Research On RSSI-based Indoor Wireless Localization Algorithms

In recent years, wireless sensor network is in rapid development stage. As thewireless sensor network technology matures, many related applications appear. Indoorpositioning is an application that is widely applied in military, industry and medical careand many other domains.RSSI based positioning has advantages of low cost, low-power consumption,supporting many nodes and so on. However, as the wireless signal is highly susceptible toenvironmental factors, positioning accuracy is not high. After analysising with actual data,it’s found that the assumption of general algorithms that the sigma propagation model isindependt and identically distributed, is no longer satisfied with the obstacles exsitingsituation. It’s necessary to screen nodes which are interferenced by obstacles. A newalgorithm based on the current RSSI-based positioning algorithm is proposed. By addingthe prior knowledge of the distribution of noise in signal propagation model, we get aposteriori probability based on Bayesian formula, and use the conjugate gradientalgorithm to solve the maximum a posteriori. The general algorithm doesn’t consider theobstacles interference. In this special situation, we add in the abnormal value judgment, byeliminating outliers to optimize algorithm. The result which is an estimate of the expectedvalue is a biased estimate, so the result should be amendedFinally, data is simulated in a variety of circumstances, and the real data is colletctedin lab of Huazhong Univesity of Science and Technology. The algorithm are verified bysimulated data and real data, and compared with several other RSSI based algorithms:maximum likelihood algorithm, weighted maximum likelihood algorithm and leastsquares, it showed that the proposed algorithm compared with other algorithm has betterpositioning accuracy.