An Immune Co-Evolutionary Algorithm-Based Scheduling Approach for Dynamic Network Load Balancing

摘要

Future Internet should be capable of adaptability to the changes of different users and network environments.An important method for addressing this challenge may apply biologically-inspired computing approaches.We focus on the scheduling problem of network load balancing and design a scheduling agent to control migration behaviors of agents in the network in order to acquire load balancing.The immune co-evolutionary algorithm (ICEA) is proposed used by the scheduling agent.As a demonstration,a network environment with scheduling agent is built that can schedule agents to migrate and accomplish tasks.According to different task amounts,simulations are done through adopting and not adopting load balancing approach.We compare response time under different conditions.The results demonstrate that the scheduling agent can solve the problem of network load balancing.