Low-latency cloud-fog network architecture and its load balancing strategy for medical big data

Abstract

In order to apply fog computing to the field of medical big data, this paper proposes a low-latency hybrid cloud-fog network architecture for medical big data, which can solve the processing delay of business in cloud computing center architecture. In this architecture, edge network equipment such as routers and switches in the hospital are used to build a “fog computing” layer between the cloud server and terminals. Then, the computing service for medical data on the cloud is moved to fog equipment for processing, which reduces the processing delay of medical businesses. Besides, it reduces the computing load on cloud servers and improves the overall robustness of network. For further optimizing the processing delay of business in the above network architecture, we study the load balancing strategy in fog computing network. Due to the global search ability of bat algorithm is strong, the convergence speed is fast and it is easy to implement. Therefore, this paper uses bat algorithm to solve the optimization problem in medical big data scenario. Bat algorithm is better than genetic algorithm and particle swarm optimization on the unconstrained optimization problems. However, it also suffers from problems such as local optimization and slow convergence. To solve this problem, we utilize load balancing to initialize bat population data that improves the quality of solution for initial samples. After getting the best bats, a Powell local search is performed on them, which speeds up the convergence of our algorithm. Finally, simulation results show that the proposed hybrid cloud-fog network architecture can reduce the processing delay of medical big data and improve user experience effectively.