Research Of Dynamic Spectrum Allocation Algorithm In Cognitive Radio Based On The Connected Components

Cognitive radio, an intelligent spectrum sharing technology, aimed at automatically searching and opportunistically utilizing the temporally unused licensed spectrum to improve spectrum utilization by detecting the surrounding wireless electromagnetic environment of the frequency domain, time domain and airspace domain, is regarded as the most suitable way to solve the increasingly scarcity of spectrum resource. Dynamic spectrum allocation is one of the most important and challenging technologies to implement cognitive radio under the constraint not to interfere with licensed users. Dynamic spectrum allocation technology is mainly dealt with solving how to share spectrum resources flexibly, efficiently, and non-interference. However, there is still large distance between the existing dynamic spectrum allocation algorithms and the high efficiency, fairness, speed goals of spectrum sharing, owing to the open and dynamic nature of the available spectrum in cognitive radio system. Therefore, it is become a hot researched problem to effectively assign spectrum dynamically over the world. Based on connected components theory, uniting demand factor, the dynamic spectrum allocation algorithms are researched in this thesis. The main researches of this thesis are as follows: (1) A new method of processing all connected components in parallel for allocating spectrum dynamically is proposed, through drawing graph connected components theory into spectrum allocation. After dividing the system total topology by using connected theory, the spectrums are distributed through processing each connected branches in parallel. Using this method to allocate spectrum, not only the distribution rate will be improved effectively, but also the original system performances can be maintained. This method, which can be applied to all spectrum allocation algorithms that are based on graph coloring model, is the effective way to solve the bottleneck problem of too much time taken in spectrum allocation.(2) A dynamic spectrum allocation algorithm based on the proposed method of parallel processing connected components is researched and simulated, for the special purpose of verifying the feasibility and effectiveness of that method. The research results show that the proposed method is fast, active, and adapting to the rapid time-varying cognitive radio environment better.(3) A satisfactory comprehensive performance of adaptive spectrum assignment algorithm is researched, aiming at solving the problems of second waste of spectral resources and distribution at a too slow speed. The spectrums are allocated in the principal way of processing connected components in parallel, under the condition of considering users’ demand factors. The simulation results show that both the system’s demand satisfaction and spectrum allocation speed are improved, while spectral efficiency is ensured, which means that the second waste of spectral resources is avoided as well as the problem of large time taken in spectrum allocation is solved efficiently by this algorithm.