Hybrid Algorithm for Deadline Constrained Resource Allocation in Cloud Computing

The users without sufficient computing resources habitually enjoy the Infrastructure as a Service (Iaas) from the cloud providers. The IaaS provider, in turn, achieves immense financial gains when the demand by the clients reaches a summit level. In the related scenarios, the task scheduling emerges a significant daunting challenge for the provider to offer the Quality of Service (QoS). The earlier-launched cloud federation model to tackle the related challenge is not found to be viable, in actual practice. Hence, the Self-adaptive Learning Particle Swarm Optimization (SLPSO) is envisioned to perk up the profit of IaaS provider with assured QoS devoid of any inter cloud agreement, though with an additional expenditure involved in the choice of the velocity strategy. In this work, the hybridization of Orthogonal Learning Particle Swarm Optimization (OLPSO) and Cuckoo Search (CS) known as the OLPSO-CS based scheduling technique is brought in for the purpose of scheduling. In the OLPSO-CS, each dimension of a particle symbolizes a task and a particle as a whole signifies all tasks priorities. The vital challenge, here, is concerned with allotment of various user tasks to incredibly enhance the income of Infrastructure as a Service (IaaS) provider concurrently ensuring the Quality-of-Service (QoS). The novel technique is well-geared to achieve superlative levels of quality scheduling solution. The engendered solution guarantees the user-level QoS and fine-tunes the IaaS provider reliability and financial giants. The improved outcomes originating from several test investigations vie with each other in establishing the fact that the OLPSO-CS algorithm significantly augments the IaaS provider gains vis-à-vis the benefits offered by the other techniques such as the standard Particle Swarm Optimization (PSO), CS, SLPSO and the OLPSO algorithms. It is also established from the test results that the hybridization of optimization algorithms is further suitable in the hybrid cloud scenario.