QoS (Quality of Service)

The Internet provides a "best effort" service — no guarantees on performance for a given end-to-end communication flow — no Quality of Service (QoS) control. However, it is useful to implement some monitoring and control of resources within an autonomous or constrained network environment, or to allow adaption to changing network conditions.

This work has considered measuring network performance and using the data to either provide some control of network resources, or to inform and enable applications and end-system to adapt their operation and behaviour to make best use of available network capacity.

Y. Huang, S. N. Bhatti. Resilient State Management in Large Scale Networks. IWQoS 2005 - 13th Intl. Workshop on Quality of Service. Passau, Germany. June 2005.
| PDF | .bib | 10.1007/11499169_31 | abstract This paper describes, briefly, ongoing research on resource reservation state management, including research motivations and initial design. Y. Huang, S. N. Bhatti. Scaleable Signaling Underlay for Overlay Networks. ICON 2004 - 12th IEEE Intl. Conf. on Networking. Singapore. Nov 2004.
| PDF | .bib | 10.1109/ICON.2004.1409187 | abstract This paper presents the design of a scalable decentralized signaling underlay infrastructure, which features a DHT based management information storage and query-based state lookup mechanism. The signaling underlay is aimed to apply a decentralized "peer-to- peer" style searching and discovering engine into the management and control plane of the overlay network, including grid networks and p2p applications, to facilitate deployment of QoS service. Y. Huang, S. N. Bhatti. Decentralized Resilient Grid Resource Management Overlay Networks. SCC 2004 - IEEE Intl. Conf. Services Computing. Shanghai, China. Sep 2004.
| PDF | .bib | 10.1109/SCC.2004.1358027 | abstract Recently, grid and peer-to-peer overlays have attracted attention from research communities as well as industry. However, the overwhelming majority of industrial, commercial and academic activities are geared around applications, servers, and middleware. There have been some efforts in the management of the underlying networks (the connectivity for the overlays), but far from enough. This paper proposes a resilient, fault-adaptive, overlay resource management framework, with lightweight state management infrastructure, taken from some ongoing work by the authors in this area. This research aims to solve the problems of: (a) managing distributed network resources in a decentralized way; (b) providing resilient QoS for highly dynamic networks. F. Saka N. Pezzi A. di Donato J. Orellana P. Clarke Y-T Li S. Dallison R. Hughes-Jones S. Bhatti R. Smith R. Tasker. Enabling advanced high performance networks and end-systems for Grid applications. AHM 2004 - e-Science All Hands Meeting. Nottingham, UK. Aug 2004.
| PDF | .bib | abstract The MB-NG project brings together users, industry, equipment providers and e-science applications. The project aims are: to construct a high-performance leading edge quality of service (QoS) network; to demonstrate end-to-end managed bandwidth services in a multi-domain environment and to investigate high performance data transport mechanisms for Grid data transfer across heterogeneous networks. We report on the major successes in the area of QoS and managed bandwidth, the achievements in the area of end-hosts and the benefits to applications. M. Rio, A. di Donato, F. Saka, N. Pezzi, R. Smith, S. N. Bhatti, P. Clarke. Quality of Service Networking for High Performance Grid Applications. Journal of Grid Computing, vol. 1, no. 4, pages 329-343. Dec 2003.
| PDF | .bib | 10.1023/B:GRID.0000037551.92756.4e | abstract This paper reports on different efforts to provide quality of service (QoS) Networking to Grid applications done in the context of the MB-NG, GRS and DataTAG EU projects. These are leading edge network research projects involving more that 50 researchers in the UK, Europe and North America, concerned with the development and testing of protocols and standards for the next generation of high speed networks. We have implemented and tested the Differentiated Services Architecture (DiffServ) in a multi-domain, 2.5 Gbits/s network (the first such deployment) defining appropriate Service Level Agreements (SLAs) to be used between administrative domains to guarantee end-to-end Quality of Service. We characterised several hardware implementations of DiffServ and concluded on their appropriateness for several network scenarios. Since current and future Grid applications will have to use modified mechanisms of congestion control we have evaluated old and new TCP implementations over a Differentiated Services Networks. These quality of service tests have also included innovative MPLS (Multi- Protocol Label Switching) experiments to establish guaranteed bandwidth connections to Grid applications in a fast and efficient way. We have also developed a software based bandwidth broker architecture for Grids based on IETF standards which allows applications to transparently request dynamic and advanced reservations and implemented it in a real experimental network. We finally report on experiences delivering Quality of Service networking to high performance applications like Particle Physics data transfer and High Performance Computation. This includes quantitative results on the performance improvements that QoS brought to real data transfers in the context of High Performance Computing. T. Henderson, S. N. Bhatti. Networked games - a QoS-sensitive application for QoS insensitive users?. RIPQOS 2003 - ACM SIGCOMM Workshop on Revisiting IP QoS: What have we learned, why do we care?. Karlsruhe, Germany. Aug 2003.
