Two-way Networks: when Adaptation is Useless

TL;DR

This paper investigates whether adaptation strategies in two-way networks improve capacity, finding that in many deterministic and Gaussian models, adaptation does not enhance capacity and simple non-adaptive schemes are sufficient.

Contribution

The paper proves that adaptation is generally useless for capacity in various two-way network models, providing new outer bounds and demonstrating the effectiveness of non-adaptive schemes.

Findings

Adaptation does not increase capacity in deterministic two-way channels.

Non-adaptive schemes achieve capacity or near-capacity in many models.

Partial adaptation is ineffective in the Gaussian two-way interference channel.

Abstract

In two-way networks, nodes act as both sources and destinations of messages. This allows for "adaptation" at or "interaction" between the nodes - a node's channel inputs may be functions of its message(s) and previously received signals. How to best adapt is key to two-way communication, rendering it challenging. However, examples exist of point-to-point channels where adaptation is not beneficial from a capacity perspective. We ask whether analogous examples exist for multi-user two-way networks. We first consider deterministic two-way channel models: the binary modulo-2 addition channel and a generalization thereof, and the linear deterministic channel. For these deterministic models we obtain the capacity region for the two-way multiple access/broadcast channel, the two-way Z channel and the two-way interference channel (IC). In all cases we permit all nodes to adapt channel inputs to past outputs (except for portions of the linear deterministic two-way IC where we only permit 2 of the 4 nodes to fully adapt). However, we show that this adaptation is useless from a capacity region perspective and capacity is achieved by strategies where the channel inputs at each use do not adapt to previous inputs. Finally, we consider the Gaussian two-way IC, and show that partial adaptation is useless when the interference is very strong. In the strong and weak interference regimes, we show that the non-adaptive Han and Kobayashi scheme utilized in parallel in both directions achieves to within a constant gap for the symmetric rate of the fully (some regimes) or partially (remaining regimes) adaptive models. The central technical contribution is the derivation of new, computable outer bounds which allow for adaptation. Inner bounds follow from non-adaptive achievability schemes of the corresponding one-way channel models.