Abstract
This paper investigates the achievable degrees-of-freedom (DoF) of large-scale interfering two-way relay networks (I-TWRN). The network model consists of $K$ pairs of communication nodes (CNs) and $N$ relay nodes (RNs), where $K \ll N$. Each pair of CNs communicates via a single relay node without a direct link, and interference among RNs is considered. Assuming only local channel state information (CSI) at each RN, the authors propose a distributed and opportunistic RN selection technique for three relaying protocols: amplify-and-forward (AF), decode-and-forward (DF), and compute-and-forward (CF). The study characterizes the asymptotically achievable DoF as $N$ increases and identifies the sufficient conditions on $N$ required to achieve optimal DoF. Simulation results demonstrate that the proposed distributed technique outperforms conventional schemes in achievable rate while maintaining low implementation complexity.
Introduction
In the evolving landscape of 5G and the Internet of Things (IoT), wireless networks are becoming increasingly dense. Managing interference in these large-scale environments is a critical challenge, particularly in two-way relay networks where multiple pairs of nodes attempt to communicate simultaneously. Traditional interference management often relies on Global Channel State Information (CSI), requiring every node to know the channel conditions of every other link in the network. However, in a system with hundreds or thousands of relays, the signaling overhead required to share this information becomes prohibitive.
This research addresses this bottleneck by exploring how a large pool of available relay nodes can be leveraged to simplify interference management. By using a distributed selection process based only on local information, the authors demonstrate that it is possible to achieve optimal Degrees-of-Freedom (DoF), effectively allowing a complex, crowded network to behave as if it were interference-free.
The Challenge of Interference in I-TWRNs
In an Interfering Two-Way Relay Network (I-TWRN), $K$ pairs of users share the same spectrum, leading to significant inter-pair interference. The primary goal is to maximize the DoF, which represents the number of independent signaling dimensions available for communication.
Existing solutions typically require centralized coordination or global CSI, which do not scale well. The authors instead propose a decentralized approach where each relay node makes an autonomous decision based on its local environment. This shifts the burden from complex signal processing and coordination to a selection problem: finding the "best" relay among many candidates.
Distributed Opportunistic Relay Selection (ORS)
The core of the proposed method is the Opportunistic Relay Selection (ORS) technique. This process operates in three distinct phases:
- Metric Calculation (Total Interference Level): Each relay node $n$ independently calculates its Total Interference Level (TIL). This metric accounts for both the interference the relay receives from other users and the "leakage" interference it would cause to other users during its transmission phase.
- Distributed Selection via Back-off Timers: To avoid centralized control, the relays employ a back-off timer mechanism. Each relay initializes a timer inversely proportional to its TIL. The relay experiencing the lowest interference (and thus having the shortest timer) expires first and broadcasts a short "flag" signal. Upon hearing this signal, all other relays stay silent, ensuring that only the most favorable relays are active.
- Relaying Protocol Execution: Once selected, the relays operate under one of three protocols:
- Amplify-and-Forward (AF): The relay amplifies the received signal and retransmits it.
- Decode-and-Forward (DF): The relay decodes the individual messages before retransmission.
- Compute-and-Forward (CF): The relay uses physical-layer network coding (lattice codes) to map the combined signals into a single linear combination, which is then forwarded.
The Decoupling Principle and Scaling Laws
The most significant theoretical insight of this work is the Decoupling Principle. The authors prove that as the number of available relays $N$ grows, the probability of finding a relay that is nearly "orthogonal" to the interference space approaches unity. Essentially, with enough candidates, the network can always find a subset of relays that are virtually isolated from the rest of the interference.
To achieve the optimal DoF of $K$, the authors derive a specific scaling law: the number of relays $N$ must grow at a rate faster than the Signal-to-Noise Ratio (SNR) raised to the power of $2(K-1)$. This provides a concrete mathematical requirement for network designers aiming to implement these large-scale systems.
Key Results and Performance
The study's findings highlight the strengths and weaknesses of different relaying strategies:
- AF and CF Superiority: Both the Amplify-and-Forward and Compute-and-Forward protocols were shown to achieve the optimal DoF of $K$. This means the network can support $K$ simultaneous two-way conversations without them interfering with each other, provided $N$ is sufficiently large.
- DF Limitations: The Decode-and-Forward protocol is limited to $K/2$ DoF. This is because DF requires the relay to decode individual messages, a task that is significantly more sensitive to interference than the signal-mixing or mapping approaches used in AF and CF.
- SNR Crossover: Numerical simulations indicate that while traditional "max-min-SNR" selection (which ignores interference) works well in low-SNR regimes, the proposed ORS technique is vastly superior as SNR increases and the system becomes interference-limited.
Conclusion
This research demonstrates that the complexity of large-scale wireless networks can be mitigated by turning the abundance of nodes into an advantage. By using distributed opportunistic selection, networks can achieve high throughput with minimal overhead and only local CSI. While challenges remain—such as the computational complexity of lattice codes for CF and the assumptions of channel reciprocity—this work provides a robust framework for the design of future high-density wireless communication systems.
Summary updated