虚拟网络影射仿真平台ALEVIN的改进与实现  

Improvement and realization of VNE simulation platform based on ALEVIN

作　　者：曹浩彤 胡帅[1] 杨龙祥[1,2] 

机构地区：[1]南京邮电大学通信与信息工程学院,江苏南京210003 [2]南京邮电大学宽带无线通信与传感网技术教育部重点实验室,江苏南京210003

出　　处：《南京邮电大学学报（自然科学版）》2017年第1期47-53,共7页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：国家重点基础研究发展计划(973计划)(2013CB329104);国家自然科学基金(61372124;61427801);江苏省高校自然科学基金(11KJA510001)资助项目

摘　　要：网络虚拟化旨在解决现代网络架构的僵化问题。虚拟网络映射是网络虚拟化的一个重要研究点,目前已经有多种虚拟网络映射算法,因此急需一个集成仿真平台用于算法的性能比较。ALEVIN是虚拟网络影射的代表性开源仿真平台,集成了许多启发式算法,但ALEVIN不能自由生成网络拓扑环境。为弥补ALEVIN该缺陷,文中在ALEVIN上设计和实现了该功能,且开源可下载。此外,文中提出了一种精确解影射算法并添加到ALEVIN中。同时启用添加的网络拓扑环境生成功能,将该精确解算法与典型的启发式算法进行了性能比较。Network virtualization aims to deal with the ossification of current network architecture. Virtual network embedding （VNE） is a key issue on network virtualization. Multiple VNE algorithms are pro- posed. A integrated simulation platform is needed to evaluate performances of various VNE algorithms . ALEVIN is a typical VNE simulation platform. However, several flaws of ALEVIN have not been solved until now. The typical one is that ALEVIN cannot randomly generate network topology using Waxman generator. To solve the problem, this paper realizes the function of randomly generating network topolo- gy. The function is added to ALEVIN and corresponding codes are available. In addition, an exact VNE algorithm is proposed and added to the ALEVIN. The performances of the exact VNE algorithm and sev- eral typical heuristic algorithms are compared by the function of randomly generating network topology.

关 键 词：虚拟网络映射 仿真平台 虚拟网络需求 网络拓扑 精确解 

分 类 号：TN929.5[电子电信—通信与信息系统]