Partho Pratim Mishra

A flexible model for resource management in virtual private networks

As IP technologies providing both tremendous capacity and the ability to establish dynamic secure associations between endpoints emerge, Virtual Private Networks (VPNs) are going through dramatic growth. The number of endpoints per VPN is growing and the communication pattern between endpoints is becoming increasingly hard to forecast. Consequently, users are demanding dependable, dynamic connectivity between endpoints, with the network expected to accommodate any traffic matrix, as long as the traffic to the endpoints does not overwhelm the rates of the respective ingress and egress links. We propose a new service interface, termed a hose, to provide the appropriate performance abstraction. A hose is characterized by the aggregate traffic to and from one endpoint in the VPN to the set of other endpoints in the VPN, and by an associated performance guarantee.Hoses provide important advantages to a VPN customer: (i) flexibility to send traffic to a set of endpoints without having to specify the detailed traffic matrix, and (ii) reduction in the size of access links through multiplexing gains obtained from the natural aggregation of the flows between endpoints. As compared with the conventional point to point (or customer-pipe) model for managing QoS, hoses provide reduction in the state information a customer must maintain. On the other hand, hoses would appear to increase the complexity of the already difficult problem of resource management to support QoS. To manage network resources in the face of this increased uncertainty, we consider both conventional statistical multiplexing techniques, and a new resizing technique based on online measurements.To study these performance issues, we run trace driven simulations, using traffic derived from AT&Ts voice network, and from a large corporate data network. From the customers perspective, we find that aggregation of traffic at the hose level provides significant multiplexing gains. From the providers perspective, we find that the statistical multiplexing and resizing techniques deal effectively with uncertainties about the traffic, providing significant gains over the conventional alternative of a mesh of statically sized customer-pipes between endpoints.

SWAN: a mobile multimedia wireless network

The SWAN (Seamless Wireless ATM Network) is an experimental indoor wireless network that instigates the combination of wireless access with multimedia networked computing in an indoor setting. It is based on room-sized pico-cells and mobile multimedia endpoints. It enables users carrying multimedia endpoints, such as personal digital assistants (PDAs), laptops, and portable multimedia terminals, to seamlessly roam while accessing multimedia data resident in a backbone wired network. The network model of SWAN consists of base stations connected by a wired asynchronous transfer mode (ATM) backbone network, and wireless ATM last hops to the mobile hosts. SWAN is one of the first systems to realize the concept of a wireless and mobile ATM network. Mobile hosts as well as base stations are embedded with custom-designed ATM adapter cards called FAWN (Flexible Adapter for Wireless Networking). FAWN uses off-the-shelf 2.4 GHz industrial, scientific, and medical (ISM) band radios. After giving an overview of the SWAN network model, and discussing the challenges in making ATM wireless and mobile, the article describes the first phase implementation of SWAN hardware and software. This initial implementation provides connectivity over the wireless last hop. We have investigated both native-mode end-to-end ATM communication across the wired ATM backbone and wireless ATM links, and transmission control protocol (TCP) and user datagram protocol (UDP) communication using Internet protocol (IP) over wireless ATM in the wireless link with IP forwarding and segmentation and reassemble modules at the base stations.

A comparison of bandwidth smoothing techniques for the transmission of prerecorded compressed video

The transfer of prerecorded, compressed video requires multimedia services to support large fluctuations in bandwidth requirements on multiple time scales. Bandwidth smoothing techniques can reduce the burstiness of a variable-bit-rate stream by prefetching data at a series of fixed rates, simplifying the allocation of resources in video servers and the communication network. Given a fixed client-side prefetch buffer several bandwidth smoothing algorithms have been introduced that are provably optimal under certain constraints. This paper presents a collection of metrics for comparing these smoothing algorithms and evaluating their cost-performance trade-offs. Due to the scarcity of available trace data, we have constructed a video capture testbed and generated a collection of twenty full-length, motion-JPEG encoded video clips. Using these video traces and a range of client buffer sizes, we investigate the interplay between the performance metrics through simulation experiments. The results highlight the unique strengths and weaknesses of each algorithm.

TOPS: an architecture for telephony over packet networks

Packet telephony is of increasing interest in both the telecommunications and Internet communities. The emergence of packet telephony will create new services, and presents an opportunity to rethink how conventional telephony services are implemented. In this paper, we present an architecture for telephony over packet networks (TOPS). TOPS allows users to move between terminals or to use mobile terminals while being reachable by the same name. TOPS users can have multiple terminals and control how calls are routed to them. TOPS allows for terminals with a range of capabilities such as support for video, whiteboard, and other media with a variety of coding formats. TOPS retains the necessary information on terminal capabilities to determine the appropriate type of communication to be established with the remote terminal. The architecture assumes that the underlying network supports the establishment of end-to-end connectivity between terminals, with an appropriate quality of service. The components of TOPS are a directory service, an application layer signaling protocol, and a logical channel abstraction for communication between end-systems. The directory service maps a users name to a set of terminals where the user may be reached. A user can control the translation operation by specifying profiles that customize how his name is mapped to a set of terminals where he can be reached. Terminal capabilities are also stored in the directory service. The application layer signaling protocol establishes and maintains call state between communicating terminals. The logical channel abstraction provides a shared end-to-end context for a calls constituent media and control streams, while isolating the applications from the details of the network transport mechanisms. In addition to supporting simple point-to-point calls, the architecture supports both centralized and decentralized conferencing. We also introduce a simple encapsulation format for voice.

