Gradient Kernelized Cat Boost Pre-Emptive Traffic Aware Data Distribution in Cloud Computing

Cloud provides a distributed computing flexible on-demand data services for the individual person and organizations to store and access their data anywhere and anytime through the internet without user. Every day a large volume of data is generated from different sources. The large volume of data occupies excessive bandwidth and it becomes directed to network traffic. The traditional processing tools are unable to solve effectively for network traffic during the large volume of data transmission. Therefore, a novel technique called Gradient kernelized Cat Boost Priority Pre-emptive traffic aware data distribution (GECOPP) method is introduced for minimizing the traffic occurrence level in the cloud environment during the large volume of data distribution. First, the server receives the number of user-requested data from the user. Then the GECOPP method classifies the user-requested data into an emergency or normal based on different resource factors such as energy, bandwidth, arrival data size, and session time by using Nesterov Gradient kernelized Cat Boost classifier. The task assigner stores the classified data in the priority queue to minimize the traffic occurrence level with lesser space complexity. Finally, a pre-emptive approach is employed for distributing the data to multiple data centers at different locations hence it minimizing the traffic occurrence level in the cloud environment. Experimental evaluation of the GECOPP Method is carried out on the different factors such as classification accuracy, space complexity, makespan, and throughput with respect to the number of data and number of cloud users. The observed results indicate that the GECOPP technique offers an efficient solution in terms of achieving higher accuracy, and throughput, and minimizing the makespan as well as space complexity than the conventional techniques.