Constant False Alarm Detection Method Adaptive to Environmental Changes

Abstract

The invention discloses a constant false alarm rate detection method adaptive to environmental changes. Correlation of corresponding positions of multiple periodic data and correlation of adjacent data of the same group of data are used for threshold detection, that is, time correlation and space correlation are used for threshold detection, and an interference memory group is initialized respectively to be spectrum information of first periodic data through an initialized part; clutter information of the current data is estimated to be a result when clutter information in former M periods and the spectrum information of the current data are weighted; if the located environment changes greatly or quickly, the weight of the current data information on clutter estimation can be properly increased; and the currently-estimated clutter information is used for setting of a later constant false alarm rate detection threshold. According to the method of the invention, in homogeneous background, a unit average constant false alarm rate is used to achieve optimal performance, and in nonhomogeneous background, an ordered statistical method is used for enabling the detection performance and the false alarm performance to be significantly improved.

Description

technical field [0001] The invention relates to a constant false alarm detection method, in particular to a constant false alarm detection method adaptive to environmental changes. Background technique [0002] There are many existing CFAR detection algorithms, such as unit average, average selection large, average selection small, ordered statistics, clutter map method, etc., and they all have their own scope of application. The unit average has high requirements on data characteristics, mainly based on the following two assumptions: a. The target is independent; b. All interference data samples in the reference window are independent and identically distributed, and are identically distributed with the interference in the unit containing the target . The main purpose of ordered statistics is to suppress the performance deterioration caused by the shadowing effect, but the false alarm rate of ordered statistics is higher. The clutter map method requires prior knowledge of...

AI Technical Summary

Problems solved by technology

[0003] The performance of the unit average is seriously degraded in the non-uniform background, the detection performance of the large average selection is very serious in the multi-target environment, and the detection performance of the small average selection is poor in the uniform clutter background and the false alarm performance in the clutter edge.

Images