Agent-based automated dynamic SLA negotiation framework in the cloud using the stochastic optimization approach

Abstract

Optimization of the broker negotiation strategy is one of the challenging issues observed in the cloud-based e-commerce negotiation framework. This strategy can be optimized either in the context of pre-request optimization or long-term optimization. Most researchers have focused on the context of pre-request optimization using various utility functions (such as time, opportunity, and competition). In the context of long-term optimization, the current state-of-the-art negotiation strategies can increase the utility value and success rate of the parties to some extent, but they do not guarantee minimization of the negotiation states (rounds) involved during the negotiation process. In addition, the existing strategies cannot react to the stochastic, rational, emotional, and unknown behaviour of the opponents due to their deterministic behaviour, which may lead to a negotiation break-off between parties. To overcome such limitations, a novel stochastic behavioural learning negotiation (SBLN) strategy is proposed to further maximize the utility value and success rate.