Research On Fiber Routing Planning Algorithm Based On Cloud Platform

With the rapid development of telecom industry,the quantity of optical fiber communication networks resource has an exponential increasing trend,Fiber optic network topology is becoming more and more complex.Telecom operators take safe and efficient management on all kinds of resources.So users will have the higher reguirement on the End-to-End services.Therefore, how to realize the management and scheduling of the optical fiber network resources has become the key problem to optimize the overall resource allocation and save the construction cost.With the development of 4G and 5G.Telecom operators will gradually pay their attention on the applications of fiber optic,and they will improve the proportion of fiber optic services. In the future,telecom operators will focus on the programme of optical fiber's route.In the paper,we have finished the anlysis on the function of fiber network resources and fiber optic cable system, and the key and difficult point in the fiber optic network resource system is found.In addition,the research status of cloud computing in large-scale data processing is analyzed.It is found that the data processing of large scale data processing is a very promising research direction.In recent years, the development of cloud computing is rapid, and spark is the most popular one.So we can study the problem based on Spark.In this paper,we mainly combine the related knowledge of cloud computing and the programme of optical fiber's route on the system of fiber network resources management,and we will study the graph search and shortest path problem in large scale graph.In this paper,we will study the cloud computing,then analyze the Map Reduce programming model and Spark programming model, and compare the advantages and disadvantages between them.Through the analysis of the programme of optical fiber's route,the fiber optic network topology is obtained,and analyze it.Abstract the optical fiber network is abstracted as a graph model, and the search problem and the shortest path problem of the optical fiber routing planning problem is converted into a graph.Then, we analyze the basic storage model in graph theory.Improve the exiting storage model and the improved storage model is suitable for the cloud platform and parallel computing.the storage model of reading and writing efficiency is also improved.Finally, the BFS algorithm is analyzed, and the parallel design of BFS algorithm is realized.At the same time, the breadth first search algorithm and the shortest path algorithm based on Spark platform are realized by combining Spark technology. the conclusion is drawn based on the Spark platform by doing experiments on the platformand and analyzing the results of the experiments.