Myopic Coding in Multiterminal Networks

TL;DR

This paper explores the trade-offs in multi-terminal networks between cooperation levels and achievable data rates, demonstrating that limited cooperation strategies like myopic decode-forward can perform comparably to full cooperation in certain regimes.

Contribution

It introduces and analyzes myopic decode-forward strategies, showing their effectiveness and scalability in multi-terminal networks compared to omniscient coding.

Findings

Myopic decode-forward achieves rates comparable to omniscient coding at low SNR.

Adding nodes to cooperation significantly increases transmission rates.

Myopic strategies maintain non-zero rates as network size grows indefinitely.

Abstract

This paper investigates the interplay between cooperation and achievable rates in multi-terminal networks. Cooperation refers to the process of nodes working together to relay data toward the destination. There is an inherent tradeoff between achievable information transmission rates and the level of cooperation, which is determined by how many nodes are involved and how the nodes encode/decode the data. We illustrate this trade-off by studying information-theoretic decode-forward based coding strategies for data transmission in multi-terminal networks. Decode-forward strategies are usually discussed in the context of omniscient coding, in which all nodes in the network fully cooperate with each other, both in encoding and decoding. In this paper, we investigate myopic coding, in which each node cooperates with only a few neighboring nodes. We show that achievable rates of myopic decode-forward can be as large as that of omniscient decode-forward in the low SNR regime. We also show that when each node has only a few cooperating neighbors, adding one node into the cooperation increases the transmission rate significantly. Furthermore, we show that myopic decode-forward can achieve non-zero rates as the network size grows without bound.