Positioning Of The GSM-R Source Of Interference And Cloning Base Station Based On TRAIN-LAND Spectrum Map

With the rapid development of railway construction, GSM-R network optimization is paid more attention. Artificial interference turns out to be one of the most important factors that affect the quality and security of GSM-R network. In GSM-R system, criminals can attack the train-ground communication system through strong interference methods such as blanket jamming and the cloning base station, thus threatening the safe operation of trains. The train-ground spectrum map can receive strength information and as well collect and store the radio frequency fingerprints into the database through periodic inspection of testing trains and the daily operation of the trains. This thesis focuses on the analysis of radio frequency fingerprint information to locate the interference source and clone base station.The traditional wireless jammer locating scheme uses spectrum analyzer to measure the received signal. During the positioning process, to determine the existence and location of the jammer, the spectrum analyzer position and antenna direction need to be adjusted and the spectral components of signals need to be analyzed continuously. The new jammer location scheme is proposed here to overcome the problems of time consumption and human resource waste of traditional one. Several illegal man-made interference types are introduced into GSM-R system by studying the classification of GSM-R system. Firstly, based on the Okumura-Hata and Winner II channel models, the simulation scenarios and radio frequency fingerprint database are built up. To improve the efficiency of the jammer locationing performance, the thesis puts forward a real time locating method based on the database and reception level data. Three-point location, centroiding algorithm and two intersection positioning algorithm are applied here. The simulation is implemented through Visual C++2005and IT++. The results show that the three-point location algorithm is optimal with least error. As a result, the three-point location algorithm is used as the basic algorithm and optimized the jammer location algorithm through improving the multipoint least mean square error. Both the time and cost consumption are decreased, which is of a practical and efficient significance.Simulation results show that the design of radio frequency fingerprint database can satisfact the system requirements and the location error is reduced by the optimal three-point location algorithm.