Optimally Probing Channel in Opportunistic Spectrum Access
Kehao Wang, Quan Liu, Qingju Fan, Qingsong Ai · IEEE Communications Letters · 2017
This letter studies an opportunistic communication system in which a transmitter probes k out of N two-state Markov channels and accesses one channel to communicate with a receiver each time. The objective is to maximize the transmitter's throughput over a finite time horizon. A joint probing and accessing policy is proposed in which the transmitter probes k channels from the second best and accesses the best if all k channels are found to be bad. Further, the policy is theoretically shown to be optimal for the case of k + 2 = N and conjectured to be optimal for k +2 <; N under certain conditions. The results show that the optimality of the proposed policy has a tight coupling with the initially available probabilities of those channels.