Load Balancing

Devices that are connected on the internet and are exchanging data with internet brokers to receive requested services are a significant part of internet users. In order to manage and account well to IoT requests maximum processing power, speed in data transfer, and proper combining services in minimum time is needed. Since there is a large number of IoT devices which have a large scale, we have to use the abilities and services of cloud environment in order to solve its problems. So, service composition in a cloud environment is paid attention recently. We want to suggest an algorithm with the approach in this research, of improving factors propounded in the service composition problem like the number of clouds involved in service, number of services examined before responding to users’ requests SP and load balance between clouds. In this paper, the factor, similarity measure, is introduced and used to find the best cloud and composition plan in each phase which in addition to improving QoS metrics propounded in previous papers, it caused improving QoS metric of load balancing between clouds, prevention of formation of a bottleneck in clouds entrance. These changes, besides the proper load balancing, have avoided the clouds stop working suddenly and satisfied the users by presenting the services faster.

A greater community of researchers widely studies fog computing as it reduces the massive data flow to the existing cloud-connected network and performs better for real-time systems that expect a quick response. As the fog layer plays a significant role in a fog-cloud system, all of the devices participating in fog computing must be balanced with appropriate load to upstretched the system performance. The proposed method is founded on a tree-based dynamic resources arrangement mechanism that refreshes the fog clusters created using Fuzzy C Mean (FCM) to increase the speed of resource allocation. With the help of Fuzzy rule-based load calculation and intra-cluster job allocation, the load inside the group is maintained. The system also has the facility of inter-cluster job forwarding, which works on demand. A novel load balancing strategy, Real-Time Flexi Forwarded Cluster Refreshing System (RTFRS) is proposed by which all the tasks can be handled efficiently within the fog cloud system. The proposed system is designed so that overall complexity is not upraised and becomes suitable for fog computing architecture with low processing capacity by maintaining the quality of service. Experimental results show that the proposed model outperforms standard methods and algorithms used in fog computing concerning average turnaround time, average waiting time, resource utilization, average failure rate, and the load on the gateway.