Resource efficiency of a Lamarckian evolution-based scheduler under horizontal scaling of computational resources

This work investigates the resource efficiency of a computational scheduling algorithm that incorporates Lamarckian evolution principles for task distribution among edge devices. The study addresses the problem of computational resource allocation with consideration of energy consumption and load balancing.
A comparison is conducted between a Lamarckian evolutionary algorithm and the NSGA-II genetic algorithm. Experimental results demonstrate that Lamarckian evolution proves effective for high-dimensional problems under limited computational budgets, yielding more accurate solutions. For low-dimensional problems, its application is not justified due to increased computational overhead.
The study concludes that the choice of scheduling method should depend on problem scale and available resources—a critical consideration for edge computing systems and distributed environments.