Localization Scheme Based on Local-Adjusted Data-Driven Statistical Error Model

Statistical error modeling techniques have been developed for acoustic signal detection-based ranging measurements in wireless ad-hoc sensor networks (WASNs), and some localization schemes are proposed based on it. However, from previous studies, we know that the scheme still has some shortcomings. Therefore, a new localization scheme based on local-adjusted data-driven statistical error model is proposed in this paper. Compared with the old one, the new one has the following characters: (1) it makes full use of that the advantage that error mode is the most accurate when the measured distances are between 15m and 35m, (2) it can adjust its error model to local environment, (3) it scales better by dividing its network into sub-networks. From the simulation and experiment results, we know that the new one improves its performance further.