Research Article
A Novel Interference Matrix Generation Algorithm in GSM Networks
@INPROCEEDINGS{10.1109/CHINACOM.2009.5339733, author={Yu Gao and Jingqun Jingqun Song and Qixun Zhang and Yunfe Ma and Jinlong Wan and Zhiyong Feng}, title={A Novel Interference Matrix Generation Algorithm in GSM Networks}, proceedings={3rd International ICST Workshop on Cognitive Radio Network}, publisher={IEEE}, proceedings_a={CRNET}, year={2009}, month={11}, keywords={requency planning GSM interference matrix mobile measurement reports traffic grid map.}, doi={10.1109/CHINACOM.2009.5339733} }- Yu Gao
Jingqun Jingqun Song
Qixun Zhang
Yunfe Ma
Jinlong Wan
Zhiyong Feng
Year: 2009
A Novel Interference Matrix Generation Algorithm in GSM Networks
CRNET
IEEE
DOI: 10.1109/CHINACOM.2009.5339733
Abstract
Recently, the Automatic Frequency Planning which is mainly based on Interference Matrix (IM) is widely used in global system for mobile communications (GSM) networks planning. Most IM generation algorithms are derived from the Mobile Measurement Reports (MMRs) received from mobile users. Since the number of bits used for MMRs is limited and the MMRs could only provide the power level of six neighboring cells according to the GSM standards, the data from these MMRs is not accurate enough to generate IM for AFP. In this paper, we proposed a new IM generation algorithm based on Carrier to Interference Ratio (CIR) simulation results and the traffic grid map data. We design a new method by using the serving cell name, the longitude and the latitude as the index to match up the CIR data to the traffic data in the IM generation process. The proposed method can not only estimate the interference of existing GSM networks but also predict the interference of future GSM networks with traffic changing constantly. The proposed method can also take advantage of both the CIR simulation results with high resolution and the traffic grid map with prediction data to generate the IM with better accuracy than existing method using MMRs. Simulation results show that our proposed method with high accurate CIR data and traffic prediction data is much better than MMRs method.