Spectrum surveying for policy based cognitive radio

Abstract

Rapid growth in wireless communication technologies and services creates high demands for radio spectrum. Current spectrum management and regulations are unable to cater these demands despite claims that radio spectrum is underutilised. Cognitive Radio (CR) aims to resolve these conflicts by enabling opportunistic spectrum access to licensed spectrum. Understanding the spectrum occupancy patterns of licensed spectrum users is important for large scale deployment of CR. This thesis aims to evaluate the 24-hour spectrum occupancy in the 30-3000 MHz frequency band of a suburban outdoor radio environment in Johor Bahru, Malaysia and its applications in the policy development for Policy Based Cognitive Radio (PBCR) concept. A spectrum survey was carried out whereby spectrum measurement was performed, and spectrum data was classified and analysed using energy detection and duty cycle based method. Findings revealed that the spectrum occupancy was 11.29% and several potential spectrum bands for future deployment of CR were identified. Distributions of the duty cycle variable were modelled using continuous probability distributions and their entropy was investigated for different occupancy levels. Results indicated that the beta and Kumaraswamy distributions were an accurate fit and investigations on the entropy distributions proposed characterisations for high, moderate and low spectrum occupancy level. The information gained was applied to demonstrate the development and testing of PBCR policies using Cognitive Radio Language. The demonstration proved the practicability of using the information from spectrum survey in the policy development of PBCR.