Harvesting MDT data: Radio environment maps for coverage analysis in cellular networks

Abstract

Despite the remarkable progress in radio access technology to support the rapidly increasing wireless data demand, coverage analysis remains as one of the indispensable topics on which mobile operators still need innovations, above all, in terms of operational efficiency together with performance. Manual coverage detection and prediction is an inefficient and costly task. In this paper we show how Radio Environment Maps (REMs) developed as part of the research on cognitive wireless networks can be used as a basis for a powerful coverage estimation and prediction solution for present-day cellular networks. Applying powerful spatial interpolation techniques on the information coming from location-aware devices, REMs provide a realistic and remote representation of the ground-truth. The proposed approach automatically identifies the number, location and shape of the existing coverage holes and therefore constitutes a perfect example of a novel application of the Cognitive Radio concept on next generation cellular networks. Results on urban and rural environments show that the use of REM brings promising gains in coverage hole detection and prediction with respect to the case where only measurements are used.