基于改进SFLA-PF算法的OFDM系统目标跟踪  

Improved tracking targets SFLA-PF algorithm based on OFDM system

作　　者：曹向东[1] 毛永毅[1] 

机构地区：[1]西安邮电大学电子工程学院,陕西西安710061

出　　处：《电视技术》2016年第3期103-106,121,共5页Video Engineering

基　　金：陕西省自然科学基金项目(2014JM2-6088)

摘　　要：在OFDM通信系统中,为了解决非线性的目标跟踪问题,提出了基于改进混合蛙跳算法(SFLA)和粒子滤波算法(PF)相结合的方法来研究动态目标跟踪技术。首先利用高斯变异的局部搜索能力强和柯西变异的全局搜索能力强等优点对混合蛙跳算法进行改进,然后用改进后的混合蛙跳算法来优化粒子滤波算法进行动态跟踪,其优点不需要重采样步骤,有效地保持了粒子的多样性和有效性。仿真结果表明,该算法能够有效实现动态目标跟踪,并且跟踪效果优于同等条件下的混合蛙跳算法和粒子滤波算法。In the OFDM communication system, in order to solve nonlinear target tracking problem, an improved SFLA （SF- LA） and particle filter （PF） is proposed based on a combination of methods to study the dynamic target tracking technology. Firstly, Ganssian mutation local search ability and Canchy mutation global search ability, etc. are used to improve SFLA, then improved SFLA is used to optimize the dynamic tracking of particle filter algorithm, its advantages are not requiting resa- mpling steps, and meanwhile effectively maintain the diversity and effectiveness of the particles. The simulation results show that the algorithm can effectively achieve dynamic target tracking, and tracking better than SFLA and particle filter algorithm under the same conditions.

关 键 词：OFDM 混合蛙跳 粒子滤波 高斯变异 柯西变异 

分 类 号：TN929.53[电子电信—通信与信息系统]