Quantized measurement fusion in wireless sensor networks with correlated sensor noises

The quantized measurement fusion problem for target tracking in wireless sensor network (WSN) with correlated sensor noises is investigated. Each sensor node quantizes the local measurements using probabilistic quantization strategy and transmits the quantized measurements to a fusion center (FC). The FC estimates the target state in a dimension compression way according the best linear unbiased estimation (BLUE) fusion rule instead of merging all the quantized messages to a vector (augmented scheme). Focuses are on tradeoff between the communication energy and the global tracking accuracy. A closed-form solution to the optimization problem for bandwidth scheduling is given, where the total energy consumption measure is minimized subject to a constraint on the mean square error (MSE) incurred by the BLUE fusion. Nonlinear Gaussian discrete-time system model following the sigma-point Kalman filtering (SPKF) principle is adopted. Simulation example illustrates the proposed scheme obtains average percentage of communication energy saving up to 93.9% compared with the uniform quantization, while keeps computational burden reduction 35% compared with the augmented scheme.