摘要
由于扫描跟踪雷达使用的天线波束在空间中的扩展特性,一个目标可被顺序扫描的波束或同时照射的堆积波束观测到;类似地,雷达信号模糊函数的扩展特性使得理想点目标的回波经过匹配滤波后的输出在时延-多普勒平面上是扩展的,这种空、时、频的扩展特性使得,与同一目标有关的多个波束的多个时延-多普勒采样单元都包含了目标信息,这种效应可以用采样单元的参数、目标参数以及模糊函数、方向图描述.传统的雷达信号处理对这些信息的利用非常有限,为此通过建立原始数据的幅度分布模型描述这些信息,并以此为基础研究基于雷达原始幅度数据进行跟踪的后验克拉美罗限(posteriorCramer-Raobound,PCRB).分析表明,与传统的先估计目标参数后跟踪的方式相比,利用这些信息进行跟踪可以改善跟踪性能.粒子滤波是实现这种跟踪方式的一种可行算法;针对传统实现方法中将回波幅度作为目标状态的一维并对之采样造成的运算量较大的问题,提出了一种避免幅度维采样的跟踪算法,仿真表明,该算法能以较低的运算量获得较好的跟踪性能.
Because of the spatial expansion of the beam utilized by TWS（track-while-scan）radar,a target can be illuminated by a sequentially scanning beam or stacked beams;similarly,the matched filer output of the echo from an ideal point target is extended due to the expansion of ambiguity function.The expansion in space and delay-Doppler domain leads to the fact that amplitude data in a specific region of the delayDoppler domain obtained by related beams are affected by the echo from the same target and can be characterized by sampling parameters,target parameters,ambiguity function and radiation pattern.This information,which is seldom utilized in traditional processing,is first modeled statistically,and then based on the model the theoretical tracking performances（gauged by posterior Cramer-Rao bound,PCRB） are intensively studied.It is shown that performance of tracking based on raw amplitude data is improved compared to tracking based on abstracted target parameters.Particle filter is a feasible algorithm for implementing this tracking scheme.In order to reduce the computation caused by sampling the extended state（signal amplitude）,an amplitude-sampling-free algorithm is proposed,and the simulations show that acceptable tracking performance can be achieved at a low computational cost with the proposed algorithm.
关键词
扫描跟踪
后验克拉美罗限
雷达原始幅度数据
粒子滤波
track-while-scan radar
posterior Cramer-Rao bound
raw amplitude data
particle filter
作者简介
作者简介：罗飞腾，男，1981年生，博士生．研究方向：雷达信号处理．E-mail：ftluo@mail．ustc．edu．cn
通讯作者：王东进，教授．Email：wangdj@ustc.edu．cn.
