军事信息系统中基于内容流行度的情报分发模型研究

Research on Intelligence Distribution Model Based-on Content Popularity in Military Information System

军事移动应用环境的特点和对情报服务的特殊需求,对传统的集中存储+发布/订阅的情报分发方式提出了新的挑战。为了提高情报服务的时效性,减少对通信带宽的占用,降低情报服务器负载,本文参考内容中心网络架构,设计了包含"情报服务中心-情报服务站-情报用户"三个层级的移动情报服务系统模型以及情报缓存节点模型;提出了基于在线学习的情报内容流行度估计函数,并在此基础上,提出了基于内容流行度的情报分发算法。最后,通过仿真试验对比分析了新提出算法和传统算法的缓存命中率和收敛速度两个性能指标,验证了新的情报分发算法的有效性。

The characteristic of tactical application and the requirement of intelligence service,challenge the traditional way of intelligence distribution.In order to improve the efficiency of intelligence service,cut the cost of communication bandwidth,and decrease the overload of intelligence server,according to the architecture of content-centric networking,we propose a kind of three\level model of tactical intelligence service system,including intelligence service center,intelligence service station and intelligence user.We develop a popularity evaluating function for intelligence contents based on online learning.We propose an intelligence distribution algorithm based on content popularity.At last,we develop a simulator that implements the proposed algorithm and several classic algorithms to evaluate the performance using the metrics of overall cache hit ratio and convergence speed.The simulation results show that the proposed algorithm outperforms the other referenced ones.