Abstract:In the infrastructure-less cognitive radio Ad-hoc networks(CRAHNs),neighbor discovery is the precondition of the execution of the protocols of MAC,topology management and routing,thus has great influence on the perfor-mance of the network.According to the environment of heterogeneous available channel sets and absence of common control channel in CRAHNs,a fast neighbor discovery algorithm based on the spectrum comparability(FNDSC) was proposed in this paper.Different from the existing synchronous algorithms that require users to switch on global channel set(GCS) in the network,a cognitive radio user following FDNSC switches on its respective available channel set(ACS) and transmits beacon messages with a probabilityλ.As GCS is generally big while the number of available channels for a specific user is probably small,the discovery process is considerably shortened.Simulation results show that the proposed algorithm reduces the discovery time by at least 47% compared with existing solutions.

Fast Neighbor Discovery in Cognitive Radio Ad-hoc Networks