Research Of Task Scheduling Based On Improved ACO Algorithm In The Cloud Environment

As an emerging technology, Cloud computing is known for academia and industry. Cloud computing is made up of a series of resources that can be virtualized and dynamically upgraded.Users can share and easily access through the network to these resources. Through dynamic extension, we can increase Cloud services. Users do not need master cloud computing technology, and only need rent cloud computing resources according to the need of individual or groups.The application group of Cloud computing is the general public, so its users are very broad, the tasks and data volume that need to deal with are huge. The system is almost in dealing with a massive task all the time, so how to reasonably distribute cloud resource and efficiently schedule tasks is very important for cloud computing. But because current development of cloud computing is not mature enough, research on task scheduling algorithm for a cloud environment is relatively less, and consideration of cloud task scheduling algorithm is unitary, we can not meet the requirements of users.Based on cloud computing architectures and present algorithm, this paper designed a kind of load balancing of ant colony optimization algorithm for task scheduling strategy, considering time, cost, bandwidth and reliability constraints. The scheduling algorithm not only fully considers QOS requirements, but also makes the whole system load stay in a state of relative balance. For maximizing the economic benefits, improving the utilization rate of resources, it makes important significance. Then this improved ant colony optimization algorithm is applied to the adaptive fault-tolerant model, so the optimization algorithm has some fault tolerance. It makes great progress in the reseach of reliability. Finally, a new task scheduling strategy began in CloudSim simulation environment. By setting different constraints, and considering load balance factor in the process of resource selection, we compared and analyzed the performance and load balancing of scheduling algorithm. The experiments show that the task scheduling algorithm in the cloud environment not only can make task processing time shorter, make executive cost lower, make the system load relatively balanced, improve resource utilization rate, but also can have certain tolerance.