视频流量逐渐在网络中占据主导地位,且视频平台大多对其进行加密传输。虽然加密传输视频可以有效保护用户隐私,但是也增加了监管有害视频传播的难度.现有的加密视频识别方法基于TCP(Transmission Control Protocol)传输协议头部信息和HT...视频流量逐渐在网络中占据主导地位,且视频平台大多对其进行加密传输。虽然加密传输视频可以有效保护用户隐私,但是也增加了监管有害视频传播的难度.现有的加密视频识别方法基于TCP(Transmission Control Protocol)传输协议头部信息和HTTP/1.1(HypertextTransferProtocol Version1.1)的传输模式,提取应用层音视频数据单元传输长度序列来实现视频识别.但是随着基于UDP(UserDatagramProtocol)的QUIC(QuickUDP Internet Connections)协议及基于QUIC实现的HTTP/3(HypertextTransferProtocol Version 3)协议应用于视频传输,已有方法不再适用.HTTP/3协议缺少类似TCP的头部信息,且使用了多路复用机制,并对几乎所有数据进行了加密,此外,视频平台开始使用多片段合并分发技术,这给从网络流量中精准识别加密视频带来了巨大挑战。本文基于HTTP/3协议中的控制信息特征,提出了从HTTP/3加密视频流中提取数据传输特征并进行修正的方法,最大程度复原出应用层音视频长度特征.面向多片段合并分发导致的海量匹配问题,本文基于明文指纹库设计了键值数据库来实现视频的快速识别.实验结果表明,本文提出的基于HTTP/3传输特性的加密视频识别方法能够在包含36万个真实视频指纹的YouTube大规模指纹库中达到接近99%的准确率,100%的精确率以及99.32%的F1得分,对传输过程中加人了填充顿的Facebook平台,在包含28万个真实视频指纹的大规模指纹库中达到95%的准确率、100%的精确率以及96.45%的F1得分,在具有同样特性的Instagram平台中,最高可达到97.57%的F1得分,且本方法在所有指纹库中的平均视频识别时间均低于0.4秒.本文的方法首次解决了使用HTTP/3传输的加密视频在大规模指纹库场景中的识别问题,具有很强的实用性和通用性.展开更多
This paper evaluates the performance of Internet Protocol Security (IPSec) based Multiprotocol Label Switching (MPLS) virtual private network (VPN) in a small to medium sized organization. The demand for security in d...This paper evaluates the performance of Internet Protocol Security (IPSec) based Multiprotocol Label Switching (MPLS) virtual private network (VPN) in a small to medium sized organization. The demand for security in data networks has been increasing owing to the high cyber attacks and potential risks associated with networks spread over distant geographical locations. The MPLS networks ride on the public network backbone that is porous and highly susceptible to attacks and so the need for reliable security mechanisms to be part of the deployment plan. The evaluation criteria concentrated on Voice over Internet Protocol (VoIP) and Video conferencing with keen interest in jitter, end to end delivery and general data flow. This study used both structured questionnaire and observation methods. The structured questionnaire was administered to a group of 70 VPN users in a company. This provided the study with precise responses. The observation method was used in data simulations using OPNET Version 14.5 Simulation software. The results show that the IPSec features increase the size of data packets by approximately 9.98% translating into approximately 90.02% effectiveness. The tests showed that the performance metrics are all well within the recommended standards. The IPSec Based MPLS Virtual private network is more stable and secure than one without IPSec.展开更多
