Email Address
Computer and Systems Architecture
Digital Communications and Networking
Library and Information Science
OS and Networks
Software Engineering
Systems Architecture
Systems and Communications
Research Projects
Organizational Units
Faculty Member
Research Interests

Search Results

Now showing 1 - 10 of 70
  • Publication
    Revision of QoS Guarantees at the Application/Network Interface
    (1993-03-01) Nahrstedt, Klara; Smith, Jonathan M
    Connection management based on Quality of Service (QoS) offers opportunities for better resource allocation in networks providing service classes. "Negotiation" describes the process of cooperatively configuring application and network resources for an application's use. Complex and long-running applications can reduce the inefficiencies of static allocations by splitting resource use into "eras" bounded by renegotiation of QoS parameters. Renegotiation can be driven by either the application or the network in order to best match application and network dynamics. A key element in this process is a translation between differing perspectives on QoS maintained by applications and network service provision. We model translation with an entity called a "broker".
  • Publication
    Traffic Characteristics of a Distributed Memory System
    (1993-02-01) Smith, Jonathan M; Farber, David J
    We believe that many distributed computing systems of the future will use distributed shared memory as a technique for interprocess communication. Thus, traffic generated by memory requests will be a major component of the traffic for any networks which connect nodes in such a system. In this paper, we study memory reference strings gathered with a tracing program we devised. We study several models. First, we look at raw reference data, as would be seen if the network were a backplane. Second, we examine references in units of "blocks", first using a one-block cache model and then with an infinite cache. Finally, we study the effect of predictive prepaging of these "blocks" on the traffic. We provide a novel representation of memory reference data which can be used to calculate interarrival distributions directly. Integrating communication with computation can be used to control both traffic and performance.
  • Publication
    The Price of Safety in an Active Network
    (1999) Alexander, D. Scott; Anagnostakis, Kostas G.; Arbaugh, William A; Keromytis, Angelos D; Smith, Jonathan M
    Lack of security is a major threat to "Active Networking," as programmability creates numerous opportunities for mischief. The point at which programmability is exposed, e.g., through the loading of code into network elements, must therefore be carefully crafted to ensure security. This paper makes two contributions. First, it describes the implementation of a solution, the Secure Active Network Environment (SANE), which is intended to operate on an active network router. The SANE architecture provides a secure bootstrap process, which includes cryptographic certificate exchange and results in execution of a module loader for introducing new code, as well as a packet execution environment. SANE thus permits a direct comparison of security implications of active packets (such as "capsules") with active extensions (used for "flows" of packets). The second contribution of the paper is a performance study using a combination of execution traces and end-to-end throughput measurements. The example code performs an "active ping" and allows us to break down costs into categories such as authentication. In our SANE implementation on 533 Mhz Alpha PCs, securing active packets effectively increases the time required to process a packet by a third. This result implies that the majority of packets must remain unauthenticated in high performance active networking solutions. We discuss some solutions which preserve security.
  • Publication
    Design, Implementation and Experiences of the OMEGA End-Point Architecture
    (1995) Nahrstedt, Klara; Smith, Jonathan M
    New cell-switched network technologies and multimedia peripherals enable distributed applications with strict real-time requirements such as remote control with feedback. Time-bounded network communications services are necessary, but not sufficient, to meet application-to-application real-time requirements. Real-time communication must be coupled with real-time computing support at the network end-points. An end-point architecture for the computation/communications coupling must be flexible and robust to support a diversity of applications. The OMEGA architecture, when coupled with cell-switched networks (or others which can make bandwidth and delay guarantees), can approximate the behavior of dedicated microcontrollers connected by dedicated circuits in support of an application. The essence of the OMEGA architecture is resource reservation and management within the set of multimedia endpoints. Communications is preceded by a call set-up period where requirements, expressed in terms of Quality of Service (QoS) parameters, are negotiated, and guarantees are made at several logical levels, such as between applications and the network subsystem, applications and the operating system, and the network subsystem and the operating system. This establishes customized connections and allocation of resources appropriate to the application requirements and OS/network capabilities. To facilitate this resource management process, a new paradigm called the 'QoS Brokerage' is used. This paradigm requires new services and protocols across all layers of the protocol stack (i.e., the higher layers of B-ISDN), as well as re-architecting the application/network interface. A prototype of OMEGA has been implemented and tested with a master/slave telerobotics application using a dedicated 155 Mbps ATM LAN. This application employs media with highly diverse QoS requirements and therefore provides a good platform for testing how closely one can approximate a dedicated circuit and controller with workstation hosts and cell-switching. Experience with this implementation has helped to identify new challenges to extending these techniques to a larger domain of applications and systems, and raises several new research questions.
  • Publication
    An Overview of the AURORA Gigabit Testbed
    (1993-02-01) Clark, David D; Davie, Bruce S; Farber, David J.; Gopal, Inder S; Kadaba, Bharath K; Sincoskie, W. David; Smith, Jonathan M; Tennenhouse, David L.
