基于HMM信道预测的认知无线网络能效优化研究

摘要

针对Overlay频谱共享模式下的认知无线网络能量优化问题,利用隐马尔可夫(Hidden Markov Model,HMM)模型建立信道预测-择优感知-接入机制.认知用户在信道感知阶段利用HMM模型进行下一时隙信道状态(忙绿/空闲)预测,根据信道预测结果进行信道择优感知-接入,降低信道感知阶段的冗余能量消耗,提升信道接入准确率和网络吞吐量,从而有效提升网络能量效率.与采用传统随机接入模式和BPNN预测模型的无线网络进行仿真对比,结果表明,采用的HMM信道模型预测准确率达91％以上,且网络能量效率得到有效提升.%Aiming at the problem of energy optimization in CR (Cognitive Radio) network in overlay spectrum sharing mode, HMM (Hidden Markov Model) is applied to establishing the channel prediction-optimal perception-access mechanism. The CR user uses HMM model to predict the next slot channel state (busy/ idle) at the channel perception stage. And based on the channel prediction results, the channel selective perception-access is implemented, the redundant energy consumption in channel sensing stage reduced, the channel access accuracy and network throughput raised, and thus the network energy efficiency effectively improved. Simulation and comparison with wireless networks with traditional random access mode and BPNN prediction model are done. The results indicate that the accuracy of HMM channel prediction is over 91％, and the network energy efficiency effectively improved.