Integrated Quality of Service for Offline and Online Services in Edge Networks via Task Offloading and Service Caching

Area of Science:

• Edge computing
• Distributed systems
• Artificial intelligence

Background:

• Edge servers host both online and offline digital twin (DT) services.
• Offline DT caching impacts resources needed for online services.
• Existing research often neglects this resource conflict.

Purpose of the Study:

• To investigate the integrated quality of service (QoS) for online and offline edge services.
• To develop a joint optimization strategy for service caching and task offloading.
• To address the resource competition between online and offline DT services on edge servers.

Main Methods:

• Formulated a utility maximization problem considering resource usage and QoS.
• Addressed the NP-hard problem by reframing it as a Markov decision process (MDP).
• Developed a joint optimization algorithm using deep Q-network (DQN) for service caching and task offloading.

Main Results:

• The proposed algorithm effectively manages resources for both online and offline services.
• Achieved a significant enhancement in overall QoS utility.
• Demonstrated at least 14.01% utility improvement over baseline algorithms.

Conclusions:

• The DQN-based joint optimization algorithm successfully balances competing demands of online and offline DT services.
• This approach enhances edge computing efficiency and service delivery.
• Addresses a critical gap in understanding resource management for hybrid edge services.