Spectrum Prediction for Supporting IoT Applications over 5G

Spectrum-prediction-based dynamic spectrum access (DSA) has been considered as a promising approach to meet the higher spectral efficiency, more system throughput, and better network connectivity requirements of 5G Internet of Things (IoT) applications. This article discusses the feasibility of applying spectrum prediction in 5G followed by proposing a 5G architecture supporting the spectrum-prediction-based DSA for IoT users. The prediction techniques, which have critical impacts on the performance of spectrum-prediction-based DSA, are reviewed in this article. The information that can be obtained from prediction is summarized, and the typical impacts of spectrum prediction on DSA are discussed. We also identify and discuss a few open research issues and challenges resulted from applying spectrum prediction for DSA in 5G.