Cost-Efficient and Latency-Aware Edge–Cloud Task Scheduling

A novel Cost-Efficient Dynamic Edge–Cloud Scheduling Strategy (CEDECSS) is developed to address the challenges imposed by the complexity and heterogeneity of edge–cloud environments. The total operational cost Tcost, comprising execution and penalty components, is minimized under resource capacity, latency, makespan, and Task Satisfaction Ratio (TSR) constraints using a multi-objective Lagrangian formulation. In CEDECSS, a linearized latency-satisfaction constraint and a cost-aware migration rule are incorporated to achieve equilibrium between performance and utilization. Experimental evaluation conducted using the San Diego Supercomputer Centre (SDSC) SP2 dataset with task loads ranging from 100 to 5000 demonstrates 60.16% shorter makespan, 95.05% lower total cost, 60.15% reduced latency, 47.57% higher virtual machine utilization, and 35.42% improvement in TSR compared to the Hybrid Particle Swarm Optimization–Genetic Algorithm. Furthermore, respective gains of 23.79%, 61.17%, 23.79%, 24.12%, and 16.25% are achieved over the Hybrid Ant Colony Optimization–Genetic Algorithm, validating its scalability and efficiency.