Abstract
In this paper, a fusion detection algorithm that focuses on decentralized CFAR (Constant False Alarm Rate) signal detection problem without prior information is proposed. In the algorithm, the threshold and test statistic of the detection fusion algorithm derive from the conventional CFAR detection method. At last a framework for decentralized CFAR signal detection is designed corresponding to the fusion algorithm. Simulation results illustrate that an almost optimal detection performance is obtained by the proposed algorithm.
Highlights
It is a significant problem to fulfil detection in distributed information fusion system
With the development of communication ability, a means of double threshold detection arises for distribution detection [9,10,11]; it must be known that the signal noise rate (SNR) or the detection performance of every detector for reaching the optimal performance [12,13,14]
We offer a series of C with the change of the Pfa in multi-element detection fusion as figure 3 and table 1: 10-2
Summary
It is a significant problem to fulfil detection in distributed information fusion system. Conventional method makes the detection fusion with the entire detectors’ result in which a little quantity of information can be used. It restricts the detection centre’s performance [4,5,6,7]. If making fusion information directly without the prior knowledge (SNR), the result must lead to a decrease performance for the entire system. There are some detection fusion methods without prior information. The number of article is rather less and their performance is unsatisfying in the detection fusion without prior information. It’s significant to make the research for the decentralize detection
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
