Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Integrated Sensing-Communication-Computation for Edge Artificial Intelligence

Dingzhu Wen, Xiaoyang Li, Yong Zhou, Yuanming Shi, Sheng Wu, Chunxiao Jiang

TL;DR

The paper tackles the challenge of delivering efficient, private, and low-latency edge AI in 6G by tightly coupling sensing, computation, and communication (ISCC) for both FEEL and edge inference. It introduces task-oriented ISCC frameworks at the application layer (digital and analog FEEL, plus multi-device inference schemes) and develops modeling and design criteria, including the discriminant gain based on the symmetric KL divergence $D_{ ext{KL}}^{ ext{sym}}(p||q)$, to optimize accuracy under resource constraints. It further develops physical-layer ISCC techniques, presenting beamforming designs for dual-functional and triple-functional signals to balance sensing accuracy and computation/communication performance via SDPs with SDR and rank-reduction methods. The work demonstrates that triply-functional designs can outperform dual-functional designs as antenna resources grow, offering a pathway toward integrated, resource-efficient edge AI deployments in future networks.

Abstract

Edge artificial intelligence (AI) has been a promising solution towards 6G to empower a series of advanced techniques such as digital twins, holographic projection, semantic communications, and auto-driving, for achieving intelligence of everything. The performance of edge AI tasks, including edge learning and edge AI inference, depends on the quality of three highly coupled processes, i.e., sensing for data acquisition, computation for information extraction, and communication for information transmission. However, these three modules need to compete for network resources for enhancing their own quality-of-services. To this end, integrated sensing-communication-computation (ISCC) is of paramount significance for improving resource utilization as well as achieving the customized goals of edge AI tasks. By investigating the interplay among the three modules, this article presents various kinds of ISCC schemes for federated edge learning tasks and edge AI inference tasks in both application and physical layers.

Integrated Sensing-Communication-Computation for Edge Artificial Intelligence

TL;DR

The paper tackles the challenge of delivering efficient, private, and low-latency edge AI in 6G by tightly coupling sensing, computation, and communication (ISCC) for both FEEL and edge inference. It introduces task-oriented ISCC frameworks at the application layer (digital and analog FEEL, plus multi-device inference schemes) and develops modeling and design criteria, including the discriminant gain based on the symmetric KL divergence , to optimize accuracy under resource constraints. It further develops physical-layer ISCC techniques, presenting beamforming designs for dual-functional and triple-functional signals to balance sensing accuracy and computation/communication performance via SDPs with SDR and rank-reduction methods. The work demonstrates that triply-functional designs can outperform dual-functional designs as antenna resources grow, offering a pathway toward integrated, resource-efficient edge AI deployments in future networks.

Abstract

Edge artificial intelligence (AI) has been a promising solution towards 6G to empower a series of advanced techniques such as digital twins, holographic projection, semantic communications, and auto-driving, for achieving intelligence of everything. The performance of edge AI tasks, including edge learning and edge AI inference, depends on the quality of three highly coupled processes, i.e., sensing for data acquisition, computation for information extraction, and communication for information transmission. However, these three modules need to compete for network resources for enhancing their own quality-of-services. To this end, integrated sensing-communication-computation (ISCC) is of paramount significance for improving resource utilization as well as achieving the customized goals of edge AI tasks. By investigating the interplay among the three modules, this article presents various kinds of ISCC schemes for federated edge learning tasks and edge AI inference tasks in both application and physical layers.
Paper Structure (18 sections, 6 figures)

This paper contains 18 sections, 6 figures.

Table of Contents

  1. Introduction
  2. ISCC for Federated Edge Learning Tasks
  3. Vision
  4. Digital ISCC for Federated Edge Learning
  5. Analog ISCC for Federated Edge Learning
  6. ISCC for Edge AI Inference Tasks
  7. Vision
  8. ISCC based Edge Inference Framework
  9. Task-oriented Design Criteria
  10. Modeling of Sensing, Computation, and Communication
  11. ISCC Schemes for Multi-device Edge Inference
  12. Narrow-view Sensing based Multi-device ISCC schemes
  13. Wide-view Sensing based Multi-device ISCC schemes
  14. Physical Layer ISCC Techniques for Supporting Edge AI
  15. Vision: Motivation and challenges
  16. ...and 3 more sections

Figures (6)

  • Figure 1: Six Usage Scenarios of 6G Proposed by IMT-2030 IMT2030.
  • Figure 2: Integrated Sensing-Computation-Communication for Edge AI.
  • Figure 3: Integrated Sensing-Computation-Communication for Federated Edge Learning.
  • Figure 4: Integrated Sensing-Computation-Communication for Edge-device Collaborative Inference.
  • Figure 5: Physical Layer ISCC for Edge AI.
  • ...and 1 more figures