为了提高实际数字视频广播的因特网协议(IP overDVB)系统的封装效率,详细分析了目前普遍采用的多协议封装(MPE)以及由IETF(the internet engineering taskforce)提出的单向轻简封装(ULE)各自的封装过程,并在此基础上给出了统一的封装效...为了提高实际数字视频广播的因特网协议(IP overDVB)系统的封装效率,详细分析了目前普遍采用的多协议封装(MPE)以及由IETF(the internet engineering taskforce)提出的单向轻简封装(ULE)各自的封装过程,并在此基础上给出了统一的封装效率计算公式。通过对局域网及卫星数据广播网2种实际网络环境中的IP包包长分布进行统计分析,详细比较了2种技术在实际系统中的封装性能,并进一步提出了提高实际系统效率的IP包封装方案。理论分析和实验仿真表明:ULE可提供比MPE较为高效的封装效率,针对实际系统提出的封装方案可最大化系统的封装效率,为优化实际系统设计提供重要的参考依据。展开更多
文摘视频流量逐渐在网络中占据主导地位,且视频平台大多对其进行加密传输。虽然加密传输视频可以有效保护用户隐私,但是也增加了监管有害视频传播的难度.现有的加密视频识别方法基于TCP(Transmission Control Protocol)传输协议头部信息和HTTP/1.1(HypertextTransferProtocol Version1.1)的传输模式,提取应用层音视频数据单元传输长度序列来实现视频识别.但是随着基于UDP(UserDatagramProtocol)的QUIC(QuickUDP Internet Connections)协议及基于QUIC实现的HTTP/3(HypertextTransferProtocol Version 3)协议应用于视频传输,已有方法不再适用.HTTP/3协议缺少类似TCP的头部信息,且使用了多路复用机制,并对几乎所有数据进行了加密,此外,视频平台开始使用多片段合并分发技术,这给从网络流量中精准识别加密视频带来了巨大挑战。本文基于HTTP/3协议中的控制信息特征,提出了从HTTP/3加密视频流中提取数据传输特征并进行修正的方法,最大程度复原出应用层音视频长度特征.面向多片段合并分发导致的海量匹配问题,本文基于明文指纹库设计了键值数据库来实现视频的快速识别.实验结果表明,本文提出的基于HTTP/3传输特性的加密视频识别方法能够在包含36万个真实视频指纹的YouTube大规模指纹库中达到接近99%的准确率,100%的精确率以及99.32%的F1得分,对传输过程中加人了填充顿的Facebook平台,在包含28万个真实视频指纹的大规模指纹库中达到95%的准确率、100%的精确率以及96.45%的F1得分,在具有同样特性的Instagram平台中,最高可达到97.57%的F1得分,且本方法在所有指纹库中的平均视频识别时间均低于0.4秒.本文的方法首次解决了使用HTTP/3传输的加密视频在大规模指纹库场景中的识别问题,具有很强的实用性和通用性.
文摘This paper evaluates the performance of Internet Protocol Security (IPSec) based Multiprotocol Label Switching (MPLS) virtual private network (VPN) in a small to medium sized organization. The demand for security in data networks has been increasing owing to the high cyber attacks and potential risks associated with networks spread over distant geographical locations. The MPLS networks ride on the public network backbone that is porous and highly susceptible to attacks and so the need for reliable security mechanisms to be part of the deployment plan. The evaluation criteria concentrated on Voice over Internet Protocol (VoIP) and Video conferencing with keen interest in jitter, end to end delivery and general data flow. This study used both structured questionnaire and observation methods. The structured questionnaire was administered to a group of 70 VPN users in a company. This provided the study with precise responses. The observation method was used in data simulations using OPNET Version 14.5 Simulation software. The results show that the IPSec features increase the size of data packets by approximately 9.98% translating into approximately 90.02% effectiveness. The tests showed that the performance metrics are all well within the recommended standards. The IPSec Based MPLS Virtual private network is more stable and secure than one without IPSec.
文摘为了提高实际数字视频广播的因特网协议(IP overDVB)系统的封装效率,详细分析了目前普遍采用的多协议封装(MPE)以及由IETF(the internet engineering taskforce)提出的单向轻简封装(ULE)各自的封装过程,并在此基础上给出了统一的封装效率计算公式。通过对局域网及卫星数据广播网2种实际网络环境中的IP包包长分布进行统计分析,详细比较了2种技术在实际系统中的封装性能,并进一步提出了提高实际系统效率的IP包封装方案。理论分析和实验仿真表明:ULE可提供比MPE较为高效的封装效率,针对实际系统提出的封装方案可最大化系统的封装效率,为优化实际系统设计提供重要的参考依据。
