Analysis And Research Of Train-ground Communication Handoff Performance For CBTC System Based On TD-LTE

Along with the rapid development of the city and the increase of urban population year by year,urban rail transit is vigorously developed at home or abroad to improve the tension of urban traffic.CBTC is the common control system of urban rail transit for the moment.The wireless communication system of CBTC mainly adopts WLAN.However,in the wake of the WLAN equipment scale enlargement which leads to interference,the operation safety of CBTC system is more challenging.TD-LTE has advantages of high speed transmission,low latency,large coverage and backward compatibility.Therefore,it is the trend of the industry to replace WLAN with TD-LTE in urban rail transit.This paper treats the wireless communication system of CBTC,including system wireless channel model,system architecture and the wireless transmission error of system.In order to improve the quality of CBTC train and ground communication,a handoff algorithm and a optimization algorithm are proposed,which reduces the packet loss rate caused by ping-pong handoff and treats handoff interruption due to long time-delay.Firstly,the paper formulates the basic concepts of TD-LTE and CBTC,then studies the wireless communication system of CBTC based on TD-LTE including system wireless channel model and system architecture.Then,handoff mechanism of TD-LTE is expounded.The wireless transmission error of system is analyzed from packet loss rate and time delay.It is analyzed that the packet loss rate in error retransmission process is mainly affected by ping-pong handoff.According to the analysis,the total system delay depends on HARQ retransmission rate and the radio subframe configuration mode.After,this paper proposes a handoff algorithm based on improved GM(1,1)grey prediction model to solve the problem of ping-pong handoff.The handoff algorithm is based on the handoff A3 event decision formula of same frequency cell in TD-LTE.An improved GM(1,1)grey prediction model is used to predict RSRP of Source eNodeB and target eNodeB.The handoff algorithm reduces the impact of inherent errors in data series predictions by optimizing estimation parameters,in order to depress RSRP fluctuation which can prevent the frequent occurrence of ping-pong handoff.The results suggest improvements to RSRP fluctuation.The results suggest that the handoff algorithm can improve RSRP fluctuation,reduce the loss rate of the system and depress ping-pong handoff.Finally,this paper presents a target pre-bearing handoff optimization algorithm based on improved GM(1,1)model.Target pre-bearing method is used to reduce the handoff delay in the handoff optimization algorithm which is on account of a handoff algorithm based on improved GM(1,1)grey prediction model.The results indicate that the handoff optimization algorithm can effectively reduce handoff delay of the system,depress the probability of system interruption and improve the handoff success rate.Overall,the results reveal that the handoff performance of the system is greatly improved in the the handoff optimization algorithm compared with the traditional algorithm.