Extending the keyword search paradigm to relational data has been an active area of research within the database and IR community during the past decade. Many approaches have been proposed, but despite numerous publications, there remains a severe lack of standardization for the evaluation of proposed search techniques. Lack of standardization has resulted in contradictory results from different evaluations, and the numerous discrepancies muddle what advantages are proffered by different approaches. In this paper, we present the most extensive empirical performance evaluation of relational keyword search techniques to appear to date in the literature. Our results indicate that many existing search techniques do not provide acceptable performance for realistic retrieval tasks. In particular, memory consumption precludes many search techniques from scaling beyond small data sets with tens of thousands of vertices. We also explore the relationship between execution time and factors varied in previous evaluations; our analysis indicates that most of these factors have relatively little impact on performance. In summary, our work confirms previous claims regarding the unacceptable performance of these search techniques and underscores the need for standardization in evaluations--standardization exemplified by the IR community.

Emergency resources are often insufficient to satisfy fully the demands for professional help and supplies after a public disaster. Furthermore, in a mass casualty situation, the emphasis shifts from ensuring the best possible outcome for each individual patient to ensuring the best possible outcome for the greatest number of patients. Historically, various manual and electronic medical triage systems have been used both under civil and military conditions to determine the order and priority of emergency treatment, transport, and best possible destination for the patients [3][4][5][6]. Unfortunately, none of those solutions has proven flexible, accurate, scalable or unobtrusive enough to meet the public's expectations [1]. We demonstrate a system for real-time patient assessment which uses mobile electronic triaging accomplished via crowdsourced and sensor-detected information. With the use of our system, emergency management professionals receive most of the information they need for preparing themselves to perform a timely and accurate treatment of their patients even before dispatching a response team to the event. During our demonstration, we will show how our system behaves with different combinations of information inputs and compare its resulting outputs with evaluations done by medical experts. The public will be given the chance to participate in real-time demos by posing as victims and providing self-reported information about their health.

As the population ages and the risk of chronic disease increases, the cost of healthcare will rise. Technology for mobile telemetry could reduce cost and improve the efficiency of treatment. In order to achieve these goals, we first need to overcome several technical challenges. We present a Mobile Biotelemetric System (MBS) that addresses these remote medical monitoring challenges. MBS comprises a custom low-power sensor node that accurately collects and analyzes electrocardiogram (ECG) data, a client service with a multifaceted policy engine that evaluates the data, and a web portal interface for visualizing ECG data streams. MBS differs from other remote monitoring systems primarily in the policy engine's ability to provide flexible, robust, and precise system communication end-to-end and to enable tradeoffs in metrics such as power and transmission frequency. We show that, given a representative set of ECG signals, policies can be set to make the operation of the hardware and software resilient against transient ECG conditions for both security and monitoring purposes. We demonstrate that our system adaptively trades off system-level metrics based on a combination of operating conditions and user input, and that our heartbeat detection algorithm performs well for challenging ECG input.

The United States Department of Defense funded a research project at the University of Virginia to increase community resilience after a crisis (e.g., natural disaster, bioterrorism attack). The premise is that the public would be better served if it had access to reliable, near real-time information concerning the "ground truth" (actual observations and pictures) of unfolding events. One way (of many) to collect those reports is through crowdsourcing. By taking advantage of the army of people connected to the internet via both wired and wireless technologies, we can easily gather information that would help a citizen make an informed decision regarding his or her health and welfare (e.g., shelter-in-place vs. flight). The widespread use of smartphones allows people to post geolocated reports and pictures with ease. Disaster management teams would likewise benefit from the input of more observers. On the other hand, the trustworthiness of these reports may be uncertain. For instance, a terrorist group might use crowdsourcing to redirect the community's emergency responses to the wrong place. We first explore how a crowdsourcing system might be used to enhance disaster management, and then we investigate three ways to enhance the trust that the citizen places in such a system.

Keyword search within databases has become a hot topic within the research community as databases store increasing amounts of information. Users require an effective method to retrieve information from these databases without learning complex query languages (viz. SQL). Despite the recent research interest, performance and search effectiveness have not received equal attention, and scoring functions in particular have become increasingly complex while providing only modest benefits with regards to the quality of search results. An analysis of the factors appearing in existing scoring functions suggests that some factors previously deemed critical to search effectiveness are at best loosely correlated with relevance. We consider a number of these different scoring factors and use machine learning to create a new scoring function that provides significantly better results than existing approaches. We simplify our scoring function by systematically removing the factors with the lowest weight and show that this version still outperforms the previous state-of-the-art in this area.

