Network server load balancing using consensus-based control algorithm

The internet data center is growing swiftly in recent years as the Cloud computing is becoming popular. An internet data center is an apt example of a distributed network with dynamic loads, and a load balancing algorithm is needed in a data center to manage and distribute the loading to all server nodes appropriately. The implementation of a load balancer brings benefits to both the data center owner and Cloud users as it prolongs the lifespan of servers and improve the quality of service. To design a load balancing algorithm, various existing algorithms are studied. An average consensus algorithm is proposed for implementation in a network of server nodes. For verification purposes, the proposed algorithm is tested with various network topologies for its adaptability to distributed communication variance to the demand for scalability. The performance results are obtained through a series of experimental simulations; the results are then presented and analyzed in comparison to the conventional algorithm.