Research And Implement Of IREA Based On Pi Calculus

Distributed parallel computing is achievements of science and technology’s process, and it can share resource, coordinate load of multiple computer and make the computer operate suited task, so that the system’s parallel computing power can be sharply improved. Task scheduling of distributed environment is an important link of distributed parallel computing, and it can highly improve the performance of parallel computing. Task scheduling of distributed environment is a classic problem of scheduling theory, many excellent and efficient algorithm appeared, and interpersonal relationships evolution algorithm is one of them, which is a dynamic clustering algorithm based on task duplication and can solve the problem of static scheduling of distributed system multitasking to get the global optimal solution.Research of Calculus in distributed systems is developing very quickly, and Pi Calculus becomes an effective way to deal with structure of formal distributed system, because of its complete theoretical system and unique advantages of distributed architecture modeling. Pict language is a parallel programming language and it’s core semantic grammar is based on Pi calculus. Pict also expand and improve Pi calculus’ s semantics.This paper propose that combine Calculus in distributed systems and task scheduling of distributed environment, and use parallel calculus to implement task scheduling algorithms. This work mainly includes:(1) Using Pi calculus to model the frontier scheduling algorithm and the figure separation algorithm, which is the first and third step of interpersonal relationships evolution algorithm.(2) Using Pict language to implement the model of frontier scheduling algorithm and the figure separation algorithm based on Pi calclus.(3) Design different experimental scheme, by proposing tasks figure with different topology to compare with traditional high-level language.The comparing of experiment shows that, using Pi calculus and Pict language to model and implement IREA can improve the efficiency of IREA’s implement, then shows that use parallel calculus and parallel programming to distributed concurrent system’s task scheduling can improve the efficiency and performance of system.