Bi-objective particle swarm optimization algorithm for the search and track tasks in the distributed multiple-input and multiple-output radar

The resource allocation strategy plays an important role in the multifunctional performance of distributed multiple-input and multiple-output radar (D-MIMO-R) systems. In applications, different functions may be conflicting, due to the competition of the same resource. In this paper, a bi-objective hybrid particle swarm optimization (BHPSO) algorithm is proposed for the simultaneous optimization of the search and track functions, under the constraint of limited active subarrays and power budget. In the BHPSO, the nonlinear time-varying coefficients are used to balance the exploitation and exploration abilities in different stages. The heuristic mapping scheme is designed to cope with the constraints. The crossover and mutation operators are devised to the break the swarm stagnation. The distance based crowding function with the local guider scheme is incorporated to preserve the swam diversity. In addition, by exploiting the unique structures of the objective functions, we fully prove that the solutions, each of which corresponds to a specific power budget for the search or track function, have a proportional relationship. Thus, the best-known Pareto subset (BKPS) can be obtained by parallel solving two convex optimization models only once. Extensive simulation results show the effectiveness and efficiency of the proposed BHPSO, in comparison with the state-of-the-art algorithms. The statistical results also indicate that the target reflectivity is a main influence factor on the resource allocation, and the BHPSO can provide competitive results on standard test functions. (C) 2020 Elsevier B.V. All rights reserved.

Automated summary

The proposed bi-objective hybrid particle swarm optimization algorithm aims to optimize the search and track functions simultaneously in D-MIMO-R systems. By balancing exploitation and exploration abilities, designing a heuristic mapping scheme to handle constraints, and using a distance-based crowding function to preserve swarm diversity, the algorithm proves to be effective and efficient compared to state-of-the-art algorithms.