As a multi-dimensional complex system, Cyber physical systems (CPS) integrates computing, network, and physical environment. How to effectively fuse the local detection results is the key to improve the sensing performance in CPS. In order to improve the spectrum sensing performance of cognitive radio, a soft fusion-based cooperative spectrum sensing using particle swarm optimization (PSO) for cognitive radio networks is proposed. To find the optimal weighting coefficient of soft fusion, the traditional PSO algorithm is investigated and improved. To further reduce the local convergence of PSO and accelerate the convergence speed of the algorithm, the immune algorithm is introduced and chaotic sequence mechanism is applied to ensure the diversity of the particles. Moreover, the inertia weight is adjusted adaptively according to the convergence of particles so as to maintain the tradeoff between the search ability and the convergence of particles. Simulation results show that the proposed algorithm can achieve fast convergence in CPS. Compared with other typical methods, it can obtain better sensing performance under different signal-to-noise ratio environments and the constraint of false alarm probability.

中文翻译：

---

基于软融合的协同频谱感知，使用粒子群优化，用于信息物理系统中的认知无线电网络

信息物理系统（Cyber​​ Physical Systems，CPS）作为一个多维度的复杂系统，集计算、网络和物理环境于一体。如何有效融合局部检测结果是提高CPS感知性能的关键。为了提高认知无线电的频谱感知性能，提出了一种基于软融合的认知无线电网络粒子群优化协同频谱感知。为了找到最优的软融合加权系数，对传统的PSO算法进行了研究和改进。为了进一步降低粒子群算法的局部收敛性，加快算法的收敛速度，引入免疫算法，应用混沌序列机制保证粒子的多样性。而且，根据粒子的收敛性自适应调整惯性权重，以保持搜索能力和粒子收敛性之间的折衷。仿真结果表明，该算法能够在CPS中实现快速收敛。与其他典型方法相比，它在不同信噪比环境和虚警概率约束下都能获得更好的感知性能。