Indoor location algorithm research based on neural network

Most of the traditional indoor location algorithms based on the distance loss model always filter the received signal strength, and then we can use the distance loss model to infer the distance between the nodes and achieve location eventually. The accuracy of the traditional indoor location algorithm is very unstable due to multipath propagation effects and complex signal attenuation law in the indoor environment. On the basis of researching wireless signal propagation model and traditional indoor location algorithm, in this paper, firstly we converted the RSSI value into signal dropout rate and calculated the dropout rate information respectively by using different transmit power. Then we predicted location of the mobile node by BP neural network. With this method, the location accuracy is improved.