Design And Realization Of Distributed Algorithms Based On Intelligence

With the widespread application of computer network, distributed system has become the basic form of computer applications. Distributed algorithm is one of the key technologies in the distributed systems, and it affects the performance of the system. As a result,it has been paid high attention by scholars and engineers.After brief introduction of distributed algorithms and smart algorithms,, the task allocation algorithm and coordinator election algorithm, as this paper's two main research topics, their current research situation are described.In this paper, particle swarm optimization algorithm, one of intelligent computation methods, is used to solve independent task scheduling problem in distributed system. The rules of adaptive inertia weight and maintaining the diversity of the population are proposed to improve the shortcoming that PSO algorithm may fall into the local optimum value easily; In order to maintain system load balance, a new fitness function is created; The improved algorithm is applied to the task scheduling problem, and after establishing a simple network model, it is tested and compared with other algorithms. The simulation results show that the improved algorithm performs well and has obvious improvement on load balance.Coordinator election is a basic question in distributed computation. The means of heuristic search and news parallel transmission were brought forward after analyzing a classical distributed election algorithm; design the corresponding election algorithm based on a established bi-directional ring election model; analyze the performance of new algorithm and compare it with classical election algorithm. The results show that the new algorithm takes less time on electing a new coordinator.The above work confirmed that introducing the intelligent thoughts to the design of distributed algorithms can ensure a fast and reasonable operation and so enhance the system efficiency.