An e-commerce sandbox is being developed to supplement a course teaching the technology and business aspects of e-commerce at the University of Virginia. The system consists of a collection of relevant course materials and a sandbox environment for students to perform guided modules and receive useful automated feedback. The primary goal of the sandbox is to allow students to safely practice real-world e-commerce skills while promoting a self-guided learning approach. The system currently provides 22 modules covering topics in PHP and MySql. The administrative interface provides tools to create arbitrary modules so that the sandbox can expand and evolve with the state of the practice. Based upon our first classroom experience, we found that use of the interactive modules increased student code quality while providing a beneficial adjunct to the lecture portion of the class.

As the value of information placed on mobile devices increases, so does the risk that the information will be lost or stolen. In dire scenarios, such as soldiers on the battlefield, there is a tension between accessing critical information quickly and protecting that information from unauthorized viewers. Lightweight body sensors that detect and process physiological information can provide an unconventional means for simultaneously securing data on a mobile device and making pertinent health information available to authorized remote viewers. In this paper we present the design, implementation, and evaluation of our three-tier Secure Mobile Computing (SMC) system. Tier one consists of a physiological sensor (initially an electrocardiograph), microcontroller, and radio (initially Bluetooth) with the form factor of a bandage, collectively termed the "patch." The patch prototype collects and processes electrocardiograph (ECG) data and transmits the processed information over the wireless channel either continuously or periodically. The primary processing functionality, the heartbeat detection algorithm, has an average accuracy of over 99.5%. Tier two is the mobile device (e.g., cell phone, PDA, or laptop). SMC makes the utility of the mobile device dependent upon receipt of the patch's telemetry signal. SMC supports a number of programmable security policies that can either lock (e.g., encrypt) or erase data if the user is incapacitated or the mobile device loses proximity to the patch. Tier three is a web service that allows authorized viewers to view the sensor information remotely. We explore how SMC manages the interfaces between the tiers to implement security policies on the mobile device.

With regard to keyword search systems for structured data, research during the past decade has largely focused on performance. Researchers have validated their work using ad hoc experiments that may not reflect real-world workloads. We illustrate the wide deviation in existing evaluations and present an evaluation framework designed to validate the next decade of research in this field. Our comparison of 9 state-of-the-art keyword search systems contradicts the retrieval effectiveness purported by existing evaluations and reinforces the need for standardized evaluation. Our results also suggest that there remains considerable room for improvement in this field. We found that many techniques cannot scale to even moderately-sized datasets that contain roughly a million tuples. Given that existing databases are considerably larger than this threshold, our results motivate the creation of new algorithms and indexing techniques that scale to meet both current and future workloads.

Website security is essential for successful e-commerce ventures, but the vital "how-to" components of security are often lacking in academic courses. This paper describes our attempt to instill an awareness of security concerns and techniques by having the students develop an Artist eXchange website, a social networking site that permits the posting and sharing of pictures, music, and text, including an end-user rating system. The six-homework set progresses through HTML, JavaScript, PHP, MySQL, file uploads, and security testing. An innovative feature is that each assignment is evaluated via automated testing, which guides the student toward detecting and correcting mistakes, especially with regard to common attack vectors.

This paper examines the requirements of wearable sensing applications and their implications for designing the next generation of body area sensors. We define key metrics for wearable sensors and discuss how body area sensors differ from generic wireless sensors. To explore the system level issues with a wearable node, we show measurements from a wearable electrocardiogram (ECG) sensor prototype. Using heart rate monitoring as an example, we show how ultra low power (ULP) circuit design must be applied to support the stringent energy and/or power demands of long life wearable sensors. Specifically, sub-threshold operation of digital circuits creates opportunities for re-thinking the entire system. We conclude that we can only reach the lower limits of power consumption through cross-hierarchy design of the entire sensor node that leverages ULP digital circuits.

In federated systems, trust management is critical for information sharing and online collaboration. Security tokens provide a way to convey and exchange trust information for security and privacy protection. Although several communication protocols have mechanisms to exchange security tokens, users may still encounter difficulties when they have to handle security tokens across heterogeneous platforms and security domains. Semantic gaps and incompatibilities are major barriers for trust information exchange in federated trust management. This paper introduces a hybrid approach with intermediary- and query-based mechanisms to resolve semantic gaps and incompatibilities for different types of trust information exchanged by security tokens, and then proposes different exchange models for different types of information. This paper also provides a comprehensive framework to exchange security tokens across security domains with suitable approaches and exchange models. The application of this framework in a healthcare environment provides a new method to interoperate trust information for security- and privacy-critical applications.

