location significantly improves system throughput and spectrum utilization.2、On the basis of spatial reuse strategy, considering business needs desires and satisfaction about competing channel among cognitive users in a heterogeneous network, leading into parameter of competition channel probability, presenting distributed spectrum access strategy and game model based on user behavior, in which minimizing the loss throughput of cognitive network to solve the optimal channel selection set that is Nash equilibrium, through theoretical analysis the existence of Nash equilibrium is proved. Simulation results show that consider user behavior and satisfaction, the premise of ensuring user QOS needs, spectrum sharing strategies is proposed to improve the system throughput and user satisfaction.3 、 For solving Nash equilibrium in the strategy random selection algorithm,selecting learning step has a reverse effect on the convergence speed and performance,an adaptive step stochastic learning algorithm is proposed, the method adaptive update learning step by laziness factor in iteration, and the convergence of the algorithm is proved. Simulation results show that: compared to the fixed step,adaptive step stochastic learning algorithm can has an effective balance between system performance and convergence speed.