Research On Signal Preference Mechanism And Dynamic Parameter Fingerprinting Technique Based On Channel State Information

With the increasing demand for location-based services,the localization technology based on wireless network signals in complex environments has developed rapidly in recent years.Channel State Information(CSI)can be used as location feature information in fingerprint-based localization systems because it can reflect the characteristics of the signal over multiple subcarriers.However,the random noise contained in the original CSI information increases the possibility of confusion when matching fingerprint data.In order to effectively reduce the noise information generated in the localization process and improve the feature resolution of fingerprint matching,this paper focuses on the CSI-based fingerprint localization technique and the signal preference mechanism to improve the stability of the localization system by proposing a new localization means,and then optimize the system's localization effect in indoor environments.The main research contents and results of this paper include the following three aspects.(1)Processing of CSI amplitude and phase information and feature fusion.To address the problems of noise interference and high information dimensionality of CSI measurements in complex environments,this paper first preprocesses the magnitude and phase information by a density-based clustering algorithm,and then proposes dynamic fusion features(DFF)as a new fingerprint formation method by considering the location specificity of the magnitude and phase information from the fused magnitude and phase information.By establishing the dynamic weight values with environment-adaptive fusion of amplitude and phase information,the location characteristics of the fingerprint are improved while reducing the data complexity.(2)A study of similarity measures for high environmental adaptability.The online CSI data collected in the online matching phase of the fingerprint localization method need to be matched with the location fingerprint data in the fingerprint database using similarity measures.In this paper,Edit Distance on Real sequence(EDR)is used to measure the similarity between fingerprint data.In order to reduce the influence of noise and improve the environmental adaptability of the similarity matching method,this paper further sets the matching threshold needed in the process of calculating EDR as a dynamic parameter that changes with the data features,and uses the improved Edit Distance on Real Sequence(IEDR)as a similarity measure.The improved Edit Distance on Real Sequence(IEDR)is used as a similarity measure.The similarity matching method is combined with the DFF fingerprint generation method in this paper to form a complete CSI fingerprint localization system DFF-EDR,which further improves the feature resolution and localization accuracy of fingerprint data in terms of both fingerprint data formation and similarity metrics.In the experimental stage,this paper develops the evaluation and comparative analysis of the localization performance in two typical indoor scenarios from three perspectives of fingerprint formation method,similarity measurement index and system as a whole for the proposed localization system.The experimental results show that the proposed DFF-EDR localization system achieves improved localization performance in terms of fingerprint formation method and similarity measurement index.(3)Signal preference mechanism under the condition of multiple wireless access nodes.Ultra-dense heterogeneous network technology is one of the main application technologies for future 5G mobile communications.In this paper,under the premise of proposing a complete localization system,the signal preference mechanism under the condition of multiple indoor wireless access nodes is developed to improve the transmission quality of CSI signals in the localization system,and to provide a theoretical reference for future localization technologies based on ultra-dense heterogeneous networks,taking into full consideration the characteristics of the current localization system.By arranging multiple wireless access nodes in a more complex indoor scenario,the performance of the localization system under a single node and after using the signal preference mechanism is compared.The results show that the signal preference mechanism under multiple nodes has further improved the performance of the localization system.