Exploiting sparsity of relay-assisted cognitive radio networks

We propose a novel protocol for secondary users based on compressive sensing principles. The secondary user assigned to one of the primary frequency bands is aided by a set of nearby secondary users, which we refer to as secondary relays. We assume that each secondary relay participates in relaying the secondary packets with certain probability. In a given time slot, each secondary relay indicates its ability of decoding the secondary packet and being a relay for the secondary transmission. Our proposed protocol exploits the sparsity of the participating relays set and is efficient as it allows dynamic relay assignment. When the primary user is inactive, the relays send their states to the secondary source which, in turn, transmits its own packet if the received number of decoding relays satisfies a predefined threshold. When the primary user is active, the relays perform cooperative beamforming to achieve a cooperative diversity gain for the secondary source while completely eliminating the interference to the primary destination. We study the performance of the system from a cross-layer point of view.