HLSA – A New Hybrid List Scheduling Algorithm for Fog Computing

A hybrid list scheduling algorithm is applied in fog computing with heterogeneity in available resources and incoming scheduling units. The scheduling units that need to be scheduled can be independent with no precedence constraints, so that tasks can be executed in parallel. On the other hand, precedence constraints can be present between tasks and represented by a Directed Acyclic Graph (DAG). Some scheduling algorithms are efficient for independent tasks, while others excel in handling dependency workflows. This paper proposes a Hybrid List Scheduling Algorithm (HLSA) for all scheduling unit types and examines the impact of incoming scheduling unit types on the performance of the proposed algorithm. HLSA assigns priority to sensitive time tasks in a cumulative way to achieve minimum latency for sensitive IoT applications in fog computing and to get minimum makespan, computation cost, and communication cost. Also, HLSA aims to achieve the highest utilization of the exploitation of gaps in processors.