Two-User SIMO Interference Channel with TIN: Improper Signaling versus Time-Sharing

Christoph Hellings, Ferhad Askerbeyli, Wolfgang Utschick

Abstract

In the two-user Gaussian interference channel with Gaussian inputs and treating interference as noise (TIN), improper complex signals can be beneficial if time-sharing is not allowed or if only the data rates are averaged over several transmit strategies (convex hull formulation). On the other hand, proper (circularly symmetric) signals have recently been shown to be optimal if coded time-sharing is considered, i.e., if both the data rates and the transmit powers are averaged. In this paper, we show that both conclusions remain the same if single-input multiple-output (SIMO) systems with multiple antennas at the receivers are considered. The proof for the case with coded time-sharing is via a novel enhanced channel concept for the two-user SIMO interference channel, which turns out to deliver a tight outer bound to the TIN rate region with coded time-sharing. The result for the case without coded time-sharing is demonstrated by studying specific examples in which a newly proposed composite real gradient-projection method for improper signaling can outperform the globally optimal proper signaling strategy. In addition, we discuss how the achievable TIN rate region with coded time-sharing can be computed numerically.

Introduction

In modern wireless communications, managing interference is a primary hurdle to increasing data rates. A common and practical approach is Treating Interference as Noise (TIN), where receivers do not attempt to decode interfering signals but instead treat them as additive noise. Traditionally, communication systems rely on proper signaling, where signals are circularly symmetric in the complex plane. However, recent theoretical advances have suggested that improper signaling—where the signal's real and imaginary parts have different variances or are correlated—can "shape" interference, potentially offering better performance in certain scenarios.

This paper investigates the trade-off between improper signaling and coded time-sharing in a two-user Single-Input Multiple-Output (SIMO) interference channel. The central question is whether the spatial degrees of freedom provided by multiple antennas at the receiver render improper signaling redundant or if they offer new avenues for optimization.

The Role of Coded Time-Sharing

A critical distinction in this research is the type of time-sharing employed.

  1. Standard Time-Sharing (Convex Hull): Only the data rates are averaged over different transmit strategies.
  2. Coded Time-Sharing: Both the data rates and the transmit powers are averaged across different time slots.

The authors demonstrate that when coded time-sharing is permitted, proper signaling is perfectly optimal. In this context, coded time-sharing acts as a "proper-signal-maker." Any rate point achievable via complex improper signaling can be matched or exceeded by a combination of proper signals and power-averaging over time.

Methodology and Mathematical Proofs

To establish these findings, the paper employs several sophisticated mathematical and algorithmic tools:

1. Channel Enhancement Concept

To prove the optimality of proper signaling under coded time-sharing, the authors introduce a novel "enhanced channel" concept. By transforming the complex SIMO channel into a real-valued representation and identifying an "enhanced" version of the channel where phases are aligned, they establish a tight outer bound for the TIN rate region. This proof confirms that the spatial dimensions of SIMO do not change the fundamental optimality of proper signaling when temporal averaging is available.

2. Optimization Algorithms

  • Mixed Monotonic Programming (MMP): To determine the globally optimal rate region for proper signaling with coded time-sharing, the authors use a branch-and-bound algorithm based on MMP. This addresses the non-convex nature of the rate equations.
  • Composite Real Gradient-Projection: For scenarios involving improper signaling—which are significantly more difficult to optimize—the authors propose a heuristic gradient-projection method. This method operates on the composite real representation of the covariance matrices to find high-performing improper signaling strategies.

Key Insights: Improper Signaling as a Spatial Alternative

While proper signaling wins when time-sharing is allowed, the paper highlights a vital role for improper signaling in "pure strategy" scenarios. If time-sharing is restricted—perhaps due to latency constraints, hardware limitations, or the complexity of synchronization—improper signaling becomes a powerful tool.

In these cases, improper signaling acts as a "poor man's time-sharing." It allows a transmitter to "back off" in one dimension (e.g., the imaginary part of the signal) to reduce the interference caused to another user. This spatial shaping allows for rate gains that exceed the best possible proper signaling strategy without requiring temporal averaging.

Results and Numerical Analysis

The paper validates its theoretical proofs through numerical simulations using synthetic channel realizations. The results confirm:

  • Theorem 1: Proper signaling without symbol extensions achieves the entire coded time-sharing rate region for the two-user SIMO IC with TIN.
  • Pure Strategy Superiority: In scenarios where time-sharing is prohibited, the proposed improper signaling heuristic consistently outperforms the globally optimal proper signaling strategy. The achievable rate region for improper signaling sits between the "proper pure" curve and the "proper time-sharing" boundary.

Conclusion and Future Work

This research provides a definitive answer for the two-user SIMO interference channel: proper signaling is sufficient if you can average over time, but improper signaling is a valuable asset if you cannot.

However, several questions remain for future exploration:

  • The K-User Case: For three or more users ($K > 2$), the phase interactions become significantly more complex. It is possible that improper signaling could provide benefits even with coded time-sharing in these larger networks.
  • Advanced Receivers: The current study assumes TIN. If receivers utilize Successive Interference Cancellation (SIC), the optimality of proper signaling may no longer hold.
  • MISO/MIMO Extensions: Extending these proofs to Multiple-Input Single-Output (MISO) or MIMO systems is non-trivial, as the optimization variables transition from scalars to matrices, complicating the channel enhancement logic.

Summary updated

Abstract

In the two-user Gaussian interference channel with Gaussian inputs and treating interference as noise (TIN), improper complex signals can be beneficial if time-sharing is not allowed or if only the data rates are averaged over several transmit strategies (convex hull formulation). On the other hand, proper (circularly symmetric) signals have recently been shown to be optimal if coded time-sharing is considered, i.e., if both the data rates and the transmit powers are averaged. In this paper, we show that both conclusions remain the same if single-input multiple-output (SIMO) systems with multiple antennas at the receivers are considered. The proof for the case with coded time-sharing is via a novel enhanced channel concept for the two-user SIMO interference channel, which turns out to deliver a tight outer bound to the TIN rate region with coded time-sharing. The result for the case without coded time-sharing is demonstrated by studying specific examples in which a newly proposed composite real gradient-projection method for improper signaling can outperform the globally optimal proper signaling strategy. In addition, we discuss how the achievable TIN rate region with coded time-sharing can be computed numerically.
BibTeX
@article{hellings2020two,
  author = {Christoph Hellings and Ferhad Askerbeyli and Wolfgang Utschick},
  title = {{Two-User SIMO Interference Channel with TIN: Improper Signaling versus Time-Sharing}},
  journal = {IEEE Trans. Signal Process.},
  year = {2020},
  doi = {10.1109/tsp.2020.3027903},
}
Figure