A Distributed Detection Fusion Algorithm Based on Adaptive Threshold
Wei Xiang · Jisuanji fangzhen · 2009
The detection fusion under Bayes criterion is a traditional method of distributed detection fusion.The prior probability of the unknown phenomenon,the false alarm probability and the miss probability of the local sensors should be given in the above-mentioned method,but these statistics are unknown or variable in the practical applications.Therefore,a distributed detection fusion algorithm based on adaptive threshold is studied under the Neyman-Pearson criterion.The algorithm can adjust the threshold of the local sensor on-line,and improve the performance of the fusion center by the local sensor detection optimal.The results of computer simulation show that the algorithm can converge rapidly,and the performance of the fusion center is improved obviously compared with the local sensor's.