Multi-objective time-frequency difference parameter sorting technology based on adaptive multi-mode particle filter

Abstract

Modern new system radar has high-speed and flexible beam scheduling capability, and the widely used high repetition rate design of signals leads to a large number of uneven TDOA ambiguity points in passive TDOA estimation, which is difficult to locate accurately. The ambiguity of TDOA is caused by the pulse pairing error of HPRF radar signal in TDOA estimation. A large amount of position information is still retained in the ambiguity, and simple removal of false locating points may lead to the breakage and loss of target track. According to the PRF group variation characteristics of PRF packet radar signals, this paper proposes a multi-target time–frequency difference parameter sorting technology based on adaptive multi-mode particle filter. The technology unifies the time difference ambiguity and multi-objective problem under the Bayesian reasoning framework, adopts multi-modal particle filter method, and uses a posteriori probability recursive estimation to achieve time difference ambiguity resolution and parameter sorting. Through multi-dimensional probability association, we can realize the association, rebirth and extinction of goals. The simulation verifies the feasibility of the proposed method, which can still maintain a high accuracy of ambiguity resolution and target sorting under the condition that five targets exist at the same time and the TDOA curve intersects.