TY - GEN
T1 - Channel and content aware 3D video scheduling with prioritized queuing
AU - Appuhami, Harsha D.
AU - Martini, Maria G.
AU - Hewage, Chaminda T.E.R.
PY - 2012/9/6
Y1 - 2012/9/6
N2 - In this paper we address 3D video delivery over wireless systems based on Orthogonal Frequency Division Multiple Access (OFDMA), by considering a Medium Access Control (MAC) layer scheduling method combined with a prioritized queuing mechanism to prioritize the most important video components/layers with the goal of improving the perceived quality of 3D video at the receiver. We consider colour plus depth 3D video and we exploit its properties in terms of importance of the different components for the perceived quality. The priority values of the Scalable Video Coding (SVC) encoded 3D video are signalled from the Application Layer to the MAC layer via cross-layer signalling. All the users attached to a Base Station feedback their sub-channel gain to the Base Station periodically via partial Channel State Information (CSI) and this information is used in the sub-channel allocation process at the scheduler. Thereby, the proposed scheduler always guarantees that the most important layers are scheduled over the sub channels with higher gain at each time slot of an OFDMA frame. Furthermore, we have established a Packet Loss Rate (PLR) threshold which is used by the scheduler to drop video layers which are worst affected by packet losses to save scarce radio resources. Simulation results show that the MAC layer PLR is improved in the prioritized colour/depth layers at the cost of slight increase in PLR in the low prioritized layers. This results in a global quality improvement for the prioritized case.
AB - In this paper we address 3D video delivery over wireless systems based on Orthogonal Frequency Division Multiple Access (OFDMA), by considering a Medium Access Control (MAC) layer scheduling method combined with a prioritized queuing mechanism to prioritize the most important video components/layers with the goal of improving the perceived quality of 3D video at the receiver. We consider colour plus depth 3D video and we exploit its properties in terms of importance of the different components for the perceived quality. The priority values of the Scalable Video Coding (SVC) encoded 3D video are signalled from the Application Layer to the MAC layer via cross-layer signalling. All the users attached to a Base Station feedback their sub-channel gain to the Base Station periodically via partial Channel State Information (CSI) and this information is used in the sub-channel allocation process at the scheduler. Thereby, the proposed scheduler always guarantees that the most important layers are scheduled over the sub channels with higher gain at each time slot of an OFDMA frame. Furthermore, we have established a Packet Loss Rate (PLR) threshold which is used by the scheduler to drop video layers which are worst affected by packet losses to save scarce radio resources. Simulation results show that the MAC layer PLR is improved in the prioritized colour/depth layers at the cost of slight increase in PLR in the low prioritized layers. This results in a global quality improvement for the prioritized case.
UR - http://www.scopus.com/inward/record.url?scp=84867770665&partnerID=8YFLogxK
U2 - 10.1109/WiAd.2012.6296554
DO - 10.1109/WiAd.2012.6296554
M3 - Conference contribution
AN - SCOPUS:84867770665
SN - 9781457721946
T3 - 2012 Wireless Advanced, WiAd 2012
SP - 159
EP - 163
BT - 2012 Wireless Advanced, WiAd 2012
T2 - 2012 Wireless Advanced, WiAd 2012
Y2 - 25 June 2012 through 27 June 2012
ER -