With the rapid expansion of the mobile GSM network scale and urban construction, the cell coverage environment is complex and protean, which influences the station signal,and causes the work load and the difficulty of network optimization more difficult. This paper designs and implements a multiunit data collection method to acquire the realtime performance data of the GSM network. In order to improve the efficiency of data processing, it uses a fast incremental update algorithm to process the performance data of the GSM network. By using the Gaussian Distribution and Mahalanobis Distance to analyse the parameter variation of the TA, the upward and downward path loss and the output power, we count abnormal data and the changes of the district coverage environment, then give network optimization advice according to the analysis results. This method implements the cell coverage environment system based on TA which can help network optimization more efficient and improve the efficiency of the GSM network optimization.