基于最小平方误差的辐射源识别

Emitter recognition based on Least Square Error model

摘要:

针对测量参数不确定所造成的辐射源不能正确识别问题，提出一种基于最小平方误差模型的识别方法。该方法首先计算辐射源的特征参数相对于数据库中相应特征参数的平方误差，然后利用最小平方误差来确定辐射源隶属度。该模型与几何距离度模型、云模型相比，在识别辐射源类型时，增加了雷达特征参数(重复频率、脉宽和天线扫描周期)的多值性和相关性，仿真结果表明，该方法的正确识别率比前2种方法分别高出5%和40%。

Abstract:

To tackle with the parameter uncertainty in emitter recognition problem, a recognition method based on least square error model is presented. First, the square error of characteristic parameters is calculated relative to those in database for the emitter, then, the degree of membership of the emitter is determined based on least square error. Unlike geometrical distance model and cloud model, the proposed model takes multi-values and correlation of each parameter(PRF,PW and ASV) into account. Simulation results show that the emitter recognition rate of this model is higher than that of the geometrical distance model and cloud model by 5% and 40% respectively.