In federated networks, trust management is critical to information sharing and online collaboration. Security tokens provide a way to convey and exchange trust-related information for security and privacy purposes. Common users encounter difficulties when they have to handle security tokens across heterogeneous domains. Semantic gaps and incompatibilities are major barriers for trust-related information exchange in federated trust management. This paper uses intermediary-based, query-based and hybrid approaches to resolve these issues for different types of information in security tokens, and proposes three exchange models accordingly. This paper also provides a comprehensive framework using Web services to exchange security tokens across security domains with suitable approaches and exchange models.

In federated and pervasive networks, trust management has become a cornerstone for information security and privacy. Although people have recognized the importance of privacy and security for their personal information, they remain uncertain when they have to define and enforce their own access control rules or have to handle indirect information. Indirect information and subjective judgment are the major sources of uncertainty in federated trust management. This paper introduces fuzzy logic into the definition and evaluation of trust, and then provides a formal representation of fuzzy rules. It also offers a set of derivation rules for analyzing and reasoning among fuzzy rules in order to enforce these rules with a certain level of uncertainty. Application of this model to a healthcare environment with pervasive computing devices across trust domains provides a new method to handle uncertainty in trust management for federated and pervasive networks

With the development of Web technology and distributed systems, online collaborations are becoming more common and more demanding. Web services now provide standard mechanisms to enable online interactions. Yet security, privacy and trust-related protection mechanisms for Web services need additional development. In an interconnected network environment, physical connections with proper security protections are required for bridging two autonomous networks. Likewise, collaborating organizations need mechanisms for bridging extant relationships among cooperating parties that provide proper protection for privacy and trust. A trust establishment mechanism for Web services must therefore ensure privacy and owner control at all times due to the subjectivity of trust relationships. This paper describes an indirect trust establishment mechanism to bridge and build new trust relationships from extant trust relationships providing privacy protection and owner control simultaneously

With the emergence of an effective infrastructure supporting Grid Computing and Web Services, service-oriented computing has been growing over the last few years, and service-oriented architectures are becoming an important computing paradigm. When different trust domains control different component services, trust management plays a critical role to smooth the collaboration among component services. The federation of these component services makes new demands for managing trust-related behavior. Although many extant trust management systems deal with intradomain trust behaviors, there is a growing need for effective strategies for managing inter-domain behaviors. In this paper we explore requirements for a federated trust management system. The purpose of this paper is not to suggest a single type of system covering all necessary features; instead, its purpose is to initiate a discussion of the requirements arising from inter-domain federation, to offer a context in which to evaluate current and future solutions, and to encourage the development of proper models and systems for federated trust management. Our discussion addresses issues arising from trust representation, trust exchange, trust establishment, trust enforcement, and trust storage.

1. With the development of web technology and distributed systems, online collaborations are becoming more common and more demanding. These collaborations require online business trust relationships among collaborating organizations. Online business trust relationships can protect the trust, integrity, and privacy of shared resources, which are the foundation for online business. Web services provide standard mechanisms to enable online interactions and further online collaborations. Yet security, privacy and trust-related protection mechanisms for web services need additional development. In an interconnected network environment, bridging extant business relationships to extend the business circle is a convenient and tempting way. Physical connections with proper privacy and security protections are required for bridging two autonomous organizations. Likewise, collaborating organizations need proper protection mechanisms for bridging extant business trust relationships among cooperating parties. These protection mechanisms must therefore ensure privacy and owner control in the entire process of bridging business trust relationships due to the subjectivity of the relationships. This paper describes an indirect trust establishment mechanism using web service enhancements to bridge and build new online business trust relationships from extant business trust relationships providing privacy protection and owner control simultaneously.

