Research On Data Collection Method Of Crowd-sensing Network Based On Edge Computing

With the rapid increase of the use of mobile devices with built-in sensors,mobile crowd sensing(MCS)has emerged as a sensing mode driven by humans.It collects,senses,and analyzes city data through millions of independent mobile sensing devices.However,due to the uncontrollability of user movement patterns,data redundancy problems are often caused,and the pressure on cloud servers increases when a large number of users submit data to the cloud server in parallel.In order to solve this problem,this paper introduces edge computing into mobile crowd sensing to process user movement information and collect perception data.The main work is summarized as follows:(1)A user selection strategy based on the prediction algorithm of user movement patterns is proposed.Because the edge server in the sensing scene can detect the user's movement pattern in real time,this strategy first uses the edge server to analyze and predict the user's movement pattern.This process includes the preprocessing of the user's movement data and the prediction algorithm of the user's stay point(This paper is referred to as DB?RF)and the LSTM-based movement pattern prediction model,and then according to the pattern prediction algorithm,users in this area can be selected for data collection for a period of time in the future,which solves the problem of data redundancy due to the uncontrollability of user movement patterns.This experiment uses the Geo Life data set to verify the user's movement pattern prediction algorithm.The experiment shows that it can predict the user's movement pattern with high accuracy and help the platform effectively select users.(2)A data collection algorithm based on edge computing is proposed.This chapter uses the edge server to process the data collected by users,at the edge node,the data is sampled by the compressed sensing algorithm,and the compressed data is transmitted to the cloud server.The CL?BP algorithm is used in the cloud server to recover the compressed data,so as to effectively restore the data in the target area.This experiment uses air quality data set to verify the effectiveness of the proposed method.By comparing the CL?BP algorithm with the OMP algorithm,it shows that the CL?BP algorithm can recover the original data with a lower error under a smaller compression ratio(ie,the amount of compressed data),effectively reduce the pressure on cloud servers.