A Novel Spectrum Handoff Technique for Long Range Applications using Adaptive Beam Selection with Machine Learning Algorithms

A cognitive radio network is an intelligent system that can detect the presence of unused spectrum space without affecting the primary user. The CRN is responsible for managing the allocation of channels, which leads to the scarcity of spectrum. This issue should be addressed in order to ensure that the network can continue to provide long-term and profitable service to its users. The issue of spectrum handoff is considered a major issue that needs to be resolved in order to improve the efficiency of the network. One of the most common factors that can affect the network's performance is the power consumption and communication delay. This paper proposes a method that can detect the availability of channels during the handover. The accuracy and network latency of various ML algorithms are evaluated through resampling techniques. The Nave Bayes Classifier and KNN algorithms performed better than their benchmarks. For a total of 500 and 100 users, respectively, the networks experienced a network latency of 10.91 and 13.08 seconds.