چكيده به لاتين
Nowadays, due to the importance of optimizing energy consumption in the modern world, using smart meter devices and Advanced Metering Infrastructure (AMI) have become popular. Additionally, with the continual expansion of smart cities and increasing use of wireless sensor networks, in order to reduce the respond delay to delay sensitive processing requests, using Fog and Edge technologies and existence of processing units which are close to the users instead of Cloud technology is felt necessary. Therefore, assuming that the memory storage has been increased and the current CPUs has been upgraded in the next generation of smart meters, they will be able to be used as a processing unit for performing distributed processing tasks which are requested by sensors or other devices such as IoT devices, mobile phones, weather sensors, traffic control devices, trackers, water or gas meters, and, etc. . Put differently, the data collector and the external processing units will come close to sensors and other devices which delivers the request for mentioned services. In this paper we are going to propose a new architecture and mechanism for resource allocation in Fog systems to use smart meters as distributed processing nodes. There are different resource allocation solutions in Fog systems like heuristic and meta-heuristic methods but because of smart meters’ differences, a new approach is needed. These differences include some responsibilities like undertaking a processing task while doing nothing (or are idle). Or while they are processing the distributed tasks, their main task schedule can arrive so they should make an interrupt and give over the distributed task to another idle node. In addition, we are going to address the resource allocation problem in Fog systems considering the clients and providers (Fog services providers) benefits. To do so, we propose a method using the Monte Carlo Tree Search (MCTS) to find the best solution.
