| With the development of the breaking code technology, on the one hand, scholars have to research the encoding algorithms deeply; on the other hand, people realized that the length of the cipher key in existing algorithms must be lengthened. Prime number, which is the parameter of many encoding algorithms which are in common use, its research value went without saying. So, to solve the serious problems such as long time testing, low testing efficiency and so on when the digit of prime number is increasing, this paper presents an optimized algorithm by adding pre-processing before running Miller-Rabin algorithm to improve the speed of testing by reducing the times of power-module operation in original algorithm.Optimized algorithm has data parallelity, so to improve the testing efficiency furthermore, we use the paralleling computing technology. One server will send parameters to several clients and these clients will begin to work almost at the same time; their workload is different form each other because of their different performance. This paper basically achieves the target of using N clients to finish the same workload within 1 N long time of the single machine. The whole designing process of the system is in an attempt to reducing the traffic furthest and the extra spending taking by communication. Finally the testing efficiency is improved by local storage and multiple thread schedulers combined with asynchronous programming module. The experimental result is satisfied.This paper adopted Winsock programming technology based on the .Net environment, multiple thread programming, asynchronous programming module and the object oriented system analyzing technology to program. Creating the static module formed of object layer, structure layer, subject layer and property layer, also the higher logic module to show the system function and realizing the distributed computing within the computer cluster network. This system runs stably, communicates well and it's reliable. |
PDF Full Text Request