    AURORA is one of five U.S. testbeds charged with exploring applications of, and technologies necessary for, networks operating at gigabit per second or higher bandwidths. AURORA is also an experiment in collaboration, where government support (through the Corporation for National Research Initiatives, which is in turn funded by DARPA and the NSF) has spurred interaction among centers of excellence in industry, academia, and government. The emphasis of the AURORA testbed, distinct from the other four testbeds, is research into the supporting technologies for gigabit networking. Our targets include new software architectures, network abstractions, hardware technologies, and applications. This paper provides an overview of the goals and methodologies employed in AURORA, and reports preliminary results from our first year of research.
  • Publication
    SwitchWare: Accelerating Network Evolution (White Paper)
    (1996) Smith, Jonathan M; Farber, David J; Gunter, Carl A; Nettles, Scott M; Feldmeier, D. C; Sincoskie, W. David
    We propose the development of a set of software technologies ("SwitchWare") which will enable rapid development and deployment of new network services. The key insight is that by making the basic network service selectable on a per user (or even per packet) basis, the need for formal standardization is eliminated. Additionally, by making the basic network service programmable, the deployment times, today constrained by capital funding limitations, are tremendously reduced (to the order of software distribution times). Finally, by constructing an advanced, robust programming environment, even the service development time can be reduced. A SwitchWare switch consists of input and output ports controlled by a software-programmable element; programs are contained in sequences of messages sent to the SwitchWare switch's input ports, which interpret the messages as programs. We call these "Switchlets". This accelerates the pace of network evolution, as evolving user needs can be immediately reflected in the network infrastructure. Immediate reconfigurability enhances the adaptability of the network infrastructure in the face of unexpected situations. We call a network built from SwitchWare switches an active network.
  • Publication
    Moving Targets: Geographically Routed Human Movement Networks
    (2010-03-01) Aviv, Adam J; Sherr, Micah; Blaze, Matt; Smith, Jonathan M
    We introduce a new communication paradigm, Human-to-human Mobile Ad hoc Networking (HuManet), that exploits smartphone capabilities and human behavior to create decentralized networks for smartphone-to-smartphone message delivery. HuManets support stealth command-and-control messaging for mobile BotNets, covert channels in the presence of an observer who monitors all cellular communication, and distributed protocols for querying the state or content of targeted mobile devices. In this paper, we introduce techniques for constructing HumaNets and describe protocols for efficiently routing and addressing messages. In contrast to flooding or broadcast schemes that saturate the network and aggressively consume phone resources (e.g., batteries), our protocols exploit human mobility patterns to significantly increase communication efficiency while limiting the exposure of HuManets to mobile service providers. Our techniques leverage properties of smartphones – in particular, their highly synchronized clocks and ability to discern location information – to construct location profiles for each device. HuManets’ fully-distributed and heuristic-based routing protocols route messages towards phones with location profiles that are similar to those of the intended receiver, enabling efficient message delivery with limited effects to end-to-end latency.
  • Publication
    Rethinking Mobile Telephony With the IMP
    (2011-01-01) DeYoung, M.; Henke, N.; Wai, G.; Smith, Jonathan M
    The recent widespread deployment of wireless LAN technology raises the question of how a mobile telephony system might instead be architected to use wireless LAN access points and the Internet to achieve similar services. In this paper, we examine an end-to-end architecture for mobile telephony, with a strong focus on endpoint issues. We have designed, implemented, and have experience using devices we call Internet Mobile Phones or IMPs. The IMP system provides encrypted wireless voice communication over 802.11B LANs. IMPs run Linux on a lightweight single-board computer running customized voice over IP software; data is encrypted with 128-bit Blowfish. The paper reports on our design decisions and the resulting implementation of the IMPs, with sufficient detail to reproduce the devices. We report our experiences with using them for several months in a laboratory environment, and close with proposals for future experiments to investigate scale and extensibility.
  • Publication
    Gigabit Networks
    (1996) Smith, Jonathan M
    This chapter summarizes what we have learned in the past decade of research into extremely high throughput networks. Such networks are colloquially referred to as "Gigabit Networks" in reference to the billion bit per second throughput regime they now operate in. The engineering challenges are in the integration of fast transmission systems and high-performance engineering workstations.
  • Publication
    Operating Systems Support for End-to-End Gbps Networking
    (1993-03-01) Smith, Jonathan M; Traw, C. Brendan S
    This paper argues that workstation host interfaces and operating systems are a crucial element in achieving end-to-end Gbps bandwidths for applications in distributed environments. We describe several host interface architectures, discuss the interaction between the interface and host operating system, and report on an ATM host interface built at the University of Pennsylvania. Concurrently designing a host interface and software support allows careful balancing of hardware and software functions. Key ideas include use of buffer management techniques to reduce copying and scheduling data transfers using clocked interrupts. Clocked interrupts also aid with bandwidth allocation. Our interface can deliver a sustained 130 Mbps bandwidth to applications, roughly OC-3c link speed. Ninety-three percent of the host hardware subsystem throughput is delivered to the application with a small measured impact on other applications processing.