Reliable Transmission of Short Packets Through Queues and Noisy Channels Under Latency and Peak-Age Violation Guarantees

Rahul Devassy, Giuseppe Durisi, Guido Carlo Ferrante, Osvaldo Simeone, Elif Uysal-Biyikoglu

TL;DR

This paper tackles the challenge of ensuring ultra-reliable, low-latency communication for short data packets in wireless systems. It introduces a novel framework that jointly analyzes queuing and transmission dynamics, providing accurate estimates for delay and "peak-age" violation probabilities, even accounting for undetected errors, which are crucial for mission-critical applications.

What Problem Does This Paper Address?

Modern wireless applications like autonomous driving, industrial automation, and the tactile internet demand extremely reliable and low-latency communication links. These "Ultra-Reliable Low-Latency Communications" (URLLC) requirements are particularly stringent for short information packets. Traditional communication theory often relies on approximations like large-deviation theory or assumes ideal conditions (like perfect error detection), which fall short when dealing with the complexities of real-world wireless channels and queuing systems, especially for short packets. Furthermore, for applications that rely on timely updates (like sensor data), "peak-age of information" – the maximum time since the last received update – is a more relevant metric than simple packet delay. This paper addresses the need for accurate analytical tools to understand and guarantee both delay and peak-age performance under realistic conditions, including the impact of queuing and imperfect wireless channels.

Methods

The core of this work lies in a joint coding-queuing analysis that leverages finite-blocklength information theory. This approach allows for precise characterization of delay and peak-age violation probabilities without relying on large-deviation approximations.

Key methodological contributions include:

  • Variable-Length Stop-Feedback (VLSF) Coding Analysis: The paper models the transmission using VLSF codes, a general framework that encompasses standard Automatic Repeat reQuest (ARQ) and Hybrid ARQ (HARQ) schemes. This allows for a unified analysis.
  • Accurate Delay Violation Probability: A novel definition of steady-state delay violation probability is proposed, which explicitly accounts for both the delay exceeding a threshold and the occurrence of undetected errors. An upper bound is derived based on the complementary cumulative distribution function (CCDF) of the steady-state delay and the undetected error probability.
  • Finite-Blocklength Bounds: The analysis utilizes finite-blocklength results to derive bounds on achievable error probabilities and stopping times for VLSF codes. This is crucial for accurately modeling short packet transmissions.
  • Saddlepoint Approximation: For practical computation, the paper employs a saddlepoint approximation method to efficiently estimate the CCDF of the steady-state delay, demonstrating its high accuracy.
  • Peak-Age of Information Analysis: The framework is extended to analyze the peak-age of information, a critical metric for applications requiring up-to-date information. This analysis considers different queue management policies.
  • Queue Management Policy Evaluation: The paper investigates the impact of various queue management policies (FCFS-1, FCFS-2, LCFS-Q, and a novel LCFS-S) on peak-age violation probability, providing insights into optimal system design.

Datasets

Not discussed in the paper. The paper focuses on theoretical analysis and numerical simulations based on a defined system model.

Key Insights and Ideas

  • Undetected Errors Matter: The presence of undetected errors, where a packet is incorrectly decoded but acknowledged, significantly impacts both delay and peak-age violation probabilities. Ignoring them leads to inaccurate performance estimates.
  • Frame Size Optimization is Nuanced: Optimizing frame size for delay violation probability is not straightforward. Small frame sizes can lead to high error rates and retransmissions, while large frame sizes can increase latency. There's a sweet spot for maximizing throughput.
  • Peak-Age vs. Delay: For applications like sensor networks, peak-age of information is a more relevant performance metric than simple packet delay. Different queue management policies can significantly impact peak-age performance.
  • LCFS-S Policy Superiority: The proposed "Last-Come-First-Serve with Preemption in Service" (LCFS-S) policy demonstrates superior performance in minimizing peak-age violation probability compared to other standard policies.
  • Saddlepoint Approximation Accuracy: The saddlepoint approximation for delay violation probability is shown to be highly accurate, offering a computationally efficient alternative to exact methods.

Key Results

  • Accurate Performance Bounds: The paper provides accurate upper bounds on delay and peak-age violation probabilities for VLSF codes, accounting for queuing and undetected errors.
  • Parameter Sensitivity: Numerical results highlight the strong dependence of delay and peak-age violation probabilities on system parameters like signal-to-noise ratio (SNR), frame size, and undetected error probability.
  • Throughput Maximization: For ARQ with perfect error detection, the analysis shows that there exists an optimal frame size that maximizes throughput under a given delay violation constraint.
  • Policy Performance: The LCFS-S queue management policy is shown to outperform FCFS-1, FCFS-2, and LCFS-Q in minimizing peak-age violation probability.
  • Comparison of ARQ and VLSF: For small frame sizes, VLSF codes generally yield lower delay violation probabilities compared to ARQ, especially when considering the impact of frame size optimization.

Limitations and Future Work

  • Focus on Point-to-Point: The current analysis is primarily for a point-to-point communication system. The authors suggest future work to extend the framework to more complex scenarios like multiple independent flows, broadcast, multiple-access, and multi-hop settings.
  • Bi-AWGN Channel Assumption: While the analysis is generalizable to other channel models, the numerical results are presented for a binary-input additive white Gaussian noise (bi-AWGN) channel. Extensions to fading channels and interference are discussed as future directions.
  • Ideal Synchronization: The frame-asynchronous setup is introduced, but the primary analysis assumes ideal synchronization. Further investigation into the impact of imperfect synchronization could be valuable.
  • Computational Complexity: While saddlepoint approximation offers efficiency, exact computation of CCDFs can still be complex. Exploring further computational optimizations might be beneficial.

Summary updated

Abstract

This paper investigates the probability that the delay and the peak-age of information exceed a desired threshold in a point-to-point communication system with short information packets. The packets are generated according to a stationary memoryless Bernoulli process, placed in a single-server queue and then transmitted over a wireless channel. A variable-length stop-feedback coding scheme-a general strategy that encompasses simple automatic repetition request (ARQ) and more sophisticated hybrid ARQ techniques as special cases-is used by the transmitter to convey the information packets to the receiver. By leveraging finite-blocklength results, the delay violation and the peak-age violation probabilities are characterized without resorting to approximations based on larg-deviation theory as in previous literature. Numerical results illuminate the dependence of delay and peak-age violation probability on system parameters such as the frame size and the undetected error probability, and on the chosen packet-management policy. The guidelines provided by our analysis are particularly useful for the design of low-latency ultra-reliable communication systems.
BibTeX
@article{devassy2019reliable,
  author = {Rahul Devassy and Giuseppe Durisi and Guido Carlo Ferrante and Osvaldo Simeone and Elif Uysal-Biyikoglu},
  title = {{Reliable Transmission of Short Packets Through Queues and Noisy Channels Under Latency and Peak-Age Violation Guarantees}},
  journal = {IEEE J. Sel. Areas Commun.},
  year = {2019},
  url = {https://openreview.net/forum?id=S9kU4mWztC},
}
Figure