12.1 INTRODUCTION

The concept of employing multiple sensors with data fusion is widely used in surveillance systems. For a large area of coverage and/or a large number of targets under consideration, a number of geographically separated receivers may be used to monitor the same volume in space, as shown in Figure 12.1. In such space diversity systems, complete observations can be transmitted by the sensors to a central processor for data processing. Diversity systems are more robust and more reliable than single sensor systems. However, the enhanced performance of these systems is essentially derived from the diversity of the system configuration, at the expense of a required large communication bandwidth between the local receivers and the central processor. Thus, due to the constraints on the bandwidth of the communication channels, distributed signal processing with a data fusion center is preferred in many situations. In such distributed detection systems, some processing of the signal is done at each sensor, which then sends partial results (compressed data) to the central processor, or in the context of distributed detection, to the data fusion center, as shown in Figure 12.2. These partial results are combined according to a suitable data fusion rule to yield the desired global result. In our case, the partial results are decisions from the individual detectors, D _i, i = 1, 2, , N, where D _i ? {0,1}. The values of D _i are combined to yield a final decision, D ₀,...