基于能量检测的异步感知算法

Asynchronous perception algorithm based on energy detection

在未来异构无线网络中,授权用户(PU)与认知用户(SU)间无法进行协作定时,导致授权用户发射机和认知用户接收机之间存在感知时间差。针对这一异步感知场景,基于贝叶斯统计估计理论提出一种全新的异步感知算法。首先,提出一种统一的动态状态空间模型,来描述可观测能量与动态授权用户状态以及未知时间差之间的关系;然后,利用随机有限集并基于最大后验概率准则设计一种迭代式估计方案;最后,通过粒子滤波以数值逼近方式得到估计结果。仿真结果表明,通过准确获取出感知时间差,所提出的异步感知算法可有效消除接收信号的信息不确定性,从而显著提高频谱感知性能。

In the future heterogeneous wireless networks,since primary user（PU） and cognitive secondary user（SU） are not coordinated to be synchronous,it will result in sense timing difference between PU＇s transmitter and SU＇s receiver.For this asynchronous sense case,a new asynchronous sensing algorithm based on Bayesian estimation theory was proposed.A unified dynamic state space model was first proposed to describe the observable energy relationship with dynamic PU state and unknown timing difference.Then,an iterative estimation scheme was designed using stochastic finite set and the rules of maximum posterior probability.Finally,approximated estimation results were obtained by using a particle filter.The simulation results show that the proposed asynchronous scheme significantly eliminates the uncertainty of the received signal information and thus improves the spectrum sensing performance by obtaining the time difference accurately.