Trellis Tone Modulation Multiple-Access for Peer Discovery in D2D Networks

Chiwoo Lim, Min Jang, Sang-Hyo Kim2018 · DOI

TL;DR

This paper introduces Trellis Tone Modulation Multiple-Access (TTMMA), a novel non-orthogonal multiple-access scheme for device-to-device (D2D) peer discovery. TTMMA significantly improves discovery range and capacity compared to traditional methods by using single-tone transmission and a message-passing demodulation algorithm, enabling more devices to be discovered within the same resource constraints.

What Problem Does This Paper Address?

In modern mobile networks, the explosion of data traffic is overwhelming core networks. Device-to-device (D2D) communication offers a solution by enabling direct communication between nearby devices, bypassing the core network. A crucial aspect of D2D is "peer discovery" – the process by which devices find each other. Existing peer discovery methods, often based on orthogonal frequency division multiple access (OFDMA), struggle with increasing device density. These methods require strict collision avoidance and detection, leading to complexity, delays, and increased overhead. When many devices try to discover each other simultaneously, orthogonal resource allocation becomes inefficient, limiting the number of discoverable devices. This paper addresses the challenge of efficiently and effectively discovering a large number of nearby D2D devices in a distributed network environment.

Methods

The core innovation is the Trellis Tone Modulation Multiple-Access (TTMMA) scheme. Unlike traditional methods that assign distinct resources, TTMMA allows multiple devices to transmit on the same resource segment, making it a non-orthogonal scheme.

Here's a breakdown of the key techniques:

  • Trellis Tone Modulation (TTM): This is a novel modulation scheme. It uses single-tone transmission, meaning each symbol is transmitted on a single frequency tone. The specific tone used for each symbol is determined by a "trellis diagram," which maps sequences of message bits to transitions between tones over time. This approach concentrates signal energy, leading to a better signal-to-noise ratio and thus a longer discovery range. It also inherently addresses the Peak-to-Average Power Ratio (PAPR) issue common in OFDM systems.
  • Non-Orthogonal Multiple Access (NOMA): By allowing multiple devices to share the same resource, TTMMA acts as a NOMA scheme. This is crucial for increasing discovery capacity, as more devices can attempt discovery simultaneously.
  • Message-Passing Demodulation: To handle the superposed signals from multiple devices transmitting on the same resource, the paper proposes a sophisticated multi-user detection algorithm based on message-passing. This algorithm efficiently separates and reconstructs the individual discovery messages from the combined signal.
  • Tone-Space Expansion and Pre-Prediction: To further enhance the demodulation process, especially for the initial symbols where superposition is more complex, techniques like tone-space expansion and pre-prediction are introduced. These methods help manage the complexity of decoding superposed signals and improve the accuracy of identifying individual device signals.
  • Successive Interference Cancellation (SIC) and Threshold Adjustment: While TTMMA aims to reduce reliance on SIC, it's mentioned as an optional enhancement for further performance gains. Threshold adjustment is also proposed to improve detection in noisy conditions.

Datasets

Not discussed in the paper. The paper focuses on theoretical development and simulation-based performance evaluation using defined channel models and parameters.

Key Insights and Ideas

  • Non-orthogonality is key for discovery capacity: The paper demonstrates that moving away from strict orthogonal resource allocation (like FDMA) and embracing non-orthogonal schemes is essential for significantly increasing the number of devices that can be discovered simultaneously.
  • Single-tone transmission for range and PAPR: Concentrating signal energy into a single tone per symbol not only improves the transmission power (and thus range) but also elegantly solves the PAPR problem inherent in OFDM, which is a significant practical advantage.
  • Message-passing as a powerful detector for NOMA: The paper highlights the effectiveness of message-passing algorithms in decoding complex superposed signals in a NOMA context, making it a viable solution for dense D2D discovery scenarios.
  • Trade-offs in trellis design: The design of the trellis diagram (number of states and degrees) involves a trade-off between transmission efficiency, bandwidth efficiency, and demodulation complexity.

Key Results

The paper presents simulation results demonstrating the superiority of TTMMA over conventional FDMA-based discovery:

  • Increased Discovery Capacity: TTMMA can discover approximately 1.5 times more devices compared to conventional FDMA-based discovery, even without using SIC.
  • Improved Range: The single-tone transmission and PAPR reduction contribute to a significantly improved discovery range.
  • Lower Computational Complexity: TTMMA achieves its performance gains with lower computational complexity compared to FDMA-based schemes, especially when SIC is involved in FDMA.
  • Performance in Different Channels: TTMMA shows robust performance across both block Rayleigh fading and TDL-A (frequency-selective) channels, outperforming FDMA in most scenarios.

Limitations and Future Work

  • Synchronization Assumption: The proposed scheme assumes synchronous transmission, which is achieved through various methods, but the paper doesn't delve deeply into the practical implementation challenges of achieving this synchronization in a fully distributed network.
  • Ideal Channel Estimation: While the paper mentions that TTMMA can detect channel gain directly, the performance evaluation assumes ideal channel estimation for SIC.
  • Future Work: The authors suggest exploring how TTMMA performance can be further improved. This could involve optimizing the trellis design, investigating more advanced message-passing variants, or exploring its application in other multiple-access scenarios beyond D2D discovery. The paper also notes that TTMMA's ability to achieve a PAPR of 1 by design makes it a promising candidate for uplink transmissions in future mobile communication systems, suggesting this as a potential area for future research.

Summary updated

BibTeX
@misc{lim2018trellis,
  author = {Chiwoo Lim and Min Jang and Sang-Hyo Kim},
  title = {{Trellis Tone Modulation Multiple-Access for Peer Discovery in D2D Networks}},
  year = {2018},
  doi = {10.3390/s18041228},
}
Figure