拟态物理学优化的认知无线电网络频谱分配

针对认知无线电网络中基于图着色模型的频谱分配问题，基于其非确定性多项式特性，以最大化网络收益总和为目标，提出了一种基于拟态物理学优化的求解算法. 在拟态物理学优化算法中，将频谱分配问题的解映射为一个具有质量的微粒，通过建立微粒的质量与其适应值之间的关系，并利用万有引力定律定义微粒间的虚拟作用力的大小，使整个群体向更好的方向运动，实现群体寻优. 给出了频谱分配问题的具体求解过程，并根据分配问题的二进制编码特点，改进了微粒的位置更新方程. 仿真实验表明：本文算法能更好地实现网络收益最大化. 关键词： 拟态物理学优化 认知无线电网络 频谱分配 网络收益

Sp ectrum allo cation of cognitive radio network based on artificial physics optimization

To study the spectrum allocation problem based on graph coloring model in cognitive radio network, an algorithm to maximize total network revenue is proposed, which is based on artificial physics optimization because of its NP-based features. In artificial physics optimization algorithm, the solution of spectrum allocation problem is mapped into a particle with mass. It establishes the relation between particle mass and its fitness value, and defines the virtual force between the particles by the law of gravity so that the entire group can move to the better direction and achieve population optimization. The detailed spectrum allocation process is given and the particle position updating equation is improved because of its binary coding features. Simulation results show that the proposed algorithm can better maximize network revenue.
Keywords： artificial physics optimization cognitive radio network spectrum allocation network revenue