Introduction
In modern wireless communications, Adaptive Modulation (AM) is a critical technique used to maximize data throughput by adjusting the transmission rate (bits per symbol) based on fluctuating channel conditions. Traditionally, AM is treated as a single-agent optimization problem where a user independently selects a modulation level based on their own signal-to-noise ratio (SNR). However, this approach proves sub-optimal in complex network topologies like the Network-Coded Two-Way Relay Channel (NC-TWRC).
In an NC-TWRC, two users communicate through a central relay (R). This process occurs in two distinct phases: the multiple access (MAC) phase and the broadcast/amplify-and-forward (AF) phase. Because the relay's performance in the second phase is intrinsically linked to the rates chosen by both users in the first phase, a decision made by one user directly impacts the efficiency of the message delivered to the other. This paper proposes a shift from selfish optimization to a coordinated supermodular game approach to address this coupling.
System Model and the Coupling Problem
The fundamental challenge in NC-TWRC is that the relay combines signals from both users. If User A chooses a high-order modulation while User B chooses a low-order one, the relay's broadcast is limited by the "weakest link," leading to wasted spectral resources.
As shown in the figure above, the relay (R) receives $x_1$ and $x_2$ in Phase I and broadcasts a combined signal $z$ in Phase II. Traditional AM schemes fail to account for this interdependency, often resulting in mismatched rates that degrade the overall system efficiency. The authors argue that users must be viewed as strategic players in a game rather than isolated agents.
The Supermodular Game Framework
The authors model the interaction as a two-player game where each user selects a bit rate (strategy) to minimize a multi-objective cost function. This function balances three key factors:
- Transmission Error ($c_e$): Penalizes high bit error rates (BER).
- Transmission Rate ($c_t$): Encourages higher data throughput.
- Coupling Cost ($c_r$): A novel term that penalizes rate mismatches between the two users.
The core mathematical breakthrough is the proof that this interaction constitutes a supermodular game. In such games, the players' strategies exhibit "strategic complementarity." This means that if one user increases their data rate, the marginal benefit for the other user to also increase their rate increases. This synergy encourages users to "keep up" with one another, naturally driving the system toward higher spectral efficiency.
To reach the optimal state, the system employs Cournot tatonnement, an iterative process where users update their strategies based on the "best response" to their partner's previous move. The authors demonstrate that this process quickly converges to a Pure Strategy Nash Equilibrium (PSNE), typically within three iterations.
Performance and Results
The proposed supermodular game approach was benchmarked against conventional single-agent AM schemes and fixed-rate strategies using Rayleigh fading channel simulations. The results indicate a substantial leap in performance without compromising reliability.
As illustrated in the performance plot, the game-theoretic approach (SupG) significantly outperforms traditional AM. At an average SNR of 6dB, the proposed method enables the relay to exchange roughly 4 more bits per symbol than conventional methods. Crucially, this gain in spectral efficiency does not come at the cost of data integrity; the Bit Error Rate (BER) remains comparable to, and in some cases better than, traditional selfish AM schemes.
Conclusion
By reimagining adaptive modulation as a supermodular game, this research provides a robust framework for coordinating users in relay-based networks. The realization of strategic complementarity allows for a significant boost in spectral efficiency while maintaining fast convergence suitable for real-time hardware. Future work may extend this model to Bayesian games to account for imperfect channel state information, further bridging the gap between theoretical game theory and practical wireless deployment.
Summary updated