Electromagnetic Situational Awareness of Cognitive Radios supported by Radio Environment Maps

Spectrum Map Construction Based on Optimized Sensor Selection and Adaptive Kriging Model

“…As shown in Fig. 9, in order to further demonstrate the broad effectiveness of the algorithm, this paper compares different SM construction algorithms based on measured data, including ABC-SA, IDW [31], NN [32], and Splines. It can be seen that the RMSE of each algorithm decreases with the increase of the number of sensors, which indicates that increasing the sample sampling rate is an important factor to improve the accuracy of the SM.…”

Gaussian mixture model based radio environment mapping for an underlay Device‐to‐Device network

“…Currently, a significant variety of contemporary and upcoming use-cases are being catered to by REMs for example, 5G and IoT network optimization, 12,13 wireless resource allocation, 14 cooperative communication, 15 transmitter localization, 16 UAV based access point placement, 17 network diagnostics, 18 Multi-hop routing, 19 and military applications. 20 In order to effectively empower all of the above mentioned and more applications, it is imperative that the REM creation process is quick. One quite simple way to achieve that through better deployment of REM sensors.…”

“…Mostly, it has been observed that the regular deployment is generally better than the random deployment of sensors. 20 As with any other physical process monitoring exercise, adding more than the adequate number of sensors is an exercise with diminishing returns in terms of REM accuracy. 27 More than adequate number of sensors would consequently lead to a higher overhead in the process of REM creation, which in turn would adversely affect the network throughput.…”

“…In Reference 26 random and square deployment has been compared with square providing better results. Mostly, it has been observed that the regular deployment is generally better than the random deployment of sensors 20 …”