On the design of a novel joint network-channel coding scheme for the multiple access relay channel

Mikel Hernáez, Pedro M. Crespo, Javier Del Ser

Abstract

This paper proposes a novel joint non-binary network-channel code for the Time-Division Decode-and-Forward Multiple Access Relay Channel (TD-DF-MARC), where the relay linearly combines — over a non-binary finite field — the coded sequences from the source nodes. A method based on an EXIT chart analysis is derived for selecting the best coefficients of the linear combination. Moreover, it is shown that for different setups of the system, different coefficients should be chosen in order to improve the performance. This conclusion contrasts with previous works where a random selection was considered. Monte Carlo simulations show that the proposed scheme outperforms, in terms of its gap to the outage probabilities, the previously published joint network-channel coding approaches. Besides, this gain is achieved by using very short-length codewords, which makes the scheme particularly attractive for low-latency applications.

Introduction

In a Multiple Access Relay Channel (MARC), two or more sources send data to a single destination with the help of an intermediate relay. Traditionally, "network coding" (combining packets at the relay) and "channel coding" (adding redundancy to fight noise) were treated as separate steps. However, information theory suggests that treating them separately is suboptimal.

The challenge lies in designing a Joint Network-Channel Code (JNCC) that allows the destination to decode everything together efficiently. Previous attempts often relied on random coefficient selection for combining data at the relay or required long codewords to be effective. This paper seeks a way to optimize these coefficients and maintain high reliability even with the short data packets required for modern low-latency communication, such as IoT or real-time control systems.

System Model and Methods

The authors focus on a Time-Division Decode-and-Forward (TD-DF) strategy. The system architecture, illustrated in Figure 1, involves source nodes performing convolutional coding and modulation before transmitting to both the relay and the destination.

System model of the proposed TD-DF-MARC scheme showing the interaction between source nodes, the relay, and the joint iterative decoder.

Key components of the proposed approach include:

  1. Non-Binary Linear Combination: Instead of simple binary XORing at the relay, the relay decodes the source messages and combines them using coefficients from a non-binary finite field ($GF(2^q)$).
  2. Factor Graph Representation: The entire system—the two source convolutional codes and the relay's network code—is represented as a single, unified factor graph.
  3. Iterative Decoding: The destination uses the Sum-Product Algorithm (SPA) to pass "soft" information (probabilities) back and forth between the sub-graphs of the sources and the relay until it converges on the correct data.
  4. EXIT Chart Optimization: This is the core innovation. The authors use EXtrinsic Information Transfer (EXIT) charts to visualize how information is exchanged during decoding. By analyzing these charts, they can predict if a decoder will get "stuck" and select specific coefficients ($h$) that keep the "decoding tunnel" open, ensuring successful convergence.

Key Results

The paper evaluates the scheme using rigorous Monte Carlo simulations under Additive White Gaussian Noise (AWGN) and Rayleigh Fading channel models. The results demonstrate several significant advantages:

  • Superior Performance: In "Scenario A" (relay equidistant from sources and destination), the scheme achieved a gap of only 1.39 dB to the outage limit, compared to 2.7 dB or 3.4 dB in prior literature.
  • Latency Advantage: The authors achieved these results using only 144 complex dimensions, whereas competing schemes required between 2,176 and 27,000 dimensions to reach their targets. This makes it ideal for short-packet communications.
  • Robustness: The authors introduced a "Relay State Information" (RSI) signaling method that effectively mitigates error propagation if the relay makes a mistake in decoding the source messages.

Conclusion

This research proves that tailored selection of coefficients, rather than random selection, significantly improves the performance of JNCC schemes. By moving to non-binary finite fields and utilizing EXIT chart analysis, the proposed method achieves high reliability with very short codewords. Future work will explore higher-order modulation (like 16-QAM) combined with Bit-Interleaved Coded Modulation (BICM) to further yield better spectral efficiency.

Summary updated

Abstract

This paper proposes a novel joint non-binary network-channel code for the Time-Division Decode-and-Forward Multiple Access Relay Channel (TD-DF-MARC), where the relay linearly combines — over a non-binary finite field — the coded sequences from the source nodes. A method based on an EXIT chart analysis is derived for selecting the best coefficients of the linear combination. Moreover, it is shown that for different setups of the system, different coefficients should be chosen in order to improve the performance. This conclusion contrasts with previous works where a random selection was considered. Monte Carlo simulations show that the proposed scheme outperforms, in terms of its gap to the outage probabilities, the previously published joint network-channel coding approaches. Besides, this gain is achieved by using very short-length codewords, which makes the scheme particularly attractive for low-latency applications.
BibTeX
@article{hernaez2012design,
  author = {Mikel Hernáez and Pedro M. Crespo and Javier Del Ser},
  title = {{On the design of a novel joint network-channel coding scheme for the multiple access relay channel}},
  journal = {IEEE Journal on Selected Areas in Communications},
  year = {2012},
  doi = {10.1109/jsac.2013.130802},
}
Figure