Impact of the number of sensors on the network cost and accuracy of the radio environment map

Spectrum management using a radio environment map (REM) to dynamically allocate frequencies may facilitate the efficient exploitation of the spectrum in a heterogeneous environment. The overhead costs of networking radios to create the map must be balanced against the impact of a poor REM due to low quality or quantity of sensor data. To study this cost-benefit balance, simulations were performed to evaluate a REM using a variable number of randomly-located sensors. It was seen that too few sensors provides insufficient coverage of the area of interest and may not provide connectivity to the data collection centre, whereas in a more dense sensor environment, sensor location was the most important factor. The results show that information about the REM's accuracy is required to mitigate false confidence and that exact replication of the ground truth is not required for dynamic spectrum management; rather, it is sufficient to identify the regions where spectrum could be reused.