An Environment-Adaptive Multi-Channel Ranging Optimization Algorithm Based on a Multi-Objective Evolutionary Model for Multipath Wireless Sensor Networks

Recently, high-precision WSN (wireless sensor network) ranging and positioning algorithms based on RSSI (Received Signal Strength Indicator) in complex indoor environments have become a popular research topic. This is because RSSI is easy to obtain and more suitable for the large-scale deployment of WSNs. However, WSN ranging and positioning algorithms using RSSI are severely affected by the presence of noise and multipath effects in complex indoor environments. To reduce multipath effects, a multi-channel ranging algorithm was developed. This algorithm must obtain accurate initial parameter values or the target-reference distance in advance; otherwise, it will fall into local optima. We propose an environment-adaptive algorithm for multi-channel ranging optimization based on an innovative evolutionary model with multiple objectives and an existing adaptive extended Kalman filter. This novel model includes a newly created objective function of the relationship between weighted multi-channel RSSI and node distance, which allows it to achieve globally optimal results without requiring extensive training to obtain accurate initial parameter values or the target-reference distance beforehand. Extensive simulations and experiments show that our proposed algorithm always has much higher ranging accuracy than the existing algorithm, regardless of whether the multi-channel RSSI is regular or the number of paths matches.