SAVVY

University:   Hochschule Osnabrück, University of Applied Sciences, Germany

Researcher: Prof. Dr.-Ing. Ralf. Tönjes

Department of Electrical Engineering & Computer Science

Laboratory for RF and Mobilcommunication

 

Abstract

Within the next few years, we will see a convergence of mobile and broadcast services. The Digital Video Broadcast standard has been extended towards DVB-H (H = handheld) to reach mobile battery powered devices. Vice versa the mobile communication standards from 3GPP and 3GPP2 have just defined a broadcast mode to distribute mobile multimedia data simultaneously to large user groups, called MBMS and BCMCS respectively. However, some requirements have to be met to achieve this goal. The radio network is a limited resource. A variety of streaming, messaging and carousel based services share the same broadcast medium. Service demands concerning bandwidth requirements and timely delivery have to be matched with radio network capacity and timely availability for optimal grade of service. The SAVVY project researches adaptive scheduling algorithms for mobile broadcast services. It will study by simulations different strategies to improve grade of service, such as exploitation of caching capabilities of mobile devices, unequal distribution of multicast users and difference of service profiles.

 

Project Objective

Within the next few years, we will see a convergence of mobile and broadcast services. The Digital Video Broadcast standard has been extended towards DVB-H (H = handheld) to reach mobile battery powered devices. Vice versa the mobile communication standards from 3GPP and 3GPP2 have just defined a broadcast mode to distribute mobile multimedia data simultaneously to large user groups, called MBMS and BCMCS respectively.

However, some requirements have to be met to achieve this goal. The radio network is a limited resource. A variety of streaming, messaging and carousel based services share the same broadcast medium. Moreover in mobile systems the spectrum is shared with valuable communication services. Service demands concerning bandwidth requirements and timely delivery have to be matched with radio network capacity and timely availability for optimal grade of service. In particular in a mobile environment which is prone to transmission errors flexible scheduling mechanisms have to be developed that reduce the access time to mobile broadcast services.

The objective of the SAVVY project is to research adaptive scheduling algorithms for mobile broadcast services. Special focus shall be given to the mobile environment where mobility and error resilience impose special demands.

 

Approach

To achieve its goals the project will exploit in particular knowledge about requirements of the services to be scheduled and feedback about current network capacity. Broadcast services can be distinguished into video and audio streaming, messaging to push data files and carousels for periodic media distribution.

To cope with the erroneous transmission it has been suggested to broadcast the data periodically. Data carousels are one realisation. However carousels introduce a waiting time. The challenge in implementing this solution is to determine the broadcast schedule in such a way that the mean waiting for the users is minimized. In the literature te use of multiple queues and data segmentation have been successfully employed to reduce waiting time. Please note that the users may have subscribed to different services and different services will have different numbers of users. Moreover the caching capabilities of mobile devices, the unequal distribution of multicast users and the difference of service profiles will be taken into account.

In case of messaging mobile broadcast systems, such as MBMS and DVB-H have defined a file repair mechanism that redistributes missing packets on request after broadcasting. For messaging services the time constraints are usually less strict. However, the repair mechanism may require a significant bandwidth in particular cells with transmission errors that reduces the free capacity consequent services. A fair priorisation scheme is required allowing coexistence of messaging with file repair.

Streaming services aim at the immediate consumption of the data. To cope with transmission errors forward error correction is employed. Hence the required bandwidth depends on the channel error probabilities and the required quality of service. Joint scheduling of streaming, messaging and carousel services is still a challenge.

The SAVVY project will study different strategies to improve grade of service. For this a simulation tool has to be selected. The simulations will compare the different approaches and evaluate the performance of the scheduling algorithms in a mobile broadcast environment.

 

Project Data

Explorative university research project

Duration: 1.4.2006 – 31.12.2008

Participants: Michael Knappmeyer (Research Staff)