Constructing accurate Radio Environment Maps with Kriging Interpolation in Cognitive Radio Networks

The increased development of information industry and relevant radio communication services is making the spectrum management problem more challenging to solve. The Radio Environment Map (REM) can be a powerful tool for solving spectrum scarcity and spectrum access problems, making context-aware resource allocation more efficient. We can optimize the accuracy of the REM using a geostatistical tool named Kriging interpolation in cognitive radio networks. In this paper, the proposed REM construction method combines residual maximum likelihood-based radio propagation parameter estimation with Kriging-based transmission power prediction. Additionally, we compare the performance of the constructed REMs with the metric of root mean square error (RMSE) in three methods: the path loss-based method, the Kriging-based method without the fit of a path loss model beforehand, and the proposed method. With the Monte Carlo simulation, the result indicates that the path loss-based method provides the most unsatisfied performance and the interpolation accuracy can be improved by 2 dB with the proposed method, which proves that our approach can effectively determine sharing conditions of radio spectrum use both in time and space.