Openflow based traffic engineering for mobile devices
Hirannaiah, Radhika M.
MetadataShow full item record
Expansion in mobile devices has led to rapid growth in data traffic. Currently, networks are complex and difficult to manage. The exponential increase in audio/video streaming and continuous demand for network connectivity has necessitated the need for traffic steering and path management. Improving seamless network connectivity with reduced power consumption has been a constant goal in the development of mobile devices. The ability to configure quality of service (QoS) levels for applications on mobile devices is primarily driven by service providers and other players such as users and enterprises who have little or no way of configuring QoS. In the OpenFlow networking approach, having a separate control and data plane offers different levels of service. This research presents an architectural model to integrate the OpenFlow switch on a mobile device. The work here provides an algorithm design and methodology for using multiple controllers to act on a single OpenFlow switch installed on a mobile device while conforming to the OpenFlow standard. By associating a controller with entities such as a service provider, enterprise, and/or user, the switch on a mobile device is programmed for flows in a mutually exclusive way. The pattern of a pipeline processing scheme is intended to enable different paths driven by flows in various tables. The design solution allows enhanced traffic engineering to accommodate reduced latency, increased QoS, and a flexible model to enable users and enterprises to set their own traffic requirements. By incorporating flows related to network services instead of individual network-related programs on a mobile device, this model may reduce the number of components and thereby power consumption. With the proposed methods, it satisfies an infrastructural need that benefits mobile devices across different players in the process.
Thesis (Ph.D.)-- Wichita State University, College of Engineering, Dept. of Electrical Engineering and Computer Science