一种基于信息熵的压缩感知流量恢复算法

A Recovery Algorithm for Compressed Sensing Flow Based on Information Entropy

以校园无线局域网为实验环境,采集大量真实用户流量数据集。利用信息熵,提出一种基于时间空间相关的压缩感知采样机制,并以此做为压缩感知的输入条件,提出一种压缩感知网络流量恢复算法,并与最近邻算法进行比较,实验证明本算法能准确恢复全网用户流量,从而更好地均衡网络负载,提高网络性能。

This thesis proposed a compressed sensing sampling mechanism related to time and space using informa- tion entropy based on the campus wireless LAN as the test environment to collect a large amount of real user traffic data. Then a recovery algorithm for compressed sensing network traffic could be achieved with the help of sampling mechanism above as the input conditions. Compared with the KNN algorithm, experiments show that this algorithm is of accurate recovery of the whole network of user traffic, so as to achieve better load balancing of the network and improve network performance.