An Autonomous Time-Dependent SLA Negotiation Strategy for Cloud Computing

Cloud service level agreement negotiation is a process of joint decision-making between cloud clients and providers to resolve their conflicting objectives. With the advances of cloud technology, operations such as discovery, scaling, monitoring and decommissioning are accomplished automatically. Therefore, negotiation between cloud clients and providers can be a bottleneck if it is carried out manually. Our objective is to propose a state-of-the-art solution to automate the negotiation process for cloud environments and specifically infrastructure as a service category. The proposed negotiation strategy is based on a time-dependent tactic. For cloud providers, the strategy uniquely considers utilization of resources when generating new offers and automatically adjusts the tactic's parameters to concede more on the price of less utilized resources. In addition, while the previous negotiation strategies in literature trust offered quality of service values regardless of their dependability, our proposed strategy is capable of assessing reliability of offers received from cloud providers. Furthermore, to find the right configuration of the time-dependent tactic in cloud computing environments, we investigate the effect of modifying parameters such as initial offer value and deadline on negotiation outputs that include ratio of deals made, and inequality index. The proposed negotiation strategy is tested with different workloads and in diverse market conditions to show how the time-dependent tactic's settings can dynamically adapt to help cloud providers increase their profits.