Predictive spectrum decision mechanisms in Cognitive Radio Networks

Abstract

Spectrum decision is an important functionality of Cognitive Radio Networks that directly effects the overall network performance of secondary users. Usually spectrum decision algorithms work on a decision cycle and produces results for the transmission cycle that comes afterwards. This situation brings out a latency and can be avoided by the spectrum prediction schemes. Considering this challenge, in this work, we propose three spectrum prediction mechanisms in order to predict the future channel usages on spectrum with the help of the history window consisting of previous spectrum decision results. More specifically, the proposed methods are based on the correlation and linear regression analysis of the previous decisions, to further forecast the future spectrum status. Since these prediction mechanisms solely depends on individual sensing histories of secondary users, they are suitable for implementation in cognitive radio ad hoc networks. We evaluate the proposed methods by re-defining the System Utility parameter and by newly deriving a Primary User Disturbance Ratio. The obtained results are compared to the generic decision fusion strategies like And, Or and Majority. The simulations results shows that the proposed Correlation Based Spectrum Prediction Scheme has better performance on varying simulation environments.