Research on Multipoint Mobile Network Positioning Technology Based on Bluetooth Mesh

Taking the large-scale dynamic positioning in the intelligent livestock industry as the demand background, in order to reduce the cost of all equipped with GNSS positioning and 4G communication chips, the research changes the location information source of some devices to Bluetooth positioning method and uses Bluetooth Mesh for networking communication and RSSI dynamic ranging positioning. In the RSSI ranging model, the Unscented Kalman Filter (UKF) is used to process the RSSI signals, which improves the accuracy by about 10% compared to the hybrid like DGMM and KGMM. For positioning, an optimization method is introduced for the influencing factors in the real environment such as time difference and distance, and the Barzilai-Borwein gradient descent method is used to solve the problem. The error of this method depends on and exceeds the positioning error of the GNSS anchor nodes in the network, which can be reduced by about 30% compared to the ML positioning method in the application context of this paper. The new weighting function likewise improves accuracy by 3%.