Unveiling the Importance of NOMA for Reducing AoI

Zhiguo Ding, Robert Schober, H. Vincent PoorICC 2023 · DOI

TL;DR

This paper demonstrates that Non-Orthogonal Multiple Access (NOMA) is a powerful tool for improving data freshness, measured by the Age of Information (AoI), in wireless networks. By implementing NOMA as a transparent "add-on" to legacy TDMA systems, the authors show that users can achieve significantly lower AoI by gaining more frequent transmission opportunities and drastically reducing their access delay.

What Problem Does This Paper Address?

In the era of 6G, applications like autonomous driving and industrial automation rely on ultra-fresh data. Traditional metrics like throughput or latency don't fully capture "freshness." Instead, researchers use Age of Information (AoI), which measures the time elapsed since the last successfully received data packet was generated.

The current standard for avoiding user interference is Orthogonal Multiple Access (OMA), specifically Time Division Multiple Access (TDMA). In TDMA, each user waits for a specific time slot to transmit. This creates two major bottlenecks for AoI:

  1. Access Delay: A user might generate data at the start of a frame but must wait several slots for their turn.
  2. Limited Opportunities: If a transmission fails due to poor channel conditions, the user must wait for the next entire frame to try again.

The authors investigate whether NOMA—a technique that allows multiple users to share the same time/frequency resources—can solve these bottlenecks without requiring a complete overhaul of existing legacy networks.

Methods

The researchers propose using Cognitive-Radio inspired NOMA (CR-NOMA) as an add-on to a legacy TDMA network.

The Setup: * Users are paired together (e.g., User A and User B). * In User A's original TDMA slot, they remain the "primary" user. However, User B is allowed to transmit simultaneously as a "secondary" user. * The base station uses Successive Interference Cancellation (SIC) to decode both signals. To ensure the system is "transparent" to the legacy setup, the secondary user's data rate is capped to ensure the primary user's performance isn't compromised.

Data Models Evaluated: 1. Generate-at-Will (GAW): A user generates a fresh packet immediately before any transmission attempt. 2. Generate-at-Request (GAR): All users generate their packets at the beginning of a frame (common in synchronized sensing/IoT).

The authors derive exact, closed-form analytical expressions for the average AoI under both models, providing a mathematical framework to evaluate NOMA's benefits against traditional TDMA.

Datasets

The paper does not use external real-world datasets. Instead, the authors validate their mathematical derivations using extensive computer simulations (Monte Carlo methods). They model wireless channels using independent and identically distributed (i.i.d.) complex Gaussian distributions to represent Rayleigh fading, a standard approach in wireless communications research.

Key Insights and Ideas

  • The "Queue Jump" Effect: In the GAR model, NOMA allows users scheduled late in a frame to transmit much earlier as secondary users. This effectively "reduces the wait," slashing the access delay that plagues OMA.
  • Redundancy as a Feature: NOMA provides a "second chance" within the same frame. If a user fails their primary slot, they can try again in their partner's slot.
  • Fairness Improvement: In TDMA, the user scheduled last in a frame naturally has a higher AoI than the first user. NOMA helps equalize this, making the network more "fair" in terms of data freshness across all devices.
  • SNR Sensitivity: The benefits of NOMA are most pronounced in the low Signal-to-Noise Ratio (SNR) regime. At very high SNR, the GAW model converges to TDMA performance because the first transmission attempt almost always succeeds, making the NOMA "second chance" unnecessary. However, for the GAR model, NOMA remains superior even at high SNR because it still reduces access delay.

Key Results

  • Significant AoI Reduction: For a system with 8 users, the simulation showed that NOMA could reduce the average AoI by more than 25% compared to TDMA (e.g., reducing normalized AoI from 28 to 20 in specific scenarios).
  • Analytical Accuracy: The derived closed-form expressions perfectly matched the simulation results, providing a reliable tool for network engineers to predict AoI performance.
  • Impact of Data Rate: The performance gain of NOMA over TDMA increases as the required data rate ($R$) increases, suggesting NOMA is particularly valuable for data-intensive status updates.

Limitations and Future Work

  • Complexity Trade-off: While NOMA improves AoI, it requires SIC at the base station, which increases computational complexity compared to simple TDMA.
  • User Pairing: The current study assumes simple two-user pairing. The authors note that the protocol can be extended to odd numbers of users or larger clusters, but the analytical complexity would increase significantly.
  • Static Scenarios: The model assumes i.i.d. channels. Future work could explore how high mobility (which causes rapid channel fluctuations) or structured interference affects these AoI gains.
  • Grant-Free Access: The authors suggest that applying these NOMA principles to grant-free transmission (where users transmit without waiting for a base station's permission) is a promising next step.

Summary updated

BibTeX
@inproceedings{ding2023unveiling,
  author = {Zhiguo Ding and Robert Schober and H. Vincent Poor},
  title = {{Unveiling the Importance of NOMA for Reducing AoI}},
  booktitle = {ICC},
  year = {2023},
  doi = {10.1109/icc45041.2023.10279047},
}
Figure