[Paper Review] Degrees of Freedom of Two-Hop Wireless Networks: "Everyone Gets the Entire Cake"

Ilan Shomorony, A. Salman Avestimehr|arXiv (Cornell University)|Oct 8, 2012

Cooperative Communication and Network Coding23 references8 citations

TL;DR

This paper introduces Aligned Network Diagonalization (AND), a novel interference management and relaying scheme for two-hop K×K×K wireless networks that achieves K degrees of freedom for both time-varying and constant channel coefficients. By jointly optimizing source, relay, and destination precoding to induce a diagonal end-to-end channel, AND eliminates interference and enables each user to achieve its full spatial multiplexing gain, effectively allowing everyone to receive the entire data cake without interference.

ABSTRACT

We show that fully connected two-hop wireless networks with K sources, K relays and K destinations have K degrees of freedom both in the case of time-varying channel coefficients and in the case of constant channel coefficients (in which case the result holds for almost all values of constant channel coefficients). Our main contribution is a new achievability scheme which we call Aligned Network Diagonalization. This scheme allows the data streams transmitted by the sources to undergo a diagonal linear transformation from the sources to the destinations, thus being received free of interference by their intended destination. In addition, we extend our scheme to multi-hop networks with fully connected hops, and multi-hop networks with MIMO nodes, for which the degrees of freedom are also fully characterized.

Motivation & Objective

• To close the gap between the achievable degrees of freedom and the cut-set outer bound in multi-hop multi-flow wireless networks.
• To demonstrate that coupled strategies for relaying and interference management outperform decoupled approaches like interference alignment or X-channel schemes.
• To characterize the degrees of freedom for fully connected two-hop networks with K sources, K relays, and K destinations under both time-varying and constant channel coefficients.
• To extend the proposed scheme to multi-hop networks with fully connected hops and MIMO-capable nodes.

Proposed method

• Proposes Aligned Network Diagonalization (AND), a linear precoding and decoding scheme that transforms the end-to-end channel into a diagonal form.
• Uses polynomial-based precoding vectors at sources and relays to align interference at relays and destinations in a structured way.
• Employs a matrix inversion technique to compute relay coefficients such that the effective channel becomes diagonal, ensuring interference-free reception.
• Applies the concept of analytic functions and linear independence of monomials to prove that the desired interference alignment is achievable for almost all channel realizations.
• Extends the scheme to multi-hop networks by applying AND recursively across hops, achieving full degrees of freedom.
• Uses determinant-based arguments and polynomial non-vanishing proofs to establish the existence of valid precoding vectors.

Experimental results

Research questions

• RQ1 Can a coupled strategy for relaying and interference management achieve the full cut-set bound of K degrees of freedom in a K×K×K two-hop network?
• RQ2 Is it possible to achieve K degrees of freedom in the presence of constant (non-time-varying) channel coefficients?
• RQ3 How does the performance of AND compare to decoupled schemes like interference alignment or X-channel strategies in terms of degrees of freedom?
• RQ4 Can the AND framework be extended to multi-hop networks with fully connected hops and MIMO-capable nodes?

Key findings

• The K×K×K two-hop wireless network achieves K degrees of freedom under time-varying channel coefficients.
• For constant channel coefficients, K degrees of freedom are achieved for almost all channel gain values, proving robustness to static fading.
• The proposed Aligned Network Diagonalization (AND) scheme achieves full degrees of freedom by inducing a diagonal end-to-end channel through joint precoding at sources, relays, and destinations.
• The AND scheme significantly outperforms decoupled strategies such as interference alignment (achieving only K/2 degrees of freedom) or X-channel schemes (achieving K²/(2K−1) degrees of freedom).
• The scheme generalizes to multi-hop networks with fully connected hops, achieving full degrees of freedom for such topologies.
• The proof relies on showing that a certain determinant of a matrix formed from monomials is non-identically zero, ensuring the existence of valid precoding vectors for interference alignment.