Search

In this study, a coherent detection method is proposed to solve the problems during the coherent integration (CI) process, including the range cell migration (RCM) and Doppler spread (DS) effects which are caused by the complex motions of the maneuvering target. Particularly, this method applies the third-order keystone transform firstly, which could remove the third-order RCM induced by the jerk motion. Then target's defined velocity can be estimated by performing the frequency-shifting cross-correlation function and scaled Fourier transform. Thereafter, utilising the estimated defined velocity to construct a velocity compensation function, which could correct the first-order RCM (FRCM). Subsequently, the matched filtering process is performed to eliminate the second-order RCM and DS effects. Finally, applying the Fourier transform with respect to the slow time to realise the CI.

Considering a multi-stage stochastic task, in which a multi-static radar system (MSRS) is applied to assist with missile interception, the authors study an optimisation problem of radar resource management. Specifically, under restriction of a fixed energy budget, the authors devote to minimise the loss, which is caused by the unsuccessfully intercepted missiles, through optimal power allocation (OPA) of MSRS within multiple stages. The design of OPA can be translated into a sequential decision-making problem. The authors formulate the problem through variable definition and modelling the missile interception procedure. As the authors need to consider the randomness of multiple coupled stages and jointly allocation power between multiple radar nodes, to solve the proposed problem is of huge computational load. The authors propose a solution that combines with reinforcement learning and particle swarm optimisation. Comparing with the uniform power allocation scheme, the simulation results demonstrate that the OPA scheme designed by the proposed method is capable to achieve preferable and more stable performance for the whole missile interception. The authors' contributions include a novel optimisation resource management model for a multi-stage stochastic task and an effective solution for the optimal resource management scheme.

The authors consider an optimisation problem of multistatic radar system (MSRS) deployment, which includes both the antenna placement and the transmitted power allocation. To improve the surveillance performance of MSRS that equipped with different detection methods, they mainly aim at two goals: (i) to improve the coverage ratio of a surveillance region; (ii) to get an even distribution of signal energy in the surveillance region. Through introducing two objective functions for the above two mentioned goals, respectively, they formulate a multi-objective optimisation problem. To overcome drawbacks caused by the significant difference between the values of these two objective functions, they propose a multi-objective particle swarm optimisation (MOPSO) algorithm with non-dominated relative crowding distance (MOPSO-NRCD). Specifically, through the non-dominative relationship comparison and normalisation of crowding distance, a novel method is proposed to select the global best solution. Moreover, a multi-swarm structure is applied to labour division of all particles. Comparing with the MOPSO with the traditional crowding distance, simulation results show that the MOPSO-NRCD can provide better candidate schemes, which are capable to satisfy more stringent performance requirements.