On the Throughput Potential of Two-Dimensional Wireless Multi-Hop Networks Using Directional Antennas

The goal of this paper is to establish the rate region of a wireless multi-hop, one-sender, two-receiver transmission chain. We examine and evaluate the advantages of utilizing directional antennas in this context. In particular, we consider a two-dimensional communication network with a single source node S, two destination nodes D₁, D₂, and 2 times (N - 1) intermediate nodes placed equidistantly between them. Multi-hop transmission is an extension of single-hop transmission that can take advantage of the reduced attenuation between closely spaced relay nodes, as well as the opportunity of spatial reuse. In two-dimensional multi-hop half-duplex transmission, each node utilizes capacity-achieving point-to-point codes to forward the most recently decoded message to its nearest neighbor in the direction of D₁ or D₂. The transmission is performed over two parallel chains and it is a mixture of two techniques, "broadcast" communication (one transmitter to two relay nodes) and interference channel. First we describe our network and the assumptions made in the scope of this study. We further evaluate the end-to-end rate region of this topology under our transmission constraints and describe the effects that take place after replacing omni-directional antennas with perfectly directional ones. Our results indicate a 5-fold end-to-end sum rate improvement over a 10-hop network with the use of 30 degrees directional beams, as compared to omni-directional transmission.