«End-to-End QoS Provision over Heterogeneous IP and non IP Broadband Wired and Wireless Network Environments A dissertation submitted in satisfaction ...»
The transmitter and the receiver reside on the same system (PC) in order to avoid issues that arise from synchronization errors or/and diﬀerences in system clocks . The video traﬃc is transmitted from the source network interface, which is connected at the ingress router of autonomous system AS1, passes through the three diﬀerent network domains and is ﬁnally returned back to the source system. For each generated packet, identiﬁed by a unique sequence number, the departure and arrival timestamps, and the type of payload that contains, are obtained. When a packet does not reach the destination, it is counted as a lost one. It is not only of interest the amount of lost packets, but also the type of content that packets have in their payload. Furthermore, not only the actual loss is important for the perceived video quality, but also the delay of packets/frames and the variation of the delay, usually referred to as packet/frame jitter. The packet/frame jitter can be addressed by so called play-out buﬀers. These buﬀers have the purpose of absorbing the jitter introduced by the network delivery delays. It is obvious that a big enough play-out buﬀer can compensate any amount of jitter. There are many proposed techniques in order to develop eﬃcient and optimized play-out buﬀer, dealing with this particular trade-oﬀ. These techniques are not within the scope of the described testbed. For our experiments the play-out buﬀer is set to 1000ms.
In order to measure the improvements in video quality by employing H.264/MPEGAVC, I use the Peak Signal to Noise Ratio (PSNR) and the Structural Similarity (SSIM)  metrics. P SN R is one of the most widespread objective metric for quality assessment and is derived from the Mean Square Error (MSE) metric, which is one of the most commonly used objective metrics to assess the application level QoS of video transmissions .
Let’s consider that the video sequence is represented by V (n, x, y) and Vor (n, x, y), where n is the frame index and x and y are the statial coordinates. The average P SN R of the decoded video sequence among frames at indices between n1 and
n2 is given by the following equation:
where V denotes the maximum greyscale value of the luminance. The average M SE of the decoded video sequence among frames at indices beteen n1 and n2
is given by:
Note that, the P SN R and M SE are well-deﬁned only for luminance values.
As it mentioned in , the Human Visual System (HVS) is much more sensitive to the sharpness of the luminance component than that of the chrominance component, therefore, I consider only the luminance P SN R.
SSIM is a Full Reference Objective Metric  for measuring the structural similarity between two image sequences exploiting the general principle that the main function of the human visual system is the extraction of structural information from the viewing ﬁeld. If v1 and v2 are two video signals, then the SSIM
is deﬁned as:
This section evaluates the performance of the proposed testbed conﬁguration through a set of four experimental cases. In this chapter, I study the performance of our framework by enabling or disabling scalable video coding, or by enabling or disabling prioritized transmission. The quality gains of scalable video coding in comparison with non-scalable video coding and the quality gains of prioritized transmission in comparison with non-prioritized transmission are compared in detail.
The ﬁrst experiment refers to a single layer MPEG-4 video stream transmission, where both DiﬀServ and BM mechanisms are not applied to the heterogeneous IP/DVB testbed. The second experiment refers to a scalable MPEG-4 FGS stream transmission of two layers, with both DiﬀServ and BM mechanisms deployed to the heterogeneous IP/DVB testbed. The BL packets are encoded using the MPEG4-FGS codec with MPEG2-TM5 rate control at 256 Kbps and the EL one encoded at 256 Kbps. By assigning high priority, Premium Service (EF) to BL, anyone can guarantee proper reception of the BL and without losses.
For the EL, I assign priorities according to the anticipated loss impact of each packet on the end-to-end video quality (considering the loss impact to itself and to dependencies). Each layer has a priority range, and each packet has diﬀerent priority according to its payload. The packets, which contain data of an I-frame are marked with the lowest drop probability (AF11), the packets which contain data of a P-frame are marked with medium drop probability (AF12) and the packets which contain data of a B-frame are marked with high drop probability (AF13). The third experiment refers to a scalable MPEG4 video stream transmission consisting of one BL and two ELs (i.e., EL1 and EL2). The encoding of BL packets remains at 256 Kbps as in the second case, while the encoding of
packets of both ELs is at 256Kbps. For this case, I use EF for transmitting BL, AF11 for transmitting EL1, and Best Eﬀort (BE) for transmitting EL2. For this case both DiﬀServ and BM mechanisms are active as in the second experiment.
