Video Streaming in Heterogeneous Networks
Research Students: Colin Bailey; Tim Porter; Mirghiasaldin Seyedebrahimi
Supervisor: Dr. X Peng
The demands for Multimedia services in heterogeneous networks, wired, wireless or the combination, are rapidly increasing. The main technical challenge in this area is how to achieve the required quality of service (QoS) and quality of experience (QoE) for the bandwidth-intense, delay-sensitive, and loss-tolerant multimedia applications at the minimum cost of network resources. This project aims to establish an optimisation model that addresses the constraints from different layers of the protocol architecture concerned, in order to attain the best trade-off between performance and cost factors for video streaming services. The research will be conducted using both analytical and empirical approaches in the following areas:
· Network protocols: Conventional network infrastructures such as the Internet, mobile communication networks, WLANs and WiMax are responsible for providing video streaming services to a wide range of users with varied QoS and QoE requirements. The functions enabled at the different layers of a specific protocol stack play important roles in ensuring the targeted performance to be achieved. Therefore, understanding the network protocols, especially TCP, UDP, IP and MAC, and their interactions is vital for establishing appropriate optimisation models to capitalise network resources and maximise performance gains. In particular, the cross-layer approach will be applied to the optimisation problems that are raised from any wired or wireless networks of running video streaming.
· Quality assessment metrics: The perceived playback video quality is an important measurement for end users’ QoE. For video streaming in TCP networks, the existing quality assessment metrics, such as PSNR for UDP networks, are not suitable for use in this environment. Other metrics, e.g. the buffer underrun probability, are used to characterise the buffer behaviour for TCP streams but unable to reflect users’ QoE. A pioneering work to introduce a new metric, namely Pause Intensity (PI), for TCP-based streaming is underway in our research group. Initial results have shown a good correlation between PI and viewers’ scores via subjective testing. Extensive studies on the analytical models and subjective assessment of PI are in progress and this quality assessment metric will be used as the object in the set-up of optimisation models.
· Performance optimisation: In order to meet the defined quality requirement in terms of the objective or subjective measurement or both, various networking and transmission technologies will be applied to maximise or minimise certain performance parameters, such as throughput (goodput), loss rate, latency and resource cost. These technologies include rate regulation, packetisation, spatial-temporal layer combination, data partitioning, erasure coding, scheduling, power adaptation, adaptive modulation and coding, and buffer design. Cross-layer optimisation will be the main strategy to apply in this work to achieve the objective for improving video quality and user fairness by optimally and dynamically utilizing the network resources.
T. Porter, and X.-H. Peng, “An objective approach to measuring video playback quality in lossy networks using TCP,” IEEE Communications Letters, Vol. 15 Issue 1, Jan. 2011, pp. 76-78.
M. Seyedebrahimi, and X.-H. Peng, “Investigation of PHY, MAC and APP Layers for Adaptive and Cross-Layer Optimization in IEEE802.11 WLANs,” in Proc. IEEE 10th Conference on Computer and Information Technology (CIT), May 2010.
C. Bailey, and X.-H. Peng, “Exploring the Effect of Buffer Behaviour on Perceived Video Quality,” accepted for IEEE 11th Conference on Scalable Computing and Communications (ScalCom), Aug. 2011.
C. Bailey, M. Seyedebrahimi, and X.-H. Peng, “Pause intensity: No-reference quality metric for video transmission in TCP networks,” in Proc. International Conference on Multimedia & Expo (ICME 2012), July 2012.
C. Bailey, Xiao-Hong Peng, “A quality driven adaptation scheme for DASH streaming,” in Proc. IEEE ChinaCom, August. 2013.
M. Seyedebrahimi, C. Bailey and X.-H. Peng, “Model and performance of a no-reference quality assessment metric for video streaming,” IEEE Trans. on Circuits and Systems for Video Technology, Vol. 23, No. 12, Dec. 2013, pp. 2034-2043.