Energy-efficient power control for contention-based synchronization in OFDMA systems with discrete powers and limited feedback

Giacomo Bacci, Luca Sanguinetti, Marco Luise, Harold Vincent PoorEURASIP J. Wirel. Commun. Netw 2013 · DOI

Abstract

This work derives a distributed and iterative algorithm for mobile terminals to control transmit powers during the synchronization procedure in OFDMA systems, such as those specified by IEEE 802.16m and 3GPP LTE standards. By formulating a finite noncooperative game with discrete action sets and limited feedback, the proposed solution maximizes energy efficiency. Numerical results demonstrate that this approach improves both energy efficiency and timing estimation accuracy compared to existing alternatives, while maintaining minimal communication overhead.

Addressing the Synchronization Bottleneck

In cellular networks, "initial ranging" or "random access" is the critical first step for a mobile device to synchronize its timing with a base station (BS). Traditional methods rely on "power ramping," where devices incrementally increase transmit power until acknowledged. While effective for connectivity, this "brute force" approach is highly energy-inefficient—a major drawback for battery-powered devices.

Previous game-theoretic research often assumed continuous power levels and perfect signal-to-interference-plus-noise ratio (SINR) knowledge. This paper bridges the gap between theory and practice by incorporating discrete power steps and limited feedback from the base station, reflecting the constraints of real-world hardware and control channels.

Methodology: The DLF-BRSA Algorithm

The authors propose the Discrete and Limited Feedback Best Response Synchronization Algorithm (DLF-BRSA). In this framework, mobile terminals are modeled as "selfish" players in a noncooperative game.

  • Players: Synchronization Terminals (STs) attempting network entry.
  • Actions: A discrete set of transmit power levels (e.g., 1 dB steps).
  • Utility: Defined as the probability of successful code detection divided by the energy consumed.
  • Feedback: The BS provides only 4 bits per frame—one bit for detection status and three bits for a quantized version of the measured SINR.

Theoretical Insights: The Braess-type Paradox

A surprising finding of this research is that restricting devices to discrete power levels can actually improve global network performance—a phenomenon akin to the Braess paradox. In continuous power games, selfish behavior often leads to a single, less efficient Nash Equilibrium. However, discrete power levels create multiple equilibria. The authors show that if devices start at their lowest power setting, the DLF-BRSA naturally converges to a socially superior outcome, maximizing the collective utility of the group.

Figure 5: The ratio of social welfare (sum of utilities) for discrete vs. continuous power games, showing how discrete steps can enhance global efficiency.

Performance Evaluation

The algorithm was evaluated through 20,000 independent realizations of a 6-tap ITU modified vehicular-A channel model, simulating realistic multipath fading and path loss.

Timing Accuracy: The DLF-BRSA significantly outperforms standard deterministic methods. For instance, at a normalized distance of 0.6, the Mean Square Error (MSE) of the timing estimate was approximately five times lower than standard power-ramping techniques.

Impact of Terminal Density: As shown in Figure 3, the Normalized Mean Square Error (NMSE) increases with the number of terminals ($K$) and the size of the power step ($\Delta$). Despite the inherent noise of discrete steps, the algorithm maintains robust performance even with coarse quantization of the SINR feedback.

Figure 3: Normalized Mean Square Error (NMSE) of the timing estimate as a function of the number of terminals K for different power step sizes.

Conclusion and Future Directions

The DLF-BRSA offers a practical, low-overhead solution for energy-efficient synchronization in modern OFDMA networks. While the algorithm successfully finds a high-performing Nash Equilibrium, future research could explore cooperative strategies to reach the absolute social optimum. Additionally, testing the algorithm in high-density small-cell environments would further validate its robustness against inter-cell interference in 5G and beyond.

Summary updated

BibTeX
@article{bacci2013energy,
  author = {Giacomo Bacci and Luca Sanguinetti and Marco Luise and Harold Vincent Poor},
  title = {{Energy-efficient power control for contention-based synchronization in OFDMA systems with discrete powers and limited feedback}},
  journal = {EURASIP J. Wirel. Commun. Netw.},
  year = {2013},
  doi = {10.1186/1687-1499-2013-192},
}
Figure