Risk-Based Non-Myopic Sensor Scheduling in Target Threat Level Assessment

Qing Dong,Ce Pang

doi:10.1109/access.2021.3078830

Abstract

The sensor scheduling problem under the target threat level assessment task was studied in this paper. Firstly, a target threat level model was established based on the Hidden Markov Model (HMM) and risk theory. Secondly, the target threat level assessment risk and the sensor radiation interception risk were combined to build the non-myopic sensor scheduling model. Thirdly, the improved branch and bound algorithm was introduced to obtain sensor management schemes from the objective function. Finally, simulations were made to illustrate the effectiveness of the proposed method and algorithm. The simulation results show that the proposed target threat level model can avoid the error transmitting from motion states of targets effectively and the improved branch and bound algorithm can achieve perfect sensor management schemes with good quality in a fast convergence rate.

Highlights

Radar, satellite, electronic supporting equipment and other kinds of sensors are applied to detect and track targets in the first step of a modern combat
The proposed improved branch and bound algorithm is shown in Tab
The non-myopic sensor scheduling method with target threat level estimation and sensor radiation risk taken into consideration has been researched in this paper

Summary

INTRODUCTION

Satellite, electronic supporting equipment and other kinds of sensors are applied to detect and track targets in the first step of a modern combat. The risk-based non-myopic sensor management method combined with the former target threat level model is further proposed. ANALYSIS ON THE RELATIONSHIP BETWEEN HMM AND TARGET THREAT LEVEL ASSESSMENT PROCESS Suppose that the variable x(k) is the state of one system at time instant k, and the state at time instant k + 1 is only related to the former adjacent state, there is: p (x(k + 1)|x(k), x(k − 1), .

3) METHOD TO ESTIMATE TARGET THREAT LEVEL

BRANCH AND BOUND ALGORITHM

SIMULATION

CONCLUSION