A high-performance real-time video/voice over IP (VV oIP) applications have been implemented using an Ubuntu Enterprise Cloud (UEC) and it is denoted UEC-VVOIP. It really outperforms the previous VVoIP via P2P connection (called SCTP-IHU) because user does not need to know what is a real IP and browsing the web with TCP-Based RMTP protocol achieves a pretty good friendly user interface. Therefore, this scheme reduces computation load and power consumption dramatically at user side. We have employed back-propagation neural network (BPNN) together with particle swarm optimization (PSO) to appropriately tune seamless handoff and analyze network traffic over time. As a result it takes about 1.597 sec for the seamless handoff between base stations under mobile wireless network. In access control the rapid facial recognition and fingerprint identification via cloud computing has been done successfully within 2.18 seconds to identify the subject. In conclusion the performance of both measures is better than the alternative VVoIP using Hadoop platform we've done before.
展开▼