基于期限的多请求数据检索问题研究

Research on deadline-based multi-requests data retrieval problem

给定一个包含多条信道的集合以及一个包含多个请求的集合,其中每一个请求包含多个请求数据项并且希望在一定期限内下载到,基于期限的多请求数据检索问题指当客户配有多条天线时寻找一个在期限内下载多个请求的数据检索序列,使得所有天线的最大访问延迟最小化。基于此,提出一种多请求的数据检索算法,以调度合适的天线检索这些请求并找到关于这些请求的检索序列,从而平衡在各天线上的访问延迟。针对单请求的数据检索,采用最大团思想寻找下载该请求中所有请求数据项的访问模式,使得检索该请求的访问延迟以及期限丢失率最小化。实验结果表明所提出的方法具有较好的性能。

Given a set of channels and a set of requests,where each request contained multiple requested data items and had a deadline, deadline-based multi-requests data retrieval problem referred that found a data retrieval sequence of downloading multiple requests in their deadlines under the constraint of the client equipped by multiple antennae, so that the maximal access latency among all antennae was minimized. A multi-requests data retrieval algorithm was proposed to schedule suitable antennae to retrieve these requests and find the data retrieval sequences about these requests,in order to balance the access latencies of all antennae. For solving single data retrieval, the maximal clique was also adopted to find the access pattern of each request for downloading all requested data items, so that the access latency and deadline miss ratio were minimized. The experimental results show that the proposed schemes have significant efficiency.