论文部分内容阅读
应用经济理论解决分布式资源分配问题一直是分布式计算领域的研究热点,其焦点是如何合理定价.目前的定价模型通常考虑资源成本、供求情况等因素对于资源价格的影响.本文通过对现有定价方法和资源分配算法的研究,提出一种基于历史交易记录和个体负载的分布式资源定价方法(HIILDPM),更符合分布式计算的特征;在此基础上设计了一种资源分配算法HIIL-CDA.结果表明,HIILDPM与现有定价方法相比具有较好的价格收敛性,较高的资源利用率和执行效率.
The application of economic theory to solve the problem of distributed resource allocation has been a hot topic in the field of distributed computing and its focus is how to reasonably pricing.Current pricing models usually take into account factors such as resource cost, supply and demand and other factors on resource prices.In this paper, Pricing method and resource allocation algorithm, a distributed resource pricing method (HIILDPM) based on historical transaction records and individual load is proposed, which is more in line with the characteristics of distributed computing. On this basis, a resource allocation algorithm HIIL- CDA.The results show that HIILDPM has better price convergence, higher resource utilization and higher efficiency compared with the existing pricing methods.