Research On Indoor Location Algorithm Based On RSSI

For a long time,location technology has been a hot research topic.In recent years,with the development of science and technology,indoor spaces have gradually become larger.And the types of indoor objects have increased.People are more and more eager to determine the location of indoor people or objects,and the demand for indoor positioning has gradually become more prominent.Among them,RSSI(Received Signal Strength Indication)has always been a popular direction of indoor location technology.This paper first introduces the purpose and significance of indoor location,and explains the importance and necessity of studying indoor location.At the same time,this paper introduces the development status of indoor location based on RSSI value through the development timeline of domestic and foreign scholars on indoor positioning technology research.Then we introduced the main content and organization of this paper.To improve the traditional RSSI ranging algorithm,for the problem that the DV-hop positioning accuracy is affected by the node distribution,the DV-hop average hop distance of the current beacon node is estimated by improving the hop weighting coefficient.For the RSSI ranging accuracy sensitive to the environmental factors,the path loss of the average hop distance of the current beacon node is used as the reference,and the distance between two nodes is calculated according to the path loss between the nodes.Finally,these beacon nodes that satisfied the hop count threshold condition are selected to estimate the unknown nodes coordinates.We call the improved method ADLDV-hop(Average hop Distance's path Loss DV-hop)algorithm.The results show that: under the same beacon nodes,communication radius and SNR,the location accuracy of this algorithm is better than BRDV-hop(Basic RSSI DV-hop)and other improved algorithms.With the development of computer technology,fingerprint location has developed rapidly.To solve the problems that BP neural network is easy to fall into local minimum value,low positioning accuracy and local oscillation in fingerprint location.By introducing a Convolutional Neural Networks(CNN)based on the Tensor Flow framework to extract the one-dimensional convolution features of the input fingerprint signal,and then combining the Inception structure to expand the depth of the neural network.The results show that under the same data set and the same number of cycles,the CNN model has higher location accuracy.By optimizing indoor location algorithms based on RSSI values in two different situations,this article demonstrates the importance of RSSI values for the development of indoor location technology.