Research On Indoor Localization Scheme And Mobility Management Strategy Based On Location Information

With the continuous development of wireless network technology and the increasing number of Internet of Thing devices connected to the network,how to provide high-quality service for users has become a major research in current wireless communication networks.Most services that mobile users need are based on location service.On the basis of the location information,personalized services can be provided to users.Therefore,how to locate users accurately in a wireless network and switch to a network that can provide optimal services is the key to improving user experience.Accurate positioning can be achieved in outdoor environments now,but the accuracy of indoor positioning methods needs to be improved urgently.In addition,it is also a major difficulty for users to effectively switch between different networks according to their own needs during the mobile process.This paper mainly studies high-precision indoor positioning algorithms and effective mobility management strategies.This paper obtains the user’s location information in the wireless network,and realizes the smooth and effective handover of the mobile user in the heterogeneous network under the premise of ensuring the user’s service quality.The specific work of the paper is as follows:(1)This paper introduces the existing indoor positioning technologies and algorithms and common mobility management technologies and solutions in detail.The related algorithm theory is given,which provides a foundation for the following research.(2)This paper proposes a fingerprinting localization scheme based on Received Signal Strength Indicator(RSSI).Based on the research on existing indoor positioning technologies and algorithms,this paper studies indoor positioning based on Wi Fi.Firstly,in order to resist the impact of multipath fading on RSSI,we build a RSSI-based fingerprinting localization model,and collect the RSSI of each access point(AP)in the actual positioning area to form an offline fingerprint database.Secondly,we propose an improved extreme learning machine(ELM)algorithm,which is ameliorated by Particle Swarm Optimization(PSO).Then we applied it to the online stages of the fingerprinting localization model.Positioning is achieved by matching the user and the RSSI in the fingerprint database.Finally,the simulation results indicate that the positioning scheme proposed in this paper reduces the positioning error effectively when it is contrast with traditional algorithm.(3)This paper proposes a multi-characteristics handover strategy based on Q-learning.In order to make users make optimal handover decisions in heterogeneous networks and alleviate the phenomenon of frequent handovers,this paper combines various user characteristic as the basis for handover decisions.Firstly,we construct a heterogeneous network in which macro cells and small cells coexist and calculate the corresponding RSSI,transmission rate and location of users in this network model.Secondly,the Q-learning algorithm is used to interact with the above characteristic and the network’s environment to make the best handover decision.Finally,form the experimental results we can know that contrasted with the scheme single characteristic,this scheme can effectively reduce the number of times of handover and obtain a higher transmission rate.