Decentralized Cooperative Spectrum Sensing for Ad-Hoc Disaster Relief Network Clusters

Disaster relief networks need to be highly adaptable and resilient in order to encompass the emergency service demands. Cognitive Radio enhanced ad-hoc architecture have been put forward as candidate to enable such networks. Spectrum sensing, the cornerstone of the Cognitive Radio paradigm, has been the target of intensive research, of which the main common conclusion was that the achievable spectrum sensing accuracy can be greatly enhanced through the use of cooperative sensing schemes. When considering applying Cognitive Radio to ad-hoc disaster relief networks, the use of spectrum sensing cooperative schemes becomes essential. A cluster based decentralized orchestration cooperative sensing scheme is proposed, where each node in the cluster decides which spectrum it should monitor, according to the past sensing decisions of all the cluster nodes. The proposed scheme performance is evaluated through a framework, which allows gauging the accuracy of multi narrow-band spectrum sensing cooperative schemes as well as to gauge the error in the estimation of each of the channels un-occupancy. Through that evaluation it is shown that the proposed decentralized scheme performance reaches the performance of the correspondent centralized scheme while outperforming the Round Robin scheme.