The Research On Spectrum Allocation Algorithm Based On Graph Coloring In Cognitive Radio

With the rapid development of information technology, the demand for spectrum is growing rapidly. However, the channel resources which can be used for allocation are limited, this phenomenon has seriously hampered the development of wireless technology, which restricts the improvement of the utilization of channel resources.In order to solve this problem, scientists introduced the technology of intelligent cognitive radio(CR), which can learn and detect the surrounding information, judge the utilization of spectrum resource. If there are some idle spectrum bands, the SU can use these bands immediately. When a PU is detected to be connected to the channel again, the SU must exit immediately. This method greatly improves the utilization of spectrum resources and reduces the number of "spectrum holes", which effectively reduces the waiting time of secondary users.Firstly, this paper introduces the key technology of cognitive radio and some common spectrum allocation models, analyzes the common algorithm of graph coloring model, and gives the results on MATLAB platform.Secondly, this paper adopts an improved algorithm based on the MMR criterion of CSGC spectrum allocation algorithm, introduces the concept of efficiency levels and branch ideas to this algorithm, redefines the label criterion. Simulation results show that the improved algorithm based on MMR criterion of CSGC spectrum allocation algorithm has greatly improved the efficiency of the poor users, greatly reduces the allocation time, and improves the fairness of the algorithm.Finally, the study of spectrum allocation algorithm is mainly based on stable network. Once the network changes, it is necessary to update the topology map, and reallocation the spectrum. This algorithm reduces the communication efficiency of other users, and increases the time cost. In view of these facts, this paper introduces a spectrum reallocation algorithm based on connected component algorithm. The simulation results show that the algorithm can greatly shorten the allocation time and optimize the system performance.