Research On Dynamic Spectrum Planning In Cognitive Radio Networks

With the rapid development of the information technology,it has been a boom in the number of communication equipment which occupies the frequency bands.The contradiction between limited spectrum resources and souring demand of spectrum has become more and more irreconcilable.In the military field,the spectrum has transferred from merely a “media” of communication to a combat domain and become an extremely important national strategic resource.As a branch of communication planning,spectrum planning was proposed to standardize the occupation order of communication equipment,improve the utilization efficiency of frequency,alleviate collisions and interruptions resulting from the competition of spectrum resources among communication equipment and enhance the quality of communication.However,the traditional static spectrum planning is not suitable for the communication under the tactical environment.The Cognitive Radio(CR)technology has enabled the communication equipment to autonomously reconstruct communication parameters based on the sensed environment information.To a certain extent,there is no difficulty in configuring and deploying a real-time and dynamic spectrum planning strategy on a common hardware.Under this situation,the research on dynamic spectrum planning in cognitive radio networks has been an urgent practical demand.Based on the communication requirements of tactical CR networks and combined with the actual deployment of the spectrum planning strategy on CR equipment,in this paper,we mainly study the dynamic spectrum planning problem in CR networks.The main contributions are summarized as follows:1.Dynamic spectrum planning of dual-hop relay based on non-exclusive relay nodes under constrained optimization model.In joint tactical communication under the battalion level,we proposed a division and reuse scheme in VHF frequency band aiming at different tactical communication networks.Further,based on this scheme,we put forward a relay frequency scheme at a non-exclusive relay node,a company node and realize the dual-hop communication between the battalion node and another company node.Then,under the constrained optimization model,we take the demand of relay,the co-site interference between the relay and other networks in the original communication system and the reuse of the spectrum band into consideration comprehensively.Ultimately,the frequency usage of relay is optimized.The communication range of the battalion node has been greatly extended.The spectrum planning of relay under this model is self-controlled by each relay node instead of a central control node.In other word,the cognitive relay company node can self-optimized its own relay spectrum band according to the relay demand,its communication environment and original communication task.It will be convenient for nodes to decide the relay strategy of spectrum to adjust to the change of relay nodes.2.Spectrum optimization of a mobile secondary user based on statistical collision performance analysis under stochastic geometry model.Considering a mobile cognitive secondary user newly access to a communication system,reusing the communication band of primary users in this system.Due to the mobility of the cognitive secondary user,the interference constrains between the secondary user and primary users change from time to time.Traditionally,the spectrum optimization based on the time slots cannot moderate the contradiction between the computational complexity and real-time planning request.We change the traditional constrained optimization model into the stochastic geometry model to analyze the frequency reuse scheme of the cognitive secondary user.Based on the proposed model,the statistical closed expression of average collision time in the No Talk Zone of primary users under the same spectrum due to the motion of the secondary user node is derived.According to this value,the frequency reuse scheme of the secondary user node is optimized.3.Research on spectrum access strategy based on Q reinforcement learning for anti-dynamic jamming.Under dynamic malicious jamming of spectrum throughout the entire communication band,a distributed multi-agent spectrum access scheme for anti-dynamic jamming is proposed.This scheme enables cognitive radio nodes to predict and evade the dynamic jamming based on the information sensed.Further,according to the information from transmission surveillance modules of these nodes,the benefit of spectrum access strategy can be fed back.Based on the benefit,the Q-table of the reinforcement learning is update to reduce the collision of spectrum usage that more than two CR nodes use the same spectrum for communication.The performance of these CR nodes to cope with dynamic malicious spectrum jamming is improved.To some extent,an intelligent spectrum access optimization of node group for anti-dynamic jamming is achieved under the proposed system.