Data access and software exchange are often achieved over insecure networks such as the public Internet. System designers are therefore forced to be proactive with regard to verifying the identity of both human users and software processes that request access to protected resources such as factory data. In this paper we show a new security architecture based upon Web services that supports authentication, authorization, and federation. Authentication verifies identity and generates a security token; authorization determines which privileges are allowed to which users; federation permits secure and reliable exchanges of identity across disparate trust domains. We illustrate how these ideas can be used to secure access to a factory Web portal and its underlying database of process data

The lack of effective trust establishment mechanisms for Web services impedes the deployment of trust models for online services. One important issue is the lack of privacy protection in trust establishment. Current Web service technology encourages a client to reveal all its attributes in a standard credential to the service provider for trust establishment. We propose a mechanism whereby the client formulates a single trust primitive by associating a subset of required attributes in a standard credential to negotiate a trust relationship. Client privacy is preserved because only those required attributes are revealed. After negotiation, a trust group element with dynamic validation is used to represent this trust relationship.

With NSF support we developed a hands-on laboratory to augment our CS453 Electronic Commerce Technologies course. We conducted weekly lab sessions that covered HTML, CSS, JavaScript, Perl, CGI, SQL, ASP, and Flash programming. Each topic was covered over a period of one to three weeks and each topic was supported by 10 to 26 individual programming exercises. Four teaching assistants provided real-time support during the laboratory period. Answers to all exercises were accessible on-line so that no student would be needlessly delayed by any particular problem. In this paper we discuss the laboratory's design goals, topical content, and the students' performance, reactions, and suggestions for improvement.

This paper presents a novel multicast protocol, MESH-R, for fully reliable data distribution in a scalable, efficient fashion. A key contribution of our solution is the receiver-driven, dynamic organization of the multicast control structure based on network performance characteristics. The MESH-R framework (1) is a fully distributed, transport-layer solution, (2) presents a robust state synchronization protocol that provides detailed end-system state for reliability, congestion control, group management, and other end-system services, and (3) achieves an efficient, low-latency error control service using a self-organizing, soft-state unicast recovery structure between multicast receivers. In order to assess the relative performance and cost of our solution in comparison to other approaches, MESH-R and three alternative protocols are implemented in a high-fidelity WAN testbed, and their performance and costs compared directly across a range of network assumptions. The results highlight the robustness and agility of MESH-R heuristics in adapting its error recovery and feedback message patterns to dynamic network conditions, and MESH-R performs well relative to the other protocols. The simulation study also uncovers some unexpected results regarding the relative performance between the four protocol designs under certain network scenarios.

The Xpress Transport Protocol (XTP) is designed to support the types of data exchanges typically found in cluster computing applications. It provides all the classic functionality of a transport protocol, plus new services such as transport multicast, multicast group management, transport layer priorities, traffic descriptions for quality of service negotiation, rate and burst control, and selectable error and flow control mechanisms. XTP has wide interconnectivity because it operates over any network layer (IP, CLNP), any datalink layer (LLC, MAC), or directly on top of the AAL of ATM. In general, XTP avoids coupling policy with mechanism; XTP offers services but the user's application defines what communications paradigm is most appropriate for its particular environment. In spite of its richer functionality, XTP's reliable data transfer modes have performance characteristics similar to TCP and XTP's unacknowledged transfer mode has performance similar to UDP. Since XTP can run simultaneously with other transport protocols, and can run over whatever network layer (if any) is provided, it represents a low risk way to exploit the increased functionality required for cluster computing without sacrificing connectedness or interoperability.

This paper addresses support for multi-party multimedia applications delivered across datagram wide-area networks such as the Internet. Our approach is to develop a novel retransmission protocol framework, MESH, that protects delay-sensitive multicast data streams against most network errors and delay variations. The MESH approach partitions the multicast group into subgroups, elects subgroup managers called active receivers, and localizes error recovery between active receivers. Novel features of our protocol include an explicit model for delay-sensitive retransmission, run-time evaluation at an active receiver to determine its retransmission partner, and the ability to handle multiple multicast data sources without building a per-source control framework. To evaluate the performance and network costs of MESH, a wide-area network simulation of digital video distribution is constructed and studied over a range of network loss rates. Simulation experiments show that both MESH and an alternative scheme, end-to-end forward error correction, provide excellent error control with MESH incurring significantly less network overhead.

