System for optimizing mutliple-target identification methods

Differential measurement methods with array of Doppler radars are used for an estimation of target trajectory. Passive and active methods give different approaches. This paper deals with a multiple-target identification algorithm and error evaluating system for multiple-target tracking. The evaluating system is primary aimed at testing of tracking algorithms but it is also able to involve other investigations of factors causing errors (PSD peak detection methods, error dependency on SNR, DTFT overlapping, window length and shape). Presented track identification algorithm is trying to estimate the physical behavior of the target by using gradient movement prediction and by defined limits for special causes.