Whispering: Joint Service Offloading and Computation Reuse in Cloud-Edge Networks

Area of Science:

• Computer Science
• Distributed Systems
• Cloud Computing

Background:

• The proliferation of Internet of Things (IoT) devices and increasing user demands strain traditional cloud computing infrastructures.
• Edge computing offers a solution by moving computation closer to data sources, addressing communication and computation challenges.
• Optimizing service migration (offloading) and computation reuse is crucial for efficient edge computing.

Purpose of the Study:

• To present Whispering, an analytical model for service offloading from cloud to edge.
• To minimize computational task completion time for user devices.
• To enhance resource utilization in edge computing environments.

Main Methods:

• Developed an analytical model named Whispering for service migration.
• Empirically investigated the impact of computation reuse for previously executed tasks.
• Proposed an adaptive task offloading scheme between edge and cloud.

Main Results:

• Whispering model demonstrated significant reductions in task completion time.
• Achieved up to 35% lower task completion times compared to edge-only execution.
• Achieved up to 97% lower task completion times when coupled with computation reuse compared to cloud-only execution.

Conclusions:

• The Whispering model effectively minimizes task completion times in edge computing.
• Computation reuse further amplifies the performance gains of the proposed offloading scheme.
• The adaptive offloading scheme enhances resource utilization and efficiency in distributed computing environments.