Communication-constrained cooperative navigation based on quantized observation

The advantages of Cooperative Navigation (CN) stem from the observation sharing. However, the applicability is constrained by the available communication bandwidth. Exploiting fewer bits representing the real-valued observation is an alternative approach. Maximum A Posteriori (MAP) and Minimum Mean Square Error (MMSE) estimators for the quantized version estimation and the state parameter estimation are employed respectively. Specially, in consideration of the communication channel noise, the MAP estimator, where the quantized versions are regarded as a series of discrete independent events, is developed as a hybrid estimator with respect to the discrete and continuous variable. The simulation results show that with the number of quantization bits increasing, the more accurate quantized version estimation is acquired. Thus, the state parameter estimation results are also closer to the case based on real-valued counterparts as inputs.