Two-hop interference channels: Impact of linear time-varying schemes

Ibrahim Issa, Silas L. Fong, Amir Salman Avestimehr

Introduction

In wireless communication, interference remains the primary bottleneck for network capacity. Multi-hop networks, which utilize relays to pass signals from sources to destinations, offer a promising architecture for managing this interference. While complex relaying strategies like "aligned interference neutralization" can theoretically double network capacity, they are often mathematically complex and difficult to implement in hardware. This research explores a simpler alternative: scalar Amplify-and-Forward (AF) relaying.

The study focuses on the "2-2-2" network topology, consisting of two sources, two relays, and two destinations. In a standard static environment with constant relay coefficients, this network collapses into a single-hop interference channel with a sum-Degree of Freedom (DoF) of 1. The authors investigate whether making these simple amplification coefficients time-varying can unlock higher performance, even when the underlying physical channel remains static.

The 2-2-2 two-hop interference channel model with sources (s), relays (u, v), and destinations (d).

Methods: Achieving 4/3 Sum-DoF

To demonstrate the potential of time-varying schemes, the authors propose a three-phase transmission strategy. This approach creates "artificial dynamics" by changing relay amplification factors across time slots to align or neutralize interference.

  1. Phase 1: Relay coefficients are chosen to cancel interference from the second source at the first destination.
  2. Phase 2: Coefficients are adjusted to cancel interference from the first source at the second destination.
  3. Phase 3: A different set of coefficients allows both destinations to "untangle" the mixed signals received in the previous phases.
The three-phase transmission strategy illustrating how symbols are combined and recovered over time.

This linear system allows each destination to recover two symbols over three time slots, achieving a DoF of 2/3 per user, or a total sum-DoF of 4/3. To prove this is the absolute limit, the authors developed a novel outer bound analysis, categorizing possible end-to-end channel matrix states and applying Fano’s inequality to show that no linear time-varying AF scheme can exceed 4/3.

Key Insights and Results

  • Artificial Dynamics: The most significant finding is that relays can create beneficial time-variation in static environments. By simply changing their multipliers, they transform a static bottleneck into a dynamic resource.
  • Performance Gain: The 4/3 sum-DoF represents a 33% increase in high-SNR capacity compared to traditional constant AF relaying.
  • The Power of Simplicity: The results prove that complex "Decode-and-Forward" strategies are not strictly necessary to beat standard interference limits; simple, time-coordinated multiplication is sufficient.

Limitations and Future Work

While this work provides a closed solution for the real-valued 2-2-2 channel, several avenues for future research remain. In complex-valued channels, the DoF could potentially reach 3/2 or 5/3. Additionally, extending these findings to $K$-user networks or MIMO (multiple antenna) systems—where a sum-DoF of $2M - 2/3$ is suggested for $M$ antennas—could significantly impact the design of 5G and 6G relaying protocols. Implementation will require addressing hardware constraints related to precise synchronization and fast coefficient switching.

Summary updated

BibTeX
@inproceedings{issa2013two,
  author = {Ibrahim Issa and Silas L. Fong and Amir Salman Avestimehr},
  title = {{Two-hop interference channels: Impact of linear time-varying schemes}},
  booktitle = {ISIT},
  year = {2013},
  doi = {10.1109/isit.2013.6620638},
}
Figure