A Simple Cooperative Diversity Method Based on Deep-Learning-Aided Relay Selection

Wei Jiang, Hans D. SchottenarXiv 2021 · DOI · arXiv

Abstract

Opportunistic relay selection (ORS) has been recognized as a simple but efficient method for mobile nodes to achieve cooperative diversity in slow fading channels. However, the wrong selection of the best relay arising from outdated channel state information (CSI) in fast time-varying channels substantially degrades its performance. With the proliferation of high-mobility applications and the adoption of higher frequency bands in 5G and beyond systems, the problem of outdated CSI will become more serious. Therefore, the design of a novel cooperative method that is applicable to not only slow fading but also fast fading is increasingly of importance. To this end, we develop and analyze a deep-learning-aided cooperative method coined predictive relay selection (PRS) in this article. It can remarkably improve the quality of CSI through fading channel prediction while retaining the simplicity of ORS by selecting a single opportunistic relay so as to avoid the complexity of multi-relay coordination and synchronization. Information-theoretic analysis and numerical results in terms of outage probability and channel capacity reveal that PRS achieves full diversity gain in slow fading wireless environments and substantially outperforms the existing schemes in fast fading channels.

Introduction: The Challenge of Fast Fading

In the race toward 5G-Advanced and 6G, wireless systems are pushing into higher frequency bands (mmWave, Terahertz) and high-mobility scenarios, such as high-speed trains and drones. While these offer massive bandwidth, they introduce a significant challenge: fast fading. In these environments, the wireless channel changes so rapidly that by the time a system measures signal quality and selects a relay, the information is often already "stale."

In cooperative networks, relays help pass signals from a source to a destination to ensure reliability—a concept known as cooperative diversity. The most efficient method is Opportunistic Relay Selection (ORS), which picks the single best relay based on current Channel State Information (CSI). However, the delay between measurement and transmission in fast-fading environments leads to "wrong selection," where the chosen relay is no longer the optimal performer. This "outdated CSI" problem severely degrades the performance of traditional relaying schemes.

System model for cooperative relaying showing the broadcasting and relaying phases.

Methodology: Predictive Relay Selection (PRS)

To address this, the proposed Predictive Relay Selection (PRS) method uses Deep Learning to forecast the future state of the wireless channel. Instead of acting on current measurements, the system acts on what the channel will be at the exact moment of transmission.

Deep Learning Architecture

The core of PRS is a Recurrent Neural Network (RNN), specifically Long Short-Term Memory (LSTM) or Gated Recurrent Units (GRU). These models are uniquely suited for time-series prediction. The predictor takes a history of CSI measurements and outputs a predicted value for the next time slot.

Architecture of the Deep Learning (DL) CSI Predictor integrated into the receiver chain.

As shown in the architecture above, the predictor is designed to be "transparent," sitting between the channel estimator and the relay selector. This allows existing ORS protocols to be upgraded to PRS with minimal structural changes. The analysis demonstrates that a PRS predictor requires less than 1% of the computational resources of a low-end commercial Digital Signal Processor (DSP), making it highly feasible for edge devices.

Key Insights and Results

The researchers derived closed-form mathematical expressions for outage probability and channel capacity, proving that PRS restores "full diversity gain." This means reliability improves exponentially as more relays are added, matching the performance of a system with perfect, instantaneous information.

Performance Highlights

  • Reliability Gains: In fast-fading scenarios (100 Hz Doppler shift), PRS outperformed traditional ORS by 12 to 15 dB.
  • Capacity Boost: In high-speed scenarios, PRS achieved nearly 3x the channel capacity of traditional ORS (e.g., 3 bps/Hz vs 1 bps/Hz at a 20dB SNR).
  • Robustness: The system remained effective despite additive noise, synchronization errors (up to 20° phase error), and varying fading statistics (Rician fading).

Conclusion and Future Work

PRS offers a powerful yet simple solution to the outdated CSI problem by leveraging the predictive capabilities of deep learning. By sticking to a single-relay selection model, it avoids the complex synchronization issues associated with multi-relay coordination.

Future research may explore online learning to allow models to adapt to new environments in real-time and extending the framework to MIMO (Multiple-Input Multiple-Output) systems. While the method is bound by the physics of the channel's coherence time, it represents a significant step forward for high-mobility 6G applications.

Reference: Wei Jiang and Hans Dieter Schotten, "A Simple Cooperative Diversity Method Based on Deep-Learning-Aided Relay Selection," IEEE Transactions on Wireless Communications (Preprint/arXiv:2102.03409).

Summary updated

Abstract

Opportunistic relay selection (ORS) has been recognized as a simple but efficient method for mobile nodes to achieve cooperative diversity in slow fading channels. However, the wrong selection of the best relay arising from outdated channel state information (CSI) in fast time-varying channels substantially degrades its performance. With the proliferation of high-mobility applications and the adoption of higher frequency bands in 5G and beyond systems, the problem of outdated CSI will become more serious. Therefore, the design of a novel cooperative method that is applicable to not only slow fading but also fast fading is increasingly of importance. To this end, we develop and analyze a deep-learning-aided cooperative method coined predictive relay selection (PRS) in this article. It can remarkably improve the quality of CSI through fading channel prediction while retaining the simplicity of ORS by selecting a single opportunistic relay so as to avoid the complexity of multi-relay coordination and synchronization. Information-theoretic analysis and numerical results in terms of outage probability and channel capacity reveal that PRS achieves full diversity gain in slow fading wireless environments and substantially outperforms the existing schemes in fast fading channels.
BibTeX
@misc{jiang2021simple,
  author = {Wei Jiang and Hans D. Schotten},
  title = {{A Simple Cooperative Diversity Method Based on Deep-Learning-Aided Relay Selection}},
  year = {2021},
  eprint = {2102.03409},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2102.03409},
}
Figure