Abstract
A spurt of progress in wireless power transfer (WPT) and mobile edge computing (MEC) provides a promising approach for Industrial Internet of Things (IIoT) to enhance the quality and productivity of manufacturing. Scheduling in such a scenario is challenging due to congested wireless channels, time-dependent energy constraints, complicated device heterogeneity, and prohibitive signaling overheads. In this paper, we first propose an online algorithm, called energy-aware resource scheduling (ERS), to maximize the system utility comprising throughput and fairness, with consideration on both system sustainability and stability. Based on Lyapunov optimization and convex optimization techniques, the proposed algorithm achieves asymptotic optimality for heterogeneous IIoT systems without prior knowledge of network state information (NSI). Subsequently, we extend the ERS algorithm to a more realistic scenario where the overhead and delay of NSI feedbacks are nonnegligible. The optimal scheduling decisions of the scenario are provided, and the optimality loss on system utility under outdated NSI is analyzed. Simulations verify our theoretical claims and demonstrate the gains of our proposed ERS algorithm over alternative benchmark schemes.
Introduction
In the era of the "smart factory," the Industrial Internet of Things (IIoT) demands high-performance data processing and sustainable energy solutions. This paper addresses the complex orchestration of IIoT networks where devices are both wirelessly powered and reliant on Mobile-Edge Computing (MEC). The authors introduce an online scheduling framework, Energy-aware Resource Scheduling (ERS), designed to balance high data throughput with hardware limitations and the reality of delayed network information.
System Model and Heterogeneity
A modern IIoT environment presents three major hurdles: sustainability, device heterogeneity, and information delay. Replacing batteries for thousands of devices is impractical; thus, they must harvest energy from wireless signals via Wireless Power Transfer (WPT). Furthermore, the network is composed of heterogeneous devices: Type-I devices have low-capacity storage and must use harvested energy immediately, while Type-II devices possess high-quality rechargeable batteries capable of storing energy for future use.
As shown in the system scenario above, an Access Point (AP) manages both the downlink WPT flow and the uplink offloading data flow. The time frame is divided into slots, where a portion $\mu_0(t)T$ is dedicated to WPT, and subsequent portions $\mu_i(t)T$ are allocated for individual devices to offload data to the edge computing circuit.
The ERS Algorithm
The researchers leverage Lyapunov Optimization to transform a long-term stochastic problem into a series of simpler, per-time-slot sub-problems. The framework includes two primary versions:
- ERS-RN (Real-time NSI): A baseline assuming the AP has instantaneous knowledge of all device states. It decouples the problem into distributed data collection at the device level and joint energy/time allocation at the AP.
- ERS-ON (Outdated NSI): A practical extension accounting for feedback intervals and signaling overhead. The AP makes decisions based on the last known state, with a mathematically proven bound on the resulting optimality loss.
By formulating these as convex optimization problems, the system can efficiently solve for the optimal sensing and offloading rates without requiring future knowledge of channel conditions.
Key Insights
- The [O(1/V), O(V)] Tradeoff: The algorithm uses a control parameter $V$. Increasing $V$ pushes the system closer to the theoretical maximum throughput but causes data queues (and thus latency) to grow linearly. This allows operators to tune the system for either speed or low latency.
- Solving the "Doubly Near-Far" Problem: In WPT-MEC systems, distant devices suffer twice: they receive less power and have weaker signals for offloading. By incorporating Proportional Fairness, ERS ensures these devices are not starved of resources.
- Exploiting Heterogeneity: The study demonstrates that Type-II devices can "save" energy during poor channel conditions and "spend" it when the channel is clear, a strategy the ERS algorithm exploits to maximize overall utility.
Performance Evaluation
The algorithm was evaluated via simulations using 3GPP-compliant parameters for path loss and bandwidth. The results confirm that ERS-ON significantly outperforms benchmark schemes such as Greedy algorithms and Homogeneous-only optimization.
As illustrated in the figure above, the proposed ERS-ON algorithm maintains a superior average throughput compared to benchmarks like EOT-ON, GAN, and PFN as the control parameter $V$ increases. Furthermore, the algorithm demonstrates high robustness to stale information, maintaining a Jain’s Fairness Index near 1.0 even with outdated network state information.
Conclusion
The ERS framework provides a robust solution for managing energy-constrained, heterogeneous IIoT networks. By achieving asymptotic optimality without requiring non-causal network knowledge, it offers a practical path toward sustainable and stable smart manufacturing. Future work may extend this model to multi-AP scenarios to address inter-cell interference in even larger industrial deployments.
Summary updated