Evaluating Opportunistic Multi-Channel MAC: Is Diversity Gain Worth the Pain?

Abstract

We evaluate the performance of an opportunistic multi-channel medium access control protocol and compare it to that of the corresponding single-channel MAC (S-MAC) and a non-opportunistic multi-channel MAC (M-MAC). We do this in three different settings: (1) an ideal scenario where no control channel is used and no sensing delay is incurred, (2) a more realistic scheme where users compete for access on a control channel using random access, and (3) a scheme similar to (2) but with a time-division multiplexing (TDM) based access scheme on the control channel. Our analysis and numerical results show that in terms of delay performance, the random access and competition on the control channel, which typically occupy a fraction of the total bandwidth, almost always wipe out the channel diversity gain, a main motivation behind an opportunistic multi-channel MAC. On the other hand opportunistic access increases bandwidth utilization which reduces the system's total busy time. As a result it helps reduce power consumption in general. When TDM is employed on the control channel, the data sub-channel sensing delay becomes the main bottleneck to attaining better performance. In this case the performance of opportunistic multi-channel MAC gets closer to that of the single-channel MAC when the channel sensing overhead is substantially reduced.