低S/N目標の追尾開始におけるMHTの性能解析

摘要

In radar tracking, detection threshold must be lowered under low SNR environments, to maintain a certain target detection probability. However that increases the number of false alarms concurrently, which occur mis-associasion in tracking. The method, using Multiple Hypothesis Tracking(MHT) to avoid mis-association with the target track and false alarms, is well known. To utilize limited power effectively, we have to optimize observation interval and signal SNR in one scan, which have trade-off relationship. However Monte Carlo simulation, which consists of thousands of trial number, is needed for the optimization. In this paper, we propose the performance prediction method, which analyzes the probabilities of track states in the initiation process.%レーダを用いた低S/N　目標の追尾開始において，検出閾値を下げて探知性能を確保し，相関決定にMHT（Multiple Hypothesis Tracking）を適用して誤警報を抑圧する方式が知られている．この方式では単位時間当りの信号電力に応じて，追尾開始性能を最適とするサンプリング間隔が存在する．しかし最適なサンプリング間隔の特定では数千試行に及ぶモンテカルロシミュレーションを実施している．この課題に対して，本稿では追尾開始性能を最適にするための，解析手法を用いたサンプリング間隔の設計手法を提案する．