计算机支持的协同学习系统中负载均衡技术研究

摘要

CSCL system enables the learners distributed in different places to improve their achievements with shareable resources. However, the enormous data transmission and computing intensity in the session of different learners can result in the heavy load of a single server. Common load balancing techniques and scheduling algorithms are firstly introduced. Then, against the characteristics of CSCL system, a novel dynamic weighted round robin scheduling algorithm is proposed based on NAT load balancing technique and TCP Handoff mechanism to fulfill the balance of CSCL system.%计算机支持的协同学习系统能够帮助分布在不同地域的学习人员共享资源、实现学习者之间的合作.然而,系统在进行学习交流时所产生的数据流量和计算强度之大,通常让单一的服务器无法承受.本文首先介绍了常用的负载均衡技术和调度算法,并针对协同学习系统的特点,结合NAT负载均衡技术和TCP传递机制,提出了一种新的动态加权轮循均衡算法,实现了系统的负载均衡.