Data Fusion Based Interference Matrix Generation for Cellular System Frequency Planning

TL;DR

This paper proposes two data fusion algorithms to improve interference matrix accuracy in cellular frequency planning by combining mobile measurement reports and drive test data.

Contribution

It introduces novel data fusion methods that enhance interference matrix generation accuracy using combined source data in cellular systems.

Findings

Fused data provides more complete interference information.

Enhanced IM accuracy validated through simulations.

Improved frequency planning outcomes.

Abstract

Interference matrix (IM) has been widely used in frequency planning/optimization of cellular systems because it describes the interaction between any two cells. IM is generated from the source data gathered from the cellular system, either mobile measurement reports (MMRs) or drive test (DT) records. IM accuracy is not satisfactory since neither MMRs nor DT records contain complete information on interference and traffic distribution. In this paper, two IM generation algorithms based on source data fusion are proposed. Data fusion in one algorithm is to reinforce MMRs data, using the frequency-domain information of DT data from the same region. Data fusion in another algorithm is to reshape DT data, using the traffic distribution information extracted from MMRs from the same region. The fused data contains more complete information so that more accurate IM can be obtained. Simulation results have validated this conclusion.