Finally, the fourth experiment adopts the setup of the third case, while it applies the prioritized packetization scheme of the second experiment to the packets of the ﬁrst EL (i.e., for this case, I use EF for transmitting BL).
Table 3.3 depicts the experimentation results in terms of PSNR and SSIM video quality metrics for eight diﬀerent video sequences.
It is obvious that for the second experimentation there is a signiﬁcant gain in video quality of 2.3dB in terms of PSNR when compared to the ﬁrst scenario. In some video sequences with many diﬀerences between scenes the video quality gain is more that 3dB.
Furthermore, in the third experiment, it is observed a gain in video quality of
1.2dB, compared to the second experiment. At the fourth scenario, the video quality, in terms of PSNR, remains at the same level.
For the Highway video sequence (consisting of 2000 frames), I measure the packet/frame losses for I-, P-, and B-frames for the four experimental cases with results presented in Table 3.4.
By isolating the losses and the delays to P- and B- frames, can be achieved signiﬁcant gains to video quality. Packet losses, which P-frame content, can aﬀect not only the decoding process of P-frames but also the B-frames. This lead to higher percentages of B-frame losses but it is a signiﬁcant aﬀect to the overall video quality. In the fourth scenario, the user can achieve the same video quality, compared to third scenario, without using only the AF11 traﬃc class of the DiﬀServ. By distributing the traﬃc to all traﬃc classes, achieving the same video quality, in the lowest price, by sending lowest traﬃc to the cost eﬀective AF11 traﬃc. From the network provider perspective, the providers network can use more eﬃcient its bandwidth, by serving more users, at the level of quality they pay.
This chapter presents that the common operation of IP DiﬀServ and DVB BM mechanisms can oﬀer quality gains for media delivery across heterogeneous IP/DVB settings. In this context, this study could constitute a potential vehicle for endto-end QoS provision. Towards this purpose, this chapter presents experimental results of an empirical study of a Linux-based heterogeneous IP/DVB network supporting continuous media applications. The development of new service categories increases the need for a diﬀerentiated, at the network level, treatment of the information packets, based on their diﬀerent association with each type of service. This brings forward the concept of diﬀerentiated QoS provisioning, that is, the possibility to guarantee the most suitable service level for every traﬃc
Several issues remain open and are currently under research. For example, a more eﬃcient mechanism for prioritized packetization of video bit stream is required. Moreover, the distribution of packet priority and a price mechanism according to DS level remains to be examined.
Scalable Video Streaming traﬃc delivery in IP/UMTS networking environment
4.1 Introduction Fixed and wireless/mobile operators are faced with the challenge of a) both creating and delivering attractive IP-based multimedia services quickly responding to fast-changing business and customer demands, and b) evolving their current underlying networking infrastructure to an architecture that can deliver such services in a highly adaptable manner with guaranteed end-to-end Quality of Service (QoS) considering networking and application aspects.
At the same time, the customer side is oﬀered IP connectivity via a wide variety of mobile/wireless access technologies. These technologies include: mobile communication networks, such as GPRS  and UMTS , the family of broadband radio access networks, like IEEE 802.11  and HIPERLAN  and wireless broadcasting technologies, like digital video broadcasting (DVBsatellite and terrestrial) .
IP technology seems to be able to resolve the inter-working amongst the diverse ﬁxed core and wireless/mobile access technologies at the network level.
For an all-IP network, the end-to-end QoS provision concerning the network perspective could be established through the appropriate mapping amongst the QoS traﬃc classes/services supported by the contributing underlying networking technologies. Building on this context, this work involves a DiﬀServ-aware IP core network and a UMTS access network and examines end-to-end QoS issues regarding scalable video streaming traﬃc delivery over such a network.
