Optimizing energy and latency in edge computing through a Boltzmann driven Bayesian framework for adaptive resource...

Area of Science:

• Edge Computing
• Resource Allocation
• Optimization Techniques

Background:

• Energy efficiency is a critical challenge in edge computing.
• Existing resource allocation algorithms often struggle to balance energy consumption and latency effectively.

Purpose of the Study:

• To develop a novel, energy-efficient resource allocation strategy for edge computing environments.
• To minimize energy consumption and latency through advanced optimization techniques.

Main Methods:

• A hybrid approach combining Boltzmann Distribution and Bayesian Optimization.
• Iterative parameter optimization using Bayesian Optimization for minimal energy and latency.
• A Boltzmann-driven probabilistic action selection mechanism with dynamic temperature adjustment for task prioritization.

Main Results:

• The proposed method significantly reduces energy consumption and average delay compared to Round-Robin and threshold-based algorithms.
• Achieved cumulative energy savings of up to 25% over baseline methods.
• Demonstrated adaptability and optimal scheduling action achievement under varying load conditions.

Conclusions:

• The integration of probabilistic selection and parameter optimization offers a new benchmark for energy-efficient resource scheduling.
• The developed framework is applicable to real-time, energy-aware edge computing applications.
• This research advances sustainable computing solutions in distributed systems through hybrid optimization.