Scheduling of Multisensor for UAV Cluster Based on Harris Hawks Optimization With an Adaptive Golden Sine Search Mechanism

Multisensor scheduling, a multi-objective optimization problem, simultaneously aims to find the optimal matching between sensors and targets in a certain period, that is, to generate the optimal multisensor scheduling scheme. Several sensor scheduling literatures proposed multisensor scheduling schemes incorporating various algorithms, but the majority of them dealt with sensors to targets in two dimensions. First, the visual relationship between sensors and UAV cluster targets is investigated, and a three-dimensional correspondence between sensors, targets, and time is designed while dividing the scheduling time. Secondly, the detection advantage of sensors to UAV cluster, sensor load, and sensor handover numbers in the scheduling time interval are summed to create multisensor scheduling objective functions for UAV cluster. Furthermore, for the UAV cluster optimization problem, a hybrid optimization algorithm is proposed, which integrates the golden sine mechanism and adaptive weight factor in Harris hawks optimization algorithm. Finally, extensive experimental and statistical analyses show that the models and algorithms proposed in this paper outperform the existing ones. It has better convergence, does not need additional computational cost, and generates better multisensor scheduling schemes.

I. Introduction

Mutisensor networks play an essential role in traffic surveillance, scene analysis, air traffic control, and battlefield reconnaissance [1]–[4], especially in the UAV cluster operation [5], which has attracted extensive attention. Many practical problems necessitate the use of a series of sensor network measurement data to estimate the system’s state; for example, in the course of combat, various types of sensors are used to achieve continuous detection and stable tracking of UAV clusters. The problem of selecting appropriate sensors to detect and track UAV cluster targets is known as a multisensor scheduling problem [6], [7], and it has long been a research focus.