[Paper Review] Ultra-Reliable and Low-Latency Wireless Communication: Tail, Risk and Scale

Mehdi Bennis, Mérouane Debbah+1 more|arXiv (Cornell University)|Jan 1, 2018

Advanced MIMO Systems OptimizationEngineering89 references994 citations

TL;DR

The paper proposes a principled framework for URLLC that emphasizes tail behavior, risk under uncertainty, and scalability, and surveys enabling techniques across network layers and verticals.

ABSTRACT

Ensuring ultra-reliable and low-latency communication (URLLC) for 5G wireless networks and beyond is of capital importance and is currently receiving tremendous attention in academia and industry. At its core, URLLC mandates a departure from expected utility-based network design approaches, in which relying on average quantities (e.g., average throughput, average delay and average response time) is no longer an option but a necessity. Instead, a principled and scalable framework which takes into account delay, reliability, packet size, network architecture, and topology (across access, edge, and core) and decision-making under uncertainty is sorely lacking. The overarching goal of this article is a first step to fill this void. Towards this vision, after providing definitions of latency and reliability, we closely examine various enablers of URLLC and their inherent tradeoffs. Subsequently, we focus our attention on a plethora of techniques and methodologies pertaining to the requirements of ultra-reliable and low-latency communication, as well as their applications through selected use cases. These results provide crisp insights for the design of low-latency and high-reliable wireless networks.

Motivation & Objective

• Define end-to-end latency and reliability concepts for URLLC and contrast with traditional 4G/5G metrics.
• Survey key enablers and their trade-offs for achieving ultra-reliability and millisecond-scale latency.
• Propose a framework and methodological tools (risk, tail, scale) to guide design and analysis of URLLC networks.
• Highlight applications and use cases motivating tail-centric, scalable network designs.

Proposed method

• Provide precise definitions of latency (E2E, user plane, control plane) and reliability across URLLC contexts.
• Identify and discuss enabling techniques (short TTI, HARQ, edge caching/computing, grant-free access, NOMA, edge AI, network slicing, multi-connectivity) and their impact on latency and reliability.
• Introduce a tail-, risk-, and scale-centric framework for URLLC analysis, including concepts from risk-sensitive learning, extreme value theory, stochastic network calculus and mean-field methods.
• Discuss fundamental trade-offs (finite vs large blocklength, spectral efficiency vs latency, energy vs latency, reliability vs latency) and cross-layer design considerations.
• Highlight potential methodologies (risk-sensitive learning, distributed AI at the edge, network slicing, proactive dropping) to manage uncertainty and tails in URLLC.

Experimental results

Research questions

• RQ1 How can URLLC requirements be formalized beyond average latency and outage probability to account for tail behavior and end-to-end reliability?
• RQ2 What ensemble of tools (risk, tail, scale) are most effective for designing scalable URLLC networks under uncertainty and diverse verticals?
• RQ3 What are the key trade-offs between latency, reliability, spectral efficiency, and energy in practical 5G/6G architectures?
• RQ4 How can edge computing, caching, and intelligent scheduling contribute to meeting stringent URLLC demands across access, edge, and core networks?

Key findings

• URLLC requires moving beyond average metrics to tail-focused and end-to-end reliability frameworks.
• Multiple enablers (short TTI, edge caching/computing, dynamic eMBB/URLLC multiplexing, grant-free access, multi-connectivity, network slicing) are crucial for meeting stringent latency and reliability targets.
• A tail-, risk-, and scale-centric perspective helps articulate the uncertainties and system-wide design choices in URLLC.
• Risk-sensitive learning and other advanced optimization approaches can mitigate large losses under uncertainty in URLLC settings.
• There are unresolved questions and open research directions around optimal diversity, resource sharing, CSI accuracy, and cross-layer design for URLLC.