سيدشايان حسيني

The industrial internet of things (IIoT) is a technology that has emerged in response to the increasing applicability of IoT in the industry sector, and is considered a significant advancement in the fourth industrial revolution. It involves transforming large and small industrial machines and equipment into smart devices that can understand their operational environment and react to events within that environment, resulting in significant progress in automating monitoring, management, and operational processes. One of the main challenges of this technology is the generation of unprocessed computational loads on a very large scale, which require powerful computing resources for processing. Smart devices, due to reasons such as energy source limitations, usage type, and limited processing resources, have a weak ability to process generated computational loads. Computation offloading techniques are introduced as a suitable solution to address the challenges in computing infrastructures such as edge computing, fog computing, and cloud computing. Software-defined networks (SDN) are a type of new network that offers a centralized point, called the SDN controller, which provides unified management of all entities within the network. The SDN controller allows the performance of each network entity to be changed at any time and place in line with defined objectives. Therefore, these networks are an ideal location for implementing load shedding because it can be more targeted and accurate based on processing requirements and network service quality. The acceptable threshold of delay for each load shedding operation in industrial areas is crucial. Although processing resources may have the ability to process a task, the result may surpass the acceptable delay threshold. Additionally, the delay in the processing queue of each processing node can also cause problems if it exceeds the acceptable threshold. The aim of this research is to propose a method for computation offloading in Industrial Internet of Things (IIoT). In this method, a greedy algorithm is used to minimize the overall system response time while considering the delay threshold and accurately computing the queuing delay. Additionally, this method has been implemented in a real testbed in combination with the software-defined controller RYU. The eva‎luation results generated by this testbed attribute better performance to this method in minimizing the overall system response time compared to the related similar method.

رايانش لبه , رايانش ابري , تخليه بار محاسباتي , شبكه نرم افزار محور , اينترنت اشياء صنعتي

edge computing , cloud computing , Computation Offloading , Software-defined networking , Industrial Internet of Thigs

Shayan Hosseini

Dr. Zeinab Movahedi