Collective Intelligence based Framework for Load Balancing of Web Servers

The paper exploits the collective intelligence referred to as ant intelligence in World Wide Web with the aim to improve the performance of online web servers by balancing the load. The central concept of this idea is that a collection of agents can individually perform relatively simple, self-centered actions, such as the selection or rejection of hyperlinks in a web page for navigation, computing the load of server and aggregate these individual actions into a common substrate. The common substrate can then be evaluated to find the best available server to perform the task. This work aims to address the challenge of distributing intelligence to World Wide Web by contributing a unique ant-based intelligent load balancing framework which is able to integrate and synthesize knowledge on a scale far beyond the capabilities of individual humans.