Revisiting Outage for Edge Inference Systems

2025-03-22

AI Summary

For 6G edge intelligence, ensuring joint end-to-end (E2E) inference accuracy and latency remains challenging. Method: This work introduces, for the first time, the notion of *inference outage (InfOut) probability*—a metric capturing the fundamental trade-off between communication overhead and inference reliability under strict latency constraints—departing from conventional channel-outage-centric communication paradigms. We establish a theoretical framework grounded in Gaussian approximation, discriminative gain distribution modeling, and E2E latency-constrained optimization, deriving a rigorous outage bound and a high-fidelity surrogate function (error < 3%). Results: Experiments demonstrate that, under identical latency constraints, the proposed approach reduces InfOut probability by up to 42% compared to communication-centric baselines, significantly enhancing E2E inference reliability.

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Reasoning under Uncertainty: Probabilistic Inference
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Cloud, edge and content delivery systems for the Web
• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

One of the key missions of sixth-generation (6G) mobile networks is to deploy large-scale artificial intelligence (AI) models at the network edge to provide remote-inference services for edge devices. The resultant platform, known as edge inference, will support a wide range of Internet-of-Things applications, such as autonomous driving, industrial automation, and augmented reality. Given the mission-critical and time-sensitive nature of these tasks, it is essential to design edge inference systems that are both reliable and capable of meeting stringent end-to-end (E2E) latency constraints. Existing studies, which primarily focus on communication reliability as characterized by channel outage probability, may fail to guarantee E2E performance, specifically in terms of E2E inference accuracy and latency. To address this limitation, we propose a theoretical framework that introduces and mathematically characterizes the inference outage (InfOut) probability, which quantifies the likelihood that the E2E inference accuracy falls below a target threshold. Under an E2E latency constraint, this framework establishes a fundamental tradeoff between communication overhead (i.e., uploading more sensor observations) and inference reliability as quantified by the InfOut probability. To find a tractable way to optimize this tradeoff, we derive accurate surrogate functions for InfOut probability by applying a Gaussian approximation to the distribution of the received discriminant gain. Experimental results demonstrate the superiority of the proposed design over conventional communication-centric approaches in terms of E2E inference reliability.

Problem

• Designing reliable edge inference systems meeting E2E latency constraints
• Addressing limitations of existing communication-centric reliability studies
• Optimizing tradeoff between communication overhead and inference reliability

Innovation

• Introduces inference outage probability for reliability
• Balances communication overhead and inference reliability
• Uses Gaussian approximation for tractable optimization