Quantizer design for distributed detection based on bhattacharyya distance

For distributed detection, quantizer design is an important issue to decrease communication bandwidth and meanwhile to achieve a good detection performance, when the communication bandwidth between local sensors and the fusion center is constrained. In this paper, we study a quantizer design method under the Swerling I target model, where the quantizers of the sensors are designed by maximizing the Bhattacharyya distance of the fusion center. The modified sequential quadratic programming (SQP) algorithm is used to solve the quantizer design problem. In the fusion center, randomized test strategy is considered to achieve the optimal detection performance and a desired false alarm probability. The detection performance of the quantization method is evaluated by the numerical results.