基于加窗隐马尔可夫模型的窄带干扰抑制技术研究

An Improved HMM Technique for Suppressing Narrow-Band Interference

针对基于隐马尔可夫模型的窄带干扰抑制技术在直接序列扩频系统中存在频谱泄漏和收敛速度较慢的问题,提出增加窗函数并选择合适的HMM训练估计时隙数的改进算法,该方法在减弱"频谱泄漏"效应的同时,加快了算法的收敛速度.仿真实验表明改进算法对快变的强窄带干扰具有较好的抑制效果.

Aiming at the Narrow-band interference in Direct-Sequence Spread Spectrum,an improved Hidden Markov Model（HMM） technique for suppressing narrow-band interference is proposed.This algorithm used the window function and chose appropriate time batches of HMM parameter estimating.It was effective to reduce the spectrum leakage effect,decrease algorithm complexity and speed up the velocity of algorithm convergence.The simulation results show that the proposed algorithm can achieved system performance,especially the fast-changing and strong narrow-band interference.