Flexible computation offloading in a fuzzy-based mobile edge orchestrator for IoT applications.

Area of Science:

• Computer Science
• Distributed Computing
• Artificial Intelligence

Background:

• Internet of Things (IoT) applications face challenges with limited internet capacity and service delays.
• Cloud computing struggles with delay-sensitive and context-aware service requirements for IoT.
• Edge computing addresses these by moving computation closer to users.

Purpose of the Study:

• To propose a flexible computation offloading method for IoT applications using a fuzzy-based mobile edge orchestrator (MEO).
• To enhance computational resource management efficiency in edge and cloud environments.
• To reduce failed tasks and improve service times for time-critical IoT applications.

Main Methods:

• Developed a fuzzy-based MEO for intelligent application placement and workload orchestration.
• Integrated end mobile devices with edge and cloud computing systems.
• Defined new input and output parameters for the fuzzy-based MEO to handle network, computation, and task requirements.

Main Results:

• The proposed fuzzy-based MEO demonstrated superior performance compared to benchmark algorithms.
• Improvements were observed in WLAN delay, service times, and reduced number of failed tasks.
• Enhanced Virtual Machine (VM) utilization was also achieved.

Conclusions:

• The fuzzy-based MEO offers an effective solution for optimizing computation offloading in IoT systems.
• The method successfully balances resource management, task requirements, and network conditions.
• This approach is beneficial for delay-sensitive and resource-intensive IoT applications.