Network architecture for mobile and wireless ATM

There is an emergent interest in providing mobile users with ubiquitous wireless access to multimedia information. In this paper we present a network architecture and protocols to achieve this goal and describe their implementation in a prototype network called SWAN. Our network model assumes end to end ATM connectivity. Thus, the key question we address is how best to enhance ATM to support host mobility and wireless access.

User-perceived performance of Web-browsing and interactive data in HFC cable access networks

Using analysis and simulation, we study the performance of Web-browsing and interactive data applications as perceived by the users of shared packet access networks. We use end-to-end page delays to characterise the user experience, and introduce a measure called equivalent circuit rate (ECR). The ECR for a user of a shared packet access network is the dedicated access circuit bit rate required to achieve the same user experience. Our analysis uses a simple closed queueing network model. The theoretical results agree well with simulations of TCP-based Web browsing workloads sharing an HFC cable data channel. We show that the mean page delay, ECR, and the number of users that can be supported are described using simple parameters that scale very well for different user and network rates. Our work also provides an effective comparison of shared and dedicated access at the application level.

On quality of service in mobile wireless networks

This paper addresses the choices that need to be made in the design of a quality of service (QoS) framework for mobile and wireless networks. An uncontrollably time varying environment, arising due to physical processes such as fading as well as systematic processes such as handoffs, is an inalienable characteristic of these networks. Therefore, the laissez faire model of purely reactive applications adapting to a black-box network does not work well for multimedia applications in the presence of wireless and mobility. However, as we empirically show, the other extreme of applying rigid quality of service models borrowed from the wired networks, with correspondingly complex network mechanisms, is also a bad idea because of the often adverse interaction between higher layer protocols with mechanisms such as rerouting and MAC control. We advocate that mobile network and wireless link layer mechanisms should be kept simple and stupid, though not oblivious of QoS, and that applications should be made more sophisticated with QoS renegotiation and adaptation capabilities. The paper describes the necessary support required in various layers of mobile wireless networks, and advocates specific solutions based on empirical validations.

Performance of a shared packet wireless network with interactive data users

This paper studies the user-perceived performance of a shared packet wireless network for interactive data applications such as Web-browsing. We have defined a new measure: the Equivalent Circuit Rate (ECR) for a user in a shared access network is the dedicated access circuit rate that would be required by the user in order to have an equivalent user experience. The ECR measure is intuitive, useful and robust. We present a simple analytical model based on a closed queueing network with a finite population of interactive data users whose traffic can be modeled as ON/OFF traffic with feedback. The analytical results are consistent with our detailed simulations, where we study TCP-based Web traffic simulations for an EDGE (Enhanced Data Rates for GSM Evolution) TDMA system using realistic workload models. Our study incorporates constant bit-rate channels, variable bit-rate channels, as well as mixed user groups with different channel conditions. We show that a shared wireless network is well characterized by simple parameters that depend only on mean statistics, for a broad range of scenarios. ECR is well approximated by R×(1−U), where U is the utilization and R is the effective shared channel rate; this result holds even if different user groups have different rates or the channel rates are variable. These results will be useful in network dimensioning, admission control, and for defining quality of service targets for different user groups.

DOSA: an architecture for providing a robust IP telephony service

An increasing number of communication services are moving to an IP-based infrastructure. Packet telephony is probably the first important real-time service that must be supported well over an IP network. The use of IP presents a tremendous opportunity for service providers to exploit endpoint intelligence to offer creative new services going far beyond the current telephony service model. However, to support telephony, signaling protocols are needed that allow the service provider to offer network-layer service differentiation and, at the same time, to control access to both the enhanced network-layer quality of service as well as other services. This paper describes the distributed open signaling architecture (DOSA), which incorporates protocols that meet these needs. A key contribution of our work is a recognition of the need for coordination between call signaling, which controls access to telephony-specific services, and resource management, which controls access to network-layer resources.

User-perceived performance of web-browsing and interactive data applications in TDMA packet wireless networks

Using analysis and simulation, we study the performance of Web-browsing and interactive data applications as perceived by the users of a shared packet access network. To characterise the user experience, we use end-to-end page delays and introduce a robust and useful measure called Equivalent Circuit Rate (ECR). The ECR for a user of a shared packet access network is the dedicated access circuit bit rate required to achieve the same user experience. Our analysis uses a simple closed queueing network model and the theoretical results agree well with simulations of TCP-based Web browsing workloads sharing a third generation EDGE (Enhanced Data Rates for GSM Evolution) cellular wireless data channel. The number of interactive data users that can be supported with acceptable performance is about 80% to 90% of the ratio of the shared channel rate to the average user rate. ECR is closely approximated by effective channel rate*(1-utilization).