Multicast scheduling for streaming video in single frequency networks
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
With development in wireless mobile communication and evolution to high capacity system like LTE, multimedia streaming over wireless channel becomes an interesting topic both in field of research and application. The bottleneck of this issue is limited available resources in the systems; therefore it is important how the system is allocating its resources to deliver different streams to the subscribers.
Transmitting data content in a mobile cell can be either uni-cast or multicast transmission. In multicasting type, data content is delivered simultaneously to all users within a group who already subscribed for that special stream. In both transmissions, an algorithm called scheduler decided which multicast group or user should be scheduled and receives the data content. Moreover the scheduler is responsible to dynamically change the transmission rate according to use-end’s channel condition. In this Master’s thesis, we analyze multicast scheduling algorithms for delivering multilayer and single-layer streams to the user-end in LTE-MBSFN network. We use three scheduler algorithms, Max-Sum, Max-Prod, RRB-Max_Sum, to set up the simulation system. We evaluate the performance of algorithms on MBSFN and single cell multicast transmission, as well as the
impact of using single-layer and multilayer streams. Comparing single cell-multicast transmission and MBSFN scenario, our simulation shows that the quality of received streams is not degraded when the cell traffic load is small on both network structures. On high cell load occasion transmitting the single layer stream in MBSFN, results in low quality of service. The streaming can be offline when the scheduler has access to the data as much as it wants (i.e. saved on the hard disk) or real-time when data contents are stored on size limited buffer. We show that sending a small data bust is not an efficient way of multicasting on MBSFN and wastes the resources. Therefore in this Master thesis, we propose a Multi-group- Max_Sum algorithm that aggregates the small packet together and forms a bigger packet to efficiently use the frequency resources. The system will approach its ideal performance (offline streaming) when using a Multi-group-Max_Sum scheduler. After specific length, the buffer size doesn’t have impact on the scheduler performance.
Place, publisher, year, edition, pages
2011. , 67 p.
EES Examensarbete / Master Thesis, XR-EE-LCN 2011:010
Engineering and Technology
IdentifiersURN: urn:nbn:se:kth:diva-53503OAI: oai:DiVA.org:kth-53503DiVA: diva2:470183
Master of Science - Wireless Systems
Fodor, Viktoria, Univ lektor/docent