A background traffic model is fundamental to packet-level network simulation since the background traffic impacts packet drop rates, queueing delays, end-to-end delay variation, and also determines available network bandwidth. In this paper, we present a statistical characterization of wide-area IP traffic based on 90-minute traces taken from a week-long trace of packets exchanged between a large campus network a state wide educational network, and a large Internet service provider. The results of this analysis can be used to provide a basis for modelling background load in simulations of wide-area packet-switched networks such as the Internet, contribute to understanding the fractal behavior of wide-area network utilization, and provide a benchmark to evaluate the accuracy of existing traffic models. The key findings of our study include the following: (1) both the aggregate packet stream and its component substreams exhibit significant long-range dependencies in agreement with other traffic studies. (2) the empirical probability distributions of packet arrivals are log-normally distributed. (3) packet sizes exhibit only short-term correlations and (4) the packet size distribution and correlation structure are independent from both network utilization and time of day.

Presents the design and implementation of a digital image capture and distribution system that supports remote ultrasound examinations and, in particular, real-time diagnosis for these examinations. The system was designed in conjunction with radiologists and staff in the Department of Radiology at the University of Virginia Hospital. Based on readily available microcomputer components, our teleultrasound system handles the acquisition, digitizing, and reliable transmission of still and moving images generated by an ultrasound machine. The digital images have a resolution of 640/spl times/480 with an 8-bit color plane, con be captured at rates up to 30 frames/sec, and are compressed and decompressed in real-time using specialized hardware. While scalable to communications networks of any transmission speed, initial deployment is envisioned for 1.5 Mbit/s T-1 leased lines. To achieve real-time still image distribution and to reduce the bandwidth necessary for motion video, the teleultrasound design employs lossy image compression based on the JPEG standard. The effects of JPEG compression on diagnostic quality are being studied in a separate signal detection study with the Department of Radiology at the University of Virginia.>

Support for digital audio and video streams (continuous media) in desktop computers is increasingly widespread and hence the demand for high-quality network distribution of continuous media streams is growing. While current local area and campus-wide networks have high-bandwidth links and powerful routers, the delay sensitivities of continuous media streams pose a challenge. Our paper provides an empirical investigation of the feasibility of transmitting real-time packet streams across a large extant campus network, the University of Virginia campus network. This network is representative of contemporary large enterprise networks consisting of Ethernet segments connected by high-performance routers to high-speed back-bones, e.g., FDDI rings. We focus on interactive packet voice. Delay and loss measurements are obtained for packet voice streams transmitted over three paths through the network. These measurements are used to examine the voice protocol parameters for this environment, including a novel retransmission-based error control proposed by Dempsey (1994).>

A new, unified approach towards the delay and error constraints that influence the quality of a digitized voice over asynchronous local computer networks is proposed. This approach, called Slack automatic repeat request (ARQ) (S-ARQ), proposes an extension of the control time mechanism for jitter control in order to provide error recovery through retransmissions in the case of packet losses in the network. While existing packet voice protocols do not use retransmission schemes, the authors' simulation studies indicate that extended control time can provide significant error coverage while remaining within the same order of magnitude as the control time required for jitter control. Simulation experiments reveal that the concept is clearly feasible within the end-to-end delay constraints in packet voice transmissions.

The Computer Networks Laboratory at the University of Virginia has implemented the SAFENET lightweight protocol suite on the Navy's Desktop Tactical Computer (DTC-2). One of the research questions was whether to embed XTP in the DTC-2's UNIX kernel or run XTP on an attached processor on the DTC-2's VMEbus. Both strategies were implemented and the attached processor approach proved to have somewhat higher performance. The hardware and software system environment is surveyed, the performance achieved at the media access controller (MAC), transport, and user levels is reported, and an analytic model for predicting throughput and latency as a functon of vendor-provided or easily measured system parameters is developed. The authors' conclusions identify both the advantages and performance concerns inherent to off-host protocol execution.

The Computer Networks Laboratory at the University of Virginia has used the Xpress Transfer Protocol (XTP) to construct the Xpress File System (XFS), a local area network peer-to-peer file sharing system for MS-DOS-based IBM personal computers. XFS is usable for a variety of purposes, most notably the real time delivery of multimedia data streams. Using specialized hardware and software it is shown that XFS is capable of maintaining sufficient file system to file system throughput to support the real time delivery of the 30 frames per second audio/video data stream generated by the Fluent system. A general discussion of the architecture and operation of XFS is provided, and an overview of the Fluent system to generate and process motion video data streams is given. The authors describe plans to extend the video mail system into a video conferencing system using XTP.

