Methods Of Active Multi-sonar Detection Fusion Based On Coordination Strategy

For dealing with the increasing demands on a variety of military and civilian applications, also for the reasonable use of multi-source information, underwater multisensor system has became hot, and the multi-sonar system is the emphasis. The monostatic sonar can complete the task in a certain area to a certain extent, but if several sonars can work together and share information with each other, they will be make the various performances of the whole system more superior in a larger geographic area, in more complicated external environment. In this paper, the problem we studied is the multisensor detection information fusion, and applying it to the multi-sonar system.The contents of this article can be roughly divided into three main parts, the first part mainly introduces the target detection methods for active sonar, and discusses the factors influencing the sonar detection performance, for example, with respect to the single model,we analysis the influence introduced by the pulse width of the signal and the bandpass width of the filter and the different methods for signal processing, in addition, we analysis the influence introduced by the class of the signal.In the second part, first of all, hard fusion decision method is studied, which is divided into two sections, the first section mainly for the sake of practical applicability, the decision rules of each local sensor are given, only to optimize the fusion rules at the fusion center to get the suboptimal solution. Then optimize the decision rules of local sensor and the fusion rule at the fusion center at the same time to get the global optimal solution. Futhermore, for the purpose of making better usage of information, we consider quantifying the local decision, and then transmitted to the fusion center, under the condition of the communication and information processing ability of the system is adequate. By this way,soft decision fusion is brought in. Otherwise, the factors influencing the multisensor system are studied.In the last part, the CFAR(Constant False Alarm Rate) target detection technology is introduced firstly, then illustrate its two basic type; CA-CFAR, OS-CFAR, and explain their application environment. Last, applying fuzzy logic to multisensor systems, and simulated the detection performances under different fusion rules.