A Genetic Algorithm for Task Scheduling in Fog Computing
Saudi Digital Library
With the increasing number of Internet of Things devices, delivering real-time services has become a challenge. Conventional cloud computing provides central resources to process the Internet of Things application services. However, it is difficult to achieve efficient resource support using cloud computing alone as some Internet of things applications require a real-time and low-latency response. As a result, fog computing has been implemented to address the shortcomings of cloud computing by extending the cloud services to the network’s edge. Both fog and cloud computing will allow for the integration and interoperability of a large number of IoT devices and services across multiple domains. To achieve reasonable management of these heterogeneous resources, an effective resource scheduling scheme is required. This dissertation introduces a latency-aware Genetic Algorithm for task scheduling in fog computing to reduce the overall service latency. The Genetic Algorithm is tested in a simulated environment that considers the dynamics of the fog infrastructure. the performance of the Genetic Algorithm in terms of latency is evaluated and compared to the performance of an Edge Placement Algorithm. The simulation results show that our proposed solution outperforms the baseline competitor in terms of latency reduction especially on dense networks.