Abstract
In this paper, we study the capacity regions of two-way diamond channels. We show that for a linear deterministic model the capacity of the diamond channel in each direction can be simultaneously achieved for all values of channel parameters, where the forward and backward channel parameters are not necessarily the same. We divide the achievability scheme into three cases, depending on the forward and backward channel parameters. For the first case, we use a reverse amplify-and-forward strategy in the relays. For the second case, we use four relay strategies based on the reverse amplify-and-forward with some modifications in terms of replacement and repetition of some stream levels. For the third case, we use two relay strategies based on performing two rounds of repetitions in a relay. The proposed schemes for deterministic channels are used to find the capacity regions within constant gaps for two special cases of the Gaussian two-way diamond channel. First, for the general Gaussian two-way relay channel with a simple coding scheme the smallest gap is achieved compared to the prior works. Then, a special symmetric Gaussian two-way diamond model is considered and the capacity region is achieved within four bits.
Introduction and Problem Formulation
In network information theory, the "diamond channel" consists of a source and destination communicating through two parallel relays. While the one-way version is well-understood, the two-way diamond channel introduces a significant challenge: bidirectional interference. When two nodes (A and B) attempt to communicate simultaneously through the same relays, their signals collide, complicating the task of reaching theoretical capacity limits.
Traditionally, researchers struggled to find a strategy that allows both directions to hit their maximum theoretical "cut-set" bounds at the same time. Prior work either focused on simpler single-relay setups or provided achievable rates that were far from the theoretical limit. This paper aims to bridge that gap by finding coding strategies that prove the cut-set bound is actually achievable in deterministic models and nearly achievable in Gaussian models.
Methodology
The authors employ a two-step approach to characterize the capacity regions:
1. Linear Deterministic Model (LDM)
The system is first analyzed using a deterministic model, which ignores noise to focus purely on signal interactions and bit-level interference. The core technique is Reverse Amplify-and-Forward (RAF). In this strategy, the relay reverses the order of received bits before retransmitting. This "flips" the signal structure, allowing source nodes to cancel their own transmitted interference more effectively.
For complex channel conditions where simple RAF is insufficient, the authors developed specialized relay strategies involving bit repetitions and reordering to align interference perfectly. This ensures that "junk" interference from the other user lines up with bits the destination already knows, making them easy to subtract.
2. Gaussian Channel Extension
These deterministic strategies are translated into the Gaussian (noisy) domain using nested lattice codes. Lattice codes allow relays to perform "compute-and-forward" operations, treating the sum of signals as a new codeword rather than just noise.
Key Results
- Deterministic Capacity: The paper proves that the capacity region for the two-way linear deterministic diamond channel is exactly defined by the cut-set bounds. Notably, node A and node B can transmit at their individual maximum capacities simultaneously without trade-offs.
- Improved Gaussian Relay Gap: For a standard two-way relay channel (single relay), the proposed scheme achieves a gap of only 1 bit from capacity, an improvement over previous schemes that maintained a 3-bit gap.
- First Diamond Channel Gap: For a symmetric reciprocal Gaussian two-way diamond channel, the authors proved the capacity region is achievable within a 4-bit gap. This represents the first constant-gap result for this specific topology.
Conclusion and Future Work
While the deterministic model is fully solved, the general (non-symmetric) Gaussian diamond channel remains a challenge due to the complex power allocation required for lattice codes. Furthermore, this work assumes full-duplex nodes. Extending these tight bounds to half-duplex constraints, which are more common in current wireless hardware, remains a vital area for future research. Additionally, the specialized relay strategies require precise knowledge of channel parameters, which may introduce overhead in rapidly fading environments.
Summary updated