Research on Multi-Terminal's AC Offloading Scheme and Multi-Server's AC Selection Scheme in IoT

Area of Science:

• Computer Science
• Electrical Engineering
• Wireless Communications

Background:

• Mobile Edge Computing (MEC) and Simultaneous Wireless Information and Power Transfer (SWIPT) are crucial for enhancing Internet of Things (IoT) device performance and sustainability.
• Existing research often overlooks multi-server architectures, focusing primarily on multi-terminal systems.

Purpose of the Study:

• To optimize computing rate and cost in a multi-terminal, multi-server, and multi-relay IoT scenario.
• To develop an efficient algorithm for resource allocation and task offloading in complex IoT environments.

Main Methods:

• Formulated equations for computing rate and computing cost in the proposed multi-server IoT system.
• Employed a modified Actor-Critic (AC) algorithm combined with convex optimization for task offloading and time allocation.
• Utilized the AC algorithm to derive a scheme for minimizing computing costs.

Main Results:

• Achieved near-optimal computing rates and computing costs, significantly reducing program execution delay.
• Demonstrated effective utilization of energy harvested via SWIPT technology, improving overall energy efficiency.
• Validated theoretical analysis through comprehensive simulation results.

Conclusions:

• The proposed deep reinforcement learning approach effectively addresses the complexities of multi-server IoT environments.
• The developed algorithm enhances both computational efficiency and energy sustainability in IoT systems.
• This research provides a robust framework for optimizing resource management in advanced IoT networks.