The Research Of QoS Architecture And Algorithm On Digital Home Gateway

With the popularization of digital appliances and networks, digital home has become a major trend of information-based technology. However, the digital products in home network are still separate. The demand of building a broadband access, interconnected and content-sharing home network is very strong. Home network achieves the interconnection of a variety of devices with home gateway as the center and middleware technology as the platform.Home network contains a great variety of digital appliances. Meanwhile home multimedia devices have been more and more diversiform and powerful, which leads to high demands for QoS management in digital home network. Home network QoS management can be divided into wired and wireless network QoS as far as the transmission medium is concerned. From the management point of view, it can be divided into integrated services and differentiated services. There are some popular home network platforms, like UPnP, OSGi, Jini, etc. UPnP is the mainstream.This paper analyzes from these perspectives and compares the advantages and disadvantages of a variety of home network QoS management technology and architecture. There are many problems of traffic management in the existing QoS architecture, such as low priority traffic starvation, congestion and priority designation.On this basis, this paper presents a class control QoS architecture on digital home gateway. This architecture improves the QoS part of UPnP, so that all the device can use QoS services including devices which do not support UPnP. This architecture deploys QoS strategy in the home gateway. It introduces the following management mechanisms: traffic prioritization authentication mechanism, user priority allocation mechanism, flow classification bandwidth control mechanism and access control mechanism. Class control QoS framework sloves QoS issues from the perspective of traffic management. It ensures relative fairness, improves bandwidth utilization, user configurable and congestion control.Finally, this paper confirms the validity of class control QoS architecture by simulation experiments. It also summarizes and gives outlook to the application of this architecture.