Research : Adaptive IPTV

Virtual RTCP — A Case Study of Monitoring and Repair for UDP-based IPTV Systems

Posted on in category adaptive-iptv

IPTV systems have seen widespread deployment, but often lack robust mechanisms for monitoring the quality of experience. This makes it difficult for network operators to ensure that their services match the quality of traditional broadcast TV systems, leading to consumer dissatisfaction. This paper, which will be presented in the Packet Video Workshop in Munich in May 2012 presents a case study of virtual RTCP, a new framework for reception quality monitoring and reporting for UDP-encapsulated MPEG video delivered over IP multicast. We show that this allows incremental deployment of reporting infrastructure, coupled with effective retransmission-based packet loss repair.

Performance Analysis of AL-FEC for RTP-based Streaming Video to Residential Users

Posted on in category adaptive-iptv

Streaming video and IPTV applications, targetting residential customers, are highly sensitive to packet loss, which can disrupt the video quality. This packet loss can be caused by congestion within the residential ISP or elsewhere within the core of the network, or by problems with the edge link. As we have previously studied, performance of residential network links can differ from backbone network performance, so it is important to evaluate the impact of these links of video quality.

To achieve acceptable user experience for these applications, numerous application-layer forward error correction (AL-FEC) schemes have been proposed. In this paper, we evaluate some of the FEC schemes developed as part of the OpenFEC project, using packet loss traces of IPTV-like traffic measured on ADSL and Cable links. We consider the effectiveness of these schemes in correcting the loss patterns present on residential links, explain why performance is different using measured loss traces compared with previous simulations using uniform random packet loss, and give recommendations for the use of FEC in streaming video applications deployed to residential Internet users.

This paper uses a modified version of the OpenFEC eperftool, version 1.2.2. The required modifications can be found in this patch.

New research student: Conor Cahir

Posted on in category adaptive-iptv

Welcome to Conor Cahir who starts work as a research student under my supervision today. Conor's work is sponsored by Cisco, and he'll be studying adaptive HTTP streaming and cache-aware TCP for streaming.

Running ns-3.11 and ns-3.12.1 on FreeBSD-8.2

Posted on in category adaptive-iptv

The following patches allow the ns-3 network simulator, versions 3.11 and 3.12.1 to compile and run on FreeBSD 8.2, with all core tests passing. The patch for ns-3.11 fixes includes and missing declarations, while that for ns-3.12.1 also updates FindSelfDirectory() in src/core/model/system-path.cc to work on FreeBSD.

crtpdumpz v1.0.1

Posted on in category adaptive-iptv

RTP packet traces for IPTV systems are large, but well structured. While they compress reasonably well with a general-purpose compression utility, such as gzip, better performance can be achieved using a compressor that understands the structure of RTP data. The crtpdumpz utility is such a compressor, built using the ideas outlined in RFC 2508.

Measurements of Real-Time Traffic to Residential Users

Posted on in category adaptive-iptv

Little performance data currently exists for streaming high-quality Internet video to residential users. Data on streaming performance provides valuable input to the design of new protocols and applications, for example when evaluating congestion control and error-correction schemes, and for sizing playout buffers in video receivers. This paper presents measurements of streaming real-time UDP traffic to a number of residential users, and discusses the basic characteristics of the data.

The following datasets are referenced from the paper. They contain measurements of CBR RTP traffic, sent from a campus machine to receivers connected to the Internet via ADSL and Cable links. Dataset-A contains only end-to-end measurements, while Dataset-B also includes some end-to-middle measurements obtained using TTL-limited probes, and packet-pair measurements from which the path capacities can be estimated. Dataset-A was collected between June and October 2009, while Dataset-B was collected between April and September 2010.

The raw datasets contain the original trace files logged by our measurement tools, compressed using crtpdumpz, and traceroute output for each trace (with IP addresses and hostnames anonymised). The processed datasets contain end-to-end queueing delay time series (both dataset-A and dataset-B), and end-to-middle delay measurements and packet-pair capacity measurements (dataset-B only).

Workshop on Adaptive Video Streaming over IP Networks

Posted on in category adaptive-iptv

I attended the Workshop on Adaptive Video Streaming over IP Networks held at Cisco Systems in Boxborough, MA, USA, on 12-13 October 2010, to present an update on our adaptive IPTV work. As usual for these workshops, the presentations were an interesting mix of academic research and industry pragmatism. The first day starting with a talk from Sameer Akshabi and Constantine Dovrolis (Georgia Tech) on some experimental evaluation of commercial HTTP streaming video systems, to understand their rate adaptation behaviour. The chunk-based HTTP pull model seems widely accepted, but there is enormous variation in the rate at which chunks are fetched, the size of the playout buffer built-up, and the way in which the applications adapt to changes in the available bandwidth. Some of the systems tested seem to work reasonably well, although they're slow to make use of increases in available bandwidth (possibly intentionally, since variable quality gives a poor user experience); others show various signs of instability. It's clear that this is a rapidly evolving area, with much still to learn.

Zhi Li and Bernd Girod from Stanford outlined some ideas on how to make a more realistic HTTP adaptive streaming simulator using ns-2. Following on from the previous presentation, they clearly described how the two control loops—TCP congestion control operating on the RTT timeframe and application-level adaptation operating over several seconds—interact, highlighting the cause of the stability problems. They also noted an interaction between TCP and client request patterns, which can cause a client that pauses for more than the TCP RTO between chunk request to go into TCP slow start, negating the advantage of keeping the connection open.

