Research And Implementation On Tracking System Of Pseudo Base Stations Based On Crowdsensing

The GSM pseudo base station utilizes the defect of GSM's one-way authentication to the mobile phone user,can be used to forge the source and content of the short message and becomes a tool for criminals to commit a crime.In order to expand the coverage of spam messages and evade the policeman's investigation,criminals often carry pseudo base stations to move around.Currently,most research is based on analyzing the pseudo base station signal to track the pseudo base station.Not only it requires professional equipment,but also the multipath effect in the urban environment results in huge positioning errors and delays.And these tracking methods cannot track the mobile pseudo base station in real time.To solve the above problems,a system to track the mobile pseudo-base station based on crowdsensing is proposed,it uses the sensors of the cell phones to collect the data of the pseudo base station in multiple points,and send the data to the system's server to analysis.The paper's main innovation is to consider that criminals usually use vehicles as an assistant tool to carry the pseudo base station to send spam message,so the pseudo base station itself is on the road.Therefore,matching the estimated trajectory of the pseudo base station with the map roads can overcome the positioning error caused by the signal analysis,then improve the positioning accuracy of the mobile pseudo base station.The main work of this article is as follows:1.A pseudo-base station tracking scheme based on crowdsensing is proposed to solve the low accuracy and poor real-time performance problems of the current tracking systems of the pseudo base.Firstly,based on the crowdsensing idea,the system includes the smart mobile terminals to monitor and collect the received signal strength indicator(RSSI)from the pseudo base station,and then send it to server along with the terminal's IMSI,location and system time,to estimate the pseudo base station position;then the estimated position sequence is matched with the roads on the map to get the trajectory that is closest to the real moving path of the pseudo base station.2.The essence of the crowdsensing based positioning method of pseudo-base stations is to using multiple anchor nodes with known locations to find out the location of the blind node with an unknown location.The positioning methods can be generally classified into two types depending on whether the method uses the electromagnetic wave equation to calculate the length between anchor and blind node or not: the former relies on the physical parameters of the wireless signal to estimate the distance,while the instability of the wireless signal will bring serious errors;the latter algorithm's accuracy depends on the selection of anchor nodes.Based on the advantages of these two methods,a positioning algorithm that is based on centroid estimation and Gaussian distribution likelihood estimation are proposed.The method is divided into two steps: 1)choose the anchor nodes with high RSSI,then use the centroid algorithm to get the location of the centroid of these chosen anchors as the estimated location;2)selecting m nodes around an anchor node to obtain the average Euclidean distance and variance to reduce the error caused by RSSI instability.Then grid the map,use the distance and variance to calculate the Gaussian distribution probability of each grid on the map.Apply the maximum likelihood estimation to take the grid with the biggest probability as the final position of the pseudo base station.According to the simulation experiment,the average error of the algorithm is only 6.51 meters and the delay is 1.61 seconds.3.In the real environment,due to the presence of measurement error and positioning error,the position sequence of the pseudo base station obtained by the above positioning algorithm still has a large deviation from its true trajectory.Considering that the pseudo base station trajectory should match the roads on the map,a new path matching algorithm based on Hidden Markov Chain to generate pseudo base station trajectory is proposed.The Hidden Markov Chain based algorithm uses spatial topological information and non-spatial attributes such as road speed limitation that is all extracted from the map to the calculation of the observation probability and transition probability,to match the position sequence with the roads.The simulation shows that even if the error ups to the limit of the positioning algorithm(the error between the estimated location and real location of the pseudo base station is 16.51 meters),the matching accuracy can still reach 90.71%,which indicates the algorithm has good robustness.4.In order to verify the effectiveness of the proposed scheme,a prototype verification system is implemented for the crowdsensing based tracking system that targets on mobile pseudo-base stations.The system consists of a cell phone application and a server that do data analysis.The cell phone application is deployed on the cell phone terminal,it can monitor and collect data on the pseudo base station that is near to the cell phone.The server computes the estimated location of the pseudo base station,generates and displays the trajectory.The simulation shows that the system has small system overhead,high accuracy in calculating the pseudo-base station trajectory,qualified real-time performance,and strong practical value.