下行卫星认知无线电门限与功率联合优化

用于卫星通信的频谱资源日益紧缺,但现存卫星系统却存在利用不足的问题。针对该问题,以最大化卫星通信中下行带内数据传输量为目标,提出了基于Underlay认知无线电的卫星信道检测门限与功率分配联合优化算法。首先分析了下行链路中信道融合检测误差、功率分配向量与数据传输量之间的数学关系,给出了信道数量及总功率均受限时信道与功率联合分配的可行性条件。之后,进行了目标模型可行性论证及融合误差分析,并将优化目标函数问题分解为检测门限与功率分配两个子优化问题。针对两个子问题,文中分别证明了加窗粒子群优化算法适用于数据融合后的门限优化,论证了信道与功率联合分配的可行性,在此基础上给出了信道与功率联合分配的最优解。引入中间量在两个子优化算法之间反复迭代,得到了本文目标函数的联合最优解。最后,仿真给出了检测门限优化性能,结果表明,改进型算法在准确度和迭代次数上均优于传统粒子群算法;比较了该联合优化算法与传统卫星通信方式及普通Underlay认知无线电用于卫星通信时的性能差异,结果表明,该联合算法能够有效提高频谱利用率;给出了算法复杂度。

Downlink joint optimization of detection threshold and power allocation for satellite cognitive radio

The frequency resuorce for satellite communication becomes rare while spectrum utilization is inefficient. To raise such a spectrum utilization efficiency, a joint optimization of detection threshold and power allocation based on Underlay cognitive radio is proposed. Firstly, the mathematical relationship among detection fusion errors, power allocation vectors and data throughput is analyzed. Correspondingly, feasibility conditions for joint allocation of channel and power are shown with the limited of channel number and transmission power. After feasibility argumentation and fusion error analysis, the joint optimization is divided into detection threshold optimization and optimal power allocation. The suitability of windowed particle swarm optimization for the downlink detection threshold is proved. The optimal solution of joint allocation of channel and power is found theoretically based on the feasibility discussion result. To get the joint optimization of detection threshold and power allocation, an intermediate variable is introduced which performs iteration between two sub arithmetic repeatedly. Simulation results show the superiority of the proposed windowed particle swarm optimization over traditional methods in tems of accuracy and iteration times. Comparing the performance among the common satellite communication, Underly cognitive radio and the joint optimization, the comparison results demonstrate the throughput improvement of the joint optimization. The computation complexity is also presented.