| PDF | .bib | 10.1145/944592.944601 | abstract Research into providing different levels of network Quality of Service (QoS) often assumes that there is a large market for QoS- sensitive applications that will be fulfilled once QoS-enabled networks have been deployed. Multiplayer networked games are an example of such an application that requires QoS, and hence will only become popular if QoS is made widely available. The prima facie evidence, however, is that games are already popular, in spite of the existing QoS-free best-effort Internet.Networked games may have become popular despite the lack of QoS because players "make do" with what is available to them. Such popularity is a double-edged sword. It may mean that there is a demand, as yet unfulfilled, from game players for QoS-enabled networks. On the other hand, it may mean that players have become accustomed to playing games without QoS, and therefore might be less willing to pay for higher QoS when it does eventually become available.In this paper we present the results of a short experiment to examine the QoS tolerances of game players. We use a set of popular First Person Shooter (FPS) game servers that are publicly available to Internet users at large. By systematically altering the network latency to the servers, we attempt to study whether degraded QoS (in the form of higher network delay) affects a user's decision to participate in the game.We find that increased network delay has an effect on a user's decision to join a game server. It appears, however, that there is no significant difference in the number of players who leave the game as a result of increased delay. We speculate that this may be due to a user's enjoyment exceeding their QoS-sensitivity, and discuss the implications of our findings with respect to providing and charging for QoS. S. N. Bhatti, S.-A. Sorensen, P. Clarke, J. Crowcroft. Network QoS for Grid Systems. IJHPCA - Intl. Journal of High Performance Computing Applications, vol. 17, no. 3, pages 219-236. Aug 2003.
| PDF | .bib | 10.1177/1094342003173009 | abstract Grid users may wish to have fine-grained control of quality of service (QoS) guarantees in a network in order to allow timely data transfer in a distributed application environment. We present a discussion of the issues and problems involved, with some critical analysis. We propose possible solutions by making reference to and analysing existing work. Also, we describe the mechanisms being proposed as part of a work-in-progress (being conducted by the authors) that uses a peer-to-peer approach to micro-manage network capacity allocations at the edge of the network, at end-sites, in a multi-domain scenario. Scheduling controllers at the end-sites are employed, which are subject to local administrative controls and have flexibility in resource allocation based on user requests for network capacity. We highlight the issues in scaling such systems to large numbers of users and the issues concerning the interfaces available to applications and end-users for accessing such services. S. N. Bhatti, P. Clarke. Larging it for the Grid: Big Networking for Big Science. AHM 2002 - e-Science All Hands Meeting. Sheffield, UK. Sep 2002.
| PDF | .bib | abstract Key words to describe the work: quality of service (QoS), network control, resource management, high-speed networking. Key Results: in progress – national and international high-speed, manageable, flexible, QoS-controlled network connectivity. How does the work advance the state-of-the-art?: to provide control of the network for users from their desktops whilst still allowing network administrators to operate the network to meet users’ needs in a scaleable with respect to user demand. Motivation (problems addressed): the need to enable connectivity for distributed applications, with requirements to access large amounts of data remotely (e.g. bio-informatics, high-energy physics, radio astronomy, socio-economic data-mining) or with requirements for real-time interaction (e.g. distributed simulation, distributed control, real- time remote visualisation, high-quality video and audio for remote language teaching or conferencing. T. Henderson, J. Crowcroft, S. N. Bhatti. Congestion pricing: Paying your way in communication networks. IEEE Internet Computing, vol. 5, no. 5, pages 85-89. Sep/Oct 2001.
| PDF | .bib | 10.1109/4236.957899 | abstract Network congestion is a fundamental problem facing Internet users today. A network where users are selfish, and thus reluctant to defer to other users, may result in the famous "tragedy of the commons", where, in the absence of controls, a shared resource is overconsumed by individuals who consider only their personal costs and not the cost to society as a whole. In terms of the Internet, the "tragedy" could be viewed as congestive collapse, resulting from overconsumption of the shared network resource. It is important to distinguish congestion pricing from other forms of network pricing. Charging network users for the congestion they cause can lead to more efficient network utilization by forcing them to take social costs into account. In a congestion-pricing framework, the congestion charge would replace usage and QoS charges. Users would pay their ISPs a subscription charge to cover fixed costs and a congestion charge only when appropriate. This pricing scheme is feasible because, in the absence of congestion, the marginal cost of a network link is practically zero. Congestion pricing can also benefit network operators. By indicating the level of congestion and the user tolerance of it in their networks, congestion pricing can inform operators about when to re-provision and increase network capacity P. Gevros, P. Kirstein, J. Crowcorft, S. N. Bhatti. Congestion Control Mechanisms and the Best Effort Service Model. IEEE Network, vol. 15, no. 3, pages 16-26. May/Jun 2001.