The morning concluded with a keynote from David D. Clark (MIT) on “After the triple play: a ten year plan” exploring the economics of the ISP market in the US. Interesting, but it's clear that the ISP marketplace in the US is very different from that in many other parts of the world.

In the afternoon, Ashish Khisti from the University of Toronto presented some theoretical models for adaptive streaming over multihop wireless networks. Then, Dina Katabi (MIT) described their SoftCast system for joint source-channel coding of wireless video for error resilience. This is a very interesting system, leveraging the best ideas from analogue video transmission and modulation to enhance digital video transmission.

The end of the first day, and the morning of the second were filled with discussion sessions and an Overview of Cisco's Video Streaming and CDN Products, and a review of Current Developments in Adaptive Streaming Standards (focussing on IETF, MPEG, and 3GPP work for HTTP streaming).

In the afternoon, we presented our measurements of UDP streaming performance of various DSL and cable access links in the UK and in Finland, including wide area distribution networks to show that loss patterns include bit error-induced losses as well as losses due to congestion and (early) data rate caps. Using a trace-driven simulation of RTP/SSM with feedback and different FEC mechanisms we then showed that, for the observed links and media transmission rates, simple parity FEC mechanisms are capable of reducing the loss rate to acceptable proportions. We concluded with some early thoughts on the applicability of our measurements to TCP-based streaming.

The final presentation, from Mostafa Ammar of Georgia Tech, focussed on hybrid CDN/P2P adaptive live streaming. All-in-all, it was an extremely interesting workshop, covering a lot of ground. Thanks to Dave Oran and Ali Begen for organising!

On the Use of RTP for Monitoring and Fault Isolation in IPTV

Posted on in category adaptive-iptv

Since the first operational IPTV networks have been deployed, service providers and operators have struggled to make their subscribers happy and satisfied with their services. To keep them as customers in the long term, they have been looking for ways to identify impairments to the perceived quality of experience. It is now well understood that this can only be achieved if the service providers have virtual eyes throughout their networks. In this article we provide an overview of the Real- Time Transport Protocol and its application to IPTV. We describe the monitoring and reporting features offered by RTP, and emphasize how they can be used to enhance subscriber QoE.

On the Scalability of RTCP-Based Network Tomography for IPTV Services

Posted on in category adaptive-iptv

Quality of experience (QoE) is an important, and admittedly overloaded, concept for the emerging IPTV services. Service providers are continuously working towards delivering a better TV experience. To this effect, cost-effective and scalable tools are highly desirable for QoE monitoring, diagnostics and reporting. In this paper, we demonstrate that the RTP Control Protocol and its network tomography extensions can satisfy the needs of the providers in collecting and reporting both detailed and summarized information using several numerical examples based on real-life scenarios.

Packet Loss Characteristics of IPTV-like Traffic on Residential Links

Posted on in category adaptive-iptv

Packet loss is one of the principal threats to quality of experience for IPTV systems. However, the packet loss characteristics of the residential access networks which carry IPTV are not widely understood. We present packet level measurements of streaming IPTV-like traffic over four residential access links, and describe the extent and nature of packet loss we encountered. We discuss the likely impact of these losses for IPTV traffic, and outline steps which can ameliorate this.

The packet traces referenced in this paper are available as ccnc-2010-eivt-data.tar.gz (246 MBytes; MD5 2a3e431e4114005c4ae7f1342b24920b). This archive contains the original trace files logged by our measurement tools, compressed using crtpdumpz, and traceroute output for each trace (with IP addresses and hostnames anonymised).

Workshop on Real-Time Video Distribution over IP Networks

Posted on in category adaptive-iptv

We presented initial measurement results at the Workshop of Real-Time Video Distribution over IP Networks held at Cisco Systems in Lawrenceville, GA, USA on 13-14 November 2008.

Cisco Atlanta campus Cisco Atlanta campus

New research student: Martin Ellis

Posted on in category adaptive-iptv

Welcome to Martin Ellis, who starts work as a new research student under my supervision today. Martin is sponsored by Cisco for his first year, and will be working on the adaptive IPTV project.

Adaptive Error Measurement, Concealment, and Repair for IPTV

Posted on in category adaptive-iptv

Internet Protocol Television (IPTV) systems are a rapidly growing converged network service. These systems replace the traditional “set-top box”, receiving broadcast or cable television, with a networked IP-centric device, to which video is streamed live or on demand. The user experience initially follows that of traditional television, but it is expected that additional interactive services will be offered over time, as the potential of the underlying converged network architecture is exploited. Key to this, however, is ensuring the television service provided matches – or exceeds – that of traditional broadcast television. Converged networks suffer from different problems than do pure data networks and dedicated real-time transmission networks, so there is a need for new algorithms and protocol mechanisms to monitor reception quality and diagnose network problems. This project aims to develop such new algorithms and protocols.

Specifically, the key areas we will address are algorithms for correlation and analysis of reception quality reports, methods of efficiently reporting reception quality, and media stream repair techniques. Building on existing standards, we will develop both per- and cross-stream analysis algorithms to locate and diagnose problems with media delivery, design new reception quality report transport, summarization, and aggregation protocols, and use these to improve quality and manageability of IPTV systems.

This project is a collaboration between Aalto University and the University of Glasgow. Funding is provided by Cisco Systems.