Research On Access Point Location Method Based On CSI

With the rapid development of wireless LAN,the security problem of wireless network has become the major concern.Rouge APs,which perform man-in-the-middle attack by pretending as a legal AP,are prevailing today.Thus,localizing the rouge APs became an urgent issue.In this paper,we proposed a new Channel State Information(CSI)based rouge AP localization method,which achieve both high accuracy(average error in 2 meters)and high efficiency.The first issue we solve is how to identify the Line-of-Sight path.The state-of-the-art CSI-based location method is to use the fine-grained feature of CSI to filter out the secondary path information and then use the weighted average of the remaining information,which is called effective channel state information to locate.However,how could we effectively remove the part of multipath affect and interference,and get the Line-of-Sight path information is still not clear.In this paper,we use skewness and kurtosis to identify the line-of-sight path,and then carry out positioning using line-of-sight path information.The second issues we solve is how to evaluate the propagation parameters efficiently and effectively.Traditional localization method usually uses anchor nodes to get the parameters of the propagation model through offline training.However,the parameters of different data sets are different,thus it fails to locate access point based on range estimation.By changing the propagation model formula,we use some simple operation to eliminate parameters for constructing the corresponding matrix of measurement data,and finally apply the l2 norm to solute the problem.We then use pure mathematical methods to solve practical problems,which can reduce the number of the propagation model parameters.Finally,it is achieved a fast location estimation and improved the positioning accuracy.We implement the LOS-Based of parameter independent localization-LBPIL system on commercial 802.11 n NICs,and then evaluate its performance in two typical indoor scenarios.From the experimental results we can find that the accuracy can be enhanced and the latency of distance calculation can be significantly reduced by using our method.Moreover,LBPIL can significantly improve the localization accuracy compared with the counterpart approach.