| PDF | .bib | 10.1109/65.923937 | abstract In the last few years there has been considerable research toward extending the Internet architecture to provide quality of service guarantees for the emerging real-time multimedia applications. QoS provision is a rather controversial endeavour. At one end of the spectrum there were proposals for reservations and per-flow state in the routers. These models did not flourish due to the network's heterogeneity the complexity of the mechanisms involved, and scalability problems. At the other end, proposals advocating that an over-provisioned best effort network will solve all the problems are not quite convincing either. The authors believe that more control is clearly needed for protecting best effort service. An important requirement is to prevent congestion collapse, keep congestion levels low, and guarantee fairness. Appropriate control structures in a best effort service network could even be used for introducing differentiation. This could be achieved without sacrificing the best effort nature of the Internet or stressing its architecture beyond its limits and original design principles. We revisit the best effort service model and the problem of congestion while focusing on the importance of cooperative resource sharing to the Internet's success, and review the congestion control principles and mechanisms which facilitate Internet resource sharing S. N. Bhatti, J. Crowcroft. QoS Sensitive Flows: Issues in IP Packet Handling. IEEE Internet Computing, vol. 4, no. 4, pages 48-57. Jul/Aug 2000.
| PDF | .bib | 10.1109/4236.865087 | abstract IP-based networks were never designed for real time traffic, yet QoS support in such networks is needed to accommodate both global use and the more demanding applications now emerging. Changes in packet handling, in particular, will help provide QoS support in IP networks. The article focuses on the issues and principles concerning router modification for IP packet handling. S. N. Bhatti, G. Knight. Enabling QoS adaptation decisions for Internet applications. Computer Networks, vol. 31, no. 7, pages 669-692. Apr 1999.
| PDF | .bib | 10.1016/S0169-7552(98)00294-3 | abstract We present a network model that allows processing of QoS (quality of service) information about media flows to enable applications to make adaptation decisions. Our model is based around a multi-dimensional spatial representation that allows QoS information describing the flow constructions and QoS parameters – flow-states – to interact with a representation of the network QoS. The model produces reports about the compatibility between the flow-states and the network QoS, indicating which flow-states the network can currently support. The simple nature of the reports allows the application to make decisions, dynamically, on which flow-state it should use. The model is relatively lightweight and scaleable. We demonstrate the use of the model by simulation of a dynamically adaptive audio tool. Our work is ongoing. S. N. Bhatti, G. Knight. QoS Assurance vs. Dynamic Adaptability for Applications. NOSSDAV 1998 - 8th Intl. Workshop on Network and Operating System Support for Digital Audio and Video. Computer Laboratory, University of Cambridge, UK. Jul 1998.
| PDF | .bib | abstract Enabling adaptability to network QoS (quality of service) is seen as a key feature for future applications. One way to approach adaptability is to build it into the net- work and allow applications to signal their requirements to the network. This means that resource reservation mechanisms must be available end-to-end, which is not always the case, especially on the Internet. Also, user preferences affect how the application is used. So the application must be dynamically adaptive, taking into account the application’s capabilities, the available network QoS and the user preferences. We suggest that it is possible to build practicable QoS summaries that capture all these inputs and allow dynamic adaptability. S. N. Bhatti, G. Knight. Notes on a QoS information model for making adaptation decisions. HIPPARCH 1998 - 4th Intl. Workshop on High Performance Protocol Architectures. Dept. of Computer Science, UCL, London, UK. Jun 1998.
| PDF | .bib | abstract We present a network model that allows processing of QoS (quality of service) information about media flows to enable applications to make adaptation decisions. Our model is based around a multi-dimensional spatial representation that allows QoS information describing the flow constructions and QoS parameters – flow-states – to interact with a representation of the network QoS. The model produces reports about the compatibility between the flow-states and the network QoS, indicating which flow-states the network can currently support. The simple nature of the reports allows the application to make decisions, dynamically, on which flow-state it should use. The model is relatively lightweight and scaleable. We demonstrate the use of the model by simulation of a dynamically adaptive audio tool. Our work is ongoing.