Quantifying performance of duplexing, multiplexing and multiple access in mesh, relay, and ad-hoc networks

Abstract

Decentralized wireless ad hoc networks have become attractive in several settings. The choice of duplexing, multiplexing, and multiple access (D/M/MA) techniques in ad hoc networks is critical to important network characteristics. Yet determining the “optimum” D/M/MA technique is still an open problem, and to our knowledge very few works directly compare available D/M/MA techniques. In this paper we extend previous comparisons of mesh and relay networks to include a simple ad hoc network. In addition, in contrast to our previous work that uses an equal peak power constraint, here we apply two different constraints: an equal average power, and an equal energy constraint. Our comparison method is illustrated with the mesh network, and results for relay networks are summarized. We then investigate D/M/MA performance for an example ad hoc network. We show that one's conclusions regarding the different D/M/MA schemes depend upon the comparison constraint, but for all three constraints, results illustrate that specific types of hybrid time-frequency resource allocation schemes are equivalent to or better than pure time division or pure frequency division in terms of data rate, signal-to-noise ratio, and throughput, in all network topologies. Moreover, the hybrid time-frequency schemes offer more flexibility than the pure time and frequency division schemes in terms of bit-error-ratio and range.