Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Goal-Oriented Wireless Communication Resource Allocation for Cyber-Physical Systems

Cheng Feng, Kedi Zheng, Yi Wang, Kaibin Huang, Qixin Chen

TL;DR

This work introduces goal-oriented wireless resource allocation for cyber-physical systems, framing data transmission as a means to advance CPS goals through an information utility gain. By reformulating bandwidth allocation as a submodular/marginal-gain decomposition and a knapsack problem, the authors enable scalable divide-and-conquer and greedy solutions with provable sub-optimality bounds. The framework is instantiated across four CPS tasks—data-driven decision making, edge learning, federated learning, and distributed optimization—with both raw and intermediate data transmissions, and validated via simulations showing improved CPS goal attainment under limited RBs compared to throughput-centric baselines. The results highlight the potential of co-designing CPS operations and wireless networks to prioritize semantically valuable data, enabling faster convergence, reduced costs, and robust performance in edge-enabled industrial systems.

Abstract

The proliferation of novel industrial applications at the wireless edge, such as smart grids and vehicle networks, demands the advancement of cyber-physical systems. The performance of CPSs is closely linked to the last-mile wireless communication networks, which often become bottlenecks due to their inherent limited resources. Current CPS operations often treat wireless communication networks as unpredictable and uncontrollable variables, ignoring the potential adaptability of wireless networks, which results in inefficient and overly conservative CPS operations. Meanwhile, current wireless communications often focus more on throughput and other transmission-related metrics instead of CPS goals. In this study, we introduce the framework of goal-oriented wireless communication resource allocations, accounting for the semantics and significance of data for CPS operation goals. This guarantees optimal CPS performance from a cybernetic standpoint. We formulate a bandwidth allocation problem aimed at maximizing the information utility gain of transmitted data brought to CPS operation goals. Since the goal-oriented bandwidth allocation problem is a large-scale combinational problem, we propose a divide-and-conquer and greedy solution algorithm. The information utility gain is first approximately decomposed into marginal utility information gains and computed in a parallel manner. Subsequently, the bandwidth allocation problem is reformulated as a knapsack problem, which can be further solved greedily with a guaranteed sub-optimality gap. We further demonstrate how our proposed goal-oriented bandwidth allocation algorithm can be applied in four potential CPS applications, including data-driven decision-making, edge learning, federated learning, and distributed optimization.

Goal-Oriented Wireless Communication Resource Allocation for Cyber-Physical Systems

TL;DR

This work introduces goal-oriented wireless resource allocation for cyber-physical systems, framing data transmission as a means to advance CPS goals through an information utility gain. By reformulating bandwidth allocation as a submodular/marginal-gain decomposition and a knapsack problem, the authors enable scalable divide-and-conquer and greedy solutions with provable sub-optimality bounds. The framework is instantiated across four CPS tasks—data-driven decision making, edge learning, federated learning, and distributed optimization—with both raw and intermediate data transmissions, and validated via simulations showing improved CPS goal attainment under limited RBs compared to throughput-centric baselines. The results highlight the potential of co-designing CPS operations and wireless networks to prioritize semantically valuable data, enabling faster convergence, reduced costs, and robust performance in edge-enabled industrial systems.

Abstract

The proliferation of novel industrial applications at the wireless edge, such as smart grids and vehicle networks, demands the advancement of cyber-physical systems. The performance of CPSs is closely linked to the last-mile wireless communication networks, which often become bottlenecks due to their inherent limited resources. Current CPS operations often treat wireless communication networks as unpredictable and uncontrollable variables, ignoring the potential adaptability of wireless networks, which results in inefficient and overly conservative CPS operations. Meanwhile, current wireless communications often focus more on throughput and other transmission-related metrics instead of CPS goals. In this study, we introduce the framework of goal-oriented wireless communication resource allocations, accounting for the semantics and significance of data for CPS operation goals. This guarantees optimal CPS performance from a cybernetic standpoint. We formulate a bandwidth allocation problem aimed at maximizing the information utility gain of transmitted data brought to CPS operation goals. Since the goal-oriented bandwidth allocation problem is a large-scale combinational problem, we propose a divide-and-conquer and greedy solution algorithm. The information utility gain is first approximately decomposed into marginal utility information gains and computed in a parallel manner. Subsequently, the bandwidth allocation problem is reformulated as a knapsack problem, which can be further solved greedily with a guaranteed sub-optimality gap. We further demonstrate how our proposed goal-oriented bandwidth allocation algorithm can be applied in four potential CPS applications, including data-driven decision-making, edge learning, federated learning, and distributed optimization.
Paper Structure (38 sections, 14 theorems, 52 equations, 8 figures, 2 tables, 4 algorithms)

This paper contains 38 sections, 14 theorems, 52 equations, 8 figures, 2 tables, 4 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. CPS Operation Considering Effects of Communication
  4. Semantic, Task-oriented, Goal-oriented, and Utility-Oriented Communications
  5. Efficient Communications for Learning
  6. Contributions and Organizations
  7. System Model
  8. Cyber-physical System Architecture
  9. Wireless Communication Model
  10. Goal-oriented RBA Problem
  11. System Goal Function $\mathcal{C} \left( \boldsymbol{z} \right)$
  12. Solution Methods
  13. Solution Challenges
  14. Divide-and-Conquer Approach
  15. Causality Problem
  16. ...and 23 more sections

Key Result

Proposition 1

When the negative CPS goal satisfies the submodular property, the total information gain can be bounded by the sum of individual information utility gain $\Delta _j$, shown as below:

Figures (8)

  • Figure 1: The architecture of the cyber-physical system. The system's ultimate objective is to minimize system goal function $\mathcal{C} \left( \boldsymbol{z} \right)$.
  • Figure 2: The communication paradigm for (a) data-driven decision making; (b) edge learning; (c) federated learning; (d) distributed optimization.
  • Figure 3: (a) The throughput and (b) the relative information utility gain of communication networks under different policies in different scenarios (data-driven decision-making).
  • Figure 4: (a) The decision cost under different policies in different scenarios; (b) the probability distribution of decision cost for the $500$ scenarios for three policies; (c) the probability distribution of decision cost for the $500$ scenarios for utility-based and hybrid based policies (data-driven decision making).
  • Figure 5: (a) The throughput and (b) the relative information utility gain of communication networks under different policies in different data collection rounds (edge learning). (c) The test accuracy under different data collection rounds. Marked points denote the points that are closest to the $0.9$ and $0.95$ accuracy level line. The numbers in the bracket denote (data collection round, test accuracy).
  • ...and 3 more figures

Theorems & Definitions (18)

  • Proposition 1
  • Proposition 2
  • Proposition 3
  • Proposition 4
  • Proposition 5
  • Proposition 6
  • Lemma 1
  • Lemma 2
  • Proposition 7
  • Proposition 8
  • ...and 8 more