The Diﬀerentiated Services (DiﬀServ)  approach proposed by IETF supports (based on the DiﬀServ Code Point (DSCP) ﬁeld of the IP header) two diﬀerent services, the Expedited Forwarding (EF) that oﬀers low packet loss and low delay/jitter and the Assured Forwarding (AF), which provides QoS guarantees better than the best-eﬀort service. Diﬀerences amongst AF services imply that a higher QoS AF class will give a better performance (faster delivery, lower loss probability) than a lower AF class .
The QoS provision in UMTS is achieved through the concept of bearers. A bearer is a service providing a particular QoS level between two deﬁned points invoking the appropriate schemes for either the creation of QoS guaranteed circuits, or the enforcement of special QoS treatments for speciﬁc packets. The selection of bearers with the appropriate characteristics constitutes the basis for the UMTS QoS provision. Each UMTS bearer is characterized by a number of quality and performance factors. The most important factor is the bearers Traﬃc Class; four traﬃc classes have been deﬁned in the scope of the UMTS framework (i.e., Conversational, Streaming, Interactive and Background). The appropriate mapping of UMTS traﬃc classes to the aforementioned DiﬀServ service classes could oﬀer a vehicle for the end-to-end QoS provision over a heterogeneous DiﬀServ/UMTS network. In this chapter, I evaluate three diﬀerent mapping approaches of traﬃc classes for the end-to-end QoS provision over a heterogeneous DiﬀServ/UMTS network .
The basic coding scheme for achieving a wide range of spatio-temporal and
quality scalability can be classiﬁed as scalable video codec. For Signal-to-Noise Ratio (SNR) scalability two approaches are the most appropriate for video delivery over heterogeneous networks, the MPEG-4 Fine Grain Scalability (FGS) video coding  and the scalable extension of H.264/MPEG-4 AVC .
The FGS feature of MPEG-4 is a promising scalable video solution to address the problem of guaranteed end-to-end QoS provision concerning the application perspective. According to MPEG-4 FGS, the Base Layer (BL) provides the basic video quality to meet the minimum user bandwidth, while the Enhancement Layer (EL) can be truncated to meet the heterogeneous network characteristics, such as available bandwidth, packet loss, and delay/jitter . In order to support ﬁne-granular SNR scalability, progressive reﬁnement (PR) slices have been introduced in the scalable extension of H.264. A base representation of the input frames of each layer is obtained by transform coding similar to H.264, and the corresponding Network Abstraction Layer (NAL) units (containing motion information and texture data) of the base layer are compatible with the single layer H.264/MPEG-4 AVC. The quality of the base representation can be improved by an additional coding of so-called PR slices. The corresponding NAL units can be arbitrarily truncated in order to support ﬁne granular quality scalability or ﬂexible bit-rate adaptation.
To address the end-to-end QoS problem scalable video streaming traﬃc delivery over a heterogeneous IP/UMTS network, the paper proposes and validates through a number of NS2-based simulation scenarios a architecture that explores the joint use of packet prioritization and scalable video coding together with the appropriate mapping of UMTS traﬃc classes to the DiﬀServ traﬃc classes.
This work extends previous authors works   taking into considerations the case of H.264/MPEG-4 AVC video streaming delivery over IP/UMTS networks.
The second case gives more complete view of the scalable video streaming over
IP/UMTS networking environments for various DiﬀServ/UMTS classes coupling.
The rest of the paper is organized as follows. In Section 4.2, the proposed scalable video coding techniques and prioritization framework for providing QoS guarantees for scalable video streaming traﬃc delivery over a heterogeneous IP/UMTS network is presented. In Section ??, I demonstrate how video-streaming applications can beneﬁt from the use of the proposed architecture. Finally, Section 4.5 draws the conclusions of this work.
4.2 Overview of the Proposed Arrhitecture