Channel Selection for Secondary Users in Decentralized Network of Unknown Size

In this letter, the challenge of optimal channel selection among the non-cooperative and unknown number of secondary (i.e., unlicensed) users (SUs) in the decentralized network is addressed. The proposed scheme consists of two phases: estimation and ranking (ER) phase and high throughput (HT) phase. In the ER phase, each SU chooses the channel randomly. Based on the sensing and collision outcomes, the proposed mathematical expression allows the SUs to independently estimate the channel ranking and the number of active SUs, U. In the HT phase, the SUs follow collision-free hopping to exploit top U channels. An upper bound on the time required for the SU orthogonalization in the HT phase, loss in throughput, and number of channel switchings are derived. Theoretical analysis and simulation results validate the superiority of the proposed scheme over existing state-of-the-art schemes.