Disorder-resistant fusion estimator design for nonlinear stochastic systems in the presence of measurement quantization

In this paper, we thoroughly study a new nonlinear disorder-resistant fusion estimation problem under both packet disorder and measurement quantization. Packet disorder, which occurs due mainly to random transmission delays (RTDs), are characterized via a series of integer-valued functions and random variables obeying certain distributions. Sensor measurements are logarithmically quantized before entering designated communication networks. The upper bounds on the filtering error covariances are acquired and later minimized by appropriately designing gain parameters of both local and fusion estimators. For fusion estimation, all local estimates and filtering covariances are integrated at a fusion center through the well-known federated criterion. Moreover, the fusion performance is examined by means of assessing the resistance of the designed estimator against the RTD-induced disorder. Simulation examples are presented to illustrate the applicability of the proposed disorder-resistant fusion estimator.