CAJ | 학술논문

多服务器环境下网络具有随机性和异构性，导致传统方法由于算法收敛效率低、易出现局部最优等弊端，大大降低网络调度的效率，无法有效完成网络调度，提出一种基于重复博弈算法的多服务器环境下网络调度方法，塑造多服务器环境下网络调度模型，依据网络调度模型以及博弈论原理，通过最优反应函数对多服务器环境下网络进行调度，将其转换成目标规划问题，采用混沌粒子群优化算法对目标规划问题进行求解，从而有效完成多服务器环境下的网络调度。实验结果说明，所提的网络调度方案可提高网络调度的速度以及效率，能够有效处理多服务器环境下网络任务调度过程中出现的问题。
다복무기배경하망락구유수궤성화이구성，도치전통방법유우산법수렴효솔저、역출현국부최우등폐단，대대강저망락조도적효솔，무법유효완성망락조도，제출일충기우중복박혁산법적다복무기배경하망락조도방법，소조다복무기배경하망락조도모형，의거망락조도모형이급박혁론원리，통과최우반응함수대다복무기배경하망락진행조도，장기전환성목표규화문제，채용혼돈입자군우화산법대목표규화문제진행구해，종이유효완성다복무기배경하적망락조도。실험결과설명，소제적망락조도방안가제고망락조도적속도이급효솔，능구유효처리다복무기배경하망락임무조도과정중출현적문제。
Much randomness of network and server environment heterogeneity, lead to the traditional methods due to the low efficiency of algorithm convergence, easy to appear the most superior local defects, greatly reduce the efficiency of network scheduling, unable to effectively complete network scheduling, an algorithm based on repeated game network scheduling method, multi-server environment shape the multi-server environment network scheduling model, the theory of network scheduling model and game theory, and through the optimal reaction function of multi-server environment network scheduling, convert it into a goal programming problem, the chaotic particle swarm optimization algorithm to solve the problems of the goal programming, thus effectively complete network scheduling server environment. Experimental results indicate that the proposed network scheduling scheme can improve the speed and efficiency of network scheduling, able to deal effectively with multi-server environment network in the process of task scheduling problems.