Implemention And Performance Analysis Of Indoor RSSI Localization Algorithm Based On Machine Learning Theory

With the help of booming artificial intelligence, machine learning, as an interdisciplinary science, involving probability, statistics,approximation theory, convex analysis, complexity theory and other disciplines, is widely used to provide intelligent solution to solve various realistic problems. LBS, as a realistic positioning service, develop rapidly in this century. Outdoor positioning based on GPS technology has been extremely mature. Indoor positioning based on RSSI matching algorithm has great room for improvement in accuracy, real-time and positioning efficiency because indoor environment is complex and difficult to model.In this paper, we combine RSSI indoor positioning technology with machine learning theory for indoor positioning is difficult to establish a mathematical model and match positioning has high computational complexity. We choose neural network algorithm to position and uses related algorithm to optimize. In addition to algorithm design, we implement BPNN and optimization algorithm via programming language.Train and test the BPNN using measured data. At last, propose a RSSI positioning scheme based on machine learning theory.