The feasibility of distributing digitized voice over FDDI using the Xpress Transfer Protocol (XTP) is examined. The requirements of such a system, the role of a communications protocol, and the overall experimental design are studied. A basic configuration is developed to measure the system throughput, latency, and jitter, and from these data the number of voice channels that a given configuration would support is computed. The basic system is then perturbed by adding up to 75 Mb/s of background synchronous load to the FDDI network, and by artificially dropping up to 10% of the transmitted packets to force retransmissions. Whether XTP's reliable transport multicast, which provides a 1-to-m transmission capability, provides a latency reduction when compared to m serial unicasts is investigated. Conclusions regarding the utility of a transport protocol for both single segment and multiple-segment LANs are presented.>

The authors discuss their experience in implementing an Abstract Syntax Notation One compiler, ASN.1 is a descriptive notation used to define the informational content of the OSI upper layer protocols. The Basic Encoding Rules are a transfer syntax that may be applied to instances of ASN.1. Collectively, these two protocols solve the disparate data representation problem within an OSI network. The authors describe the purpose and basic features of both ASN.1 and BER, their theoretical disposition within the OSI Reference Model, their practical application that includes the details of this implementation, and finally some insights into the protocols themselves is offered.>

An international group of researchers, led by Dr. G. Chesson at Silicon Graphics Inc., has developed a new transport+network protocol for modern distributed computing systems. Extant protocols (e.g. TCP, TP4, Delta-t, NETBLT, GAM-T-103, VMTP, URP) were first examined to understand their solutions to common data transfer problems. Then, a new protocol called XTP (Xpress Transfer Protocol) was developed which both reuses ideas known to be good and breaks new ground in the definition and delivery of communications services. XTP is specifically designed for implementation in hardware as the Protocol Engine which will connect to a commercial FDDI chipset. The author reports first on the philosophy underlying XTP's design, and then on his experiences in producing a software implementation.>

A unique multicast service that exploits features in the next-generation, real-time transfer layer protocol, the Xpress transfer protocol (XTP), is described. As a modular part of the service interface routines, the multicast facility, the Multidriver, exploits features in XTP in order to provide two multicast services, one unreliable and one reliable. In its reliable mode the service offers error, flow, and rate-controlled multidestination delivery of arbitrary-sized messages, with provision for the coordination of reliable reverse channels. Performance measurements on a single-segment Proteon ProNET-4 4-Mb/s 802.5 token ring with heterogeneous nodes are discussed.>

The Xpress transfer protocol (XTP), a transport and network layer protocol suitable for use as a SAFENET transfer service in the US Navy's emerging survivable adaptable fiber optic embedded network (SAFENET) specification, is described. The background protocol, the implementation environment, the communications architecture, the user interface, some working XTP demonstration programs, and some measured performance results are described. The experience with XTP shows that the implementation is stable; operational in PC and VME bus environments; and capable of being transmitted reliably over Ethernet, token ring, and fiber distributed data interface (FDDI) networks.>

The authors examine the ISO Open Systems Interconnection philosophy and the error detection mechanisms at each layer. They trace the progress of a message through each of these layers and comment on the overhead of error detection. They also offer some performance results at the transport layer. They argue that the error detection rate in modern LANs is not improved by mechanisms above the MAC, but rather that performance is severely degraded.>

A study has been initiated to determine whether the xpress transfer protocol (XTP) is a suitable candidate for use alongside a full ISO stack. A description is given of initial plans for implementing XTP and for comparing its performance to ISO TP4. Performance graphs for the basic datalink service which underlies the protocol stack are presented.>

A discussion is given on whether the Ada model of concurrency is suitable for implementing the seven-layer OSI reference model. Using the communications model introduced by R.J.A. Buhr (1984), they determine the overhead introduced by Ada when the model is implemented on two single-processor machines, a VAX 11/785 and a Rational 1000. The authors then calculate a lower bound on expected message delay. A novel model using server tasks is proposed and shown to have better performance. The authors investigate performance on an eight-processor Sequent Model 821 and a 14-processor Encore Multimax 320 by implementing the Buhr model, the server task model and a third model which abandons the Ada rendezvous in favor of procedure calls. They determine the Ada overhead per message as a function of the number of processors and calculate lower bounds on expected message delay attributable to Ada overhead for all three communications models.>

The authors theorize that, in a normal environment where the network is not overloaded, no packet priority levels should be implemented, and that the use of multiple priorities will generally degrade overall performance. They suggest that for priority operation to be meaningful, priority service must be given at the point of the service bottleneck, which generally is not network access, but rather higher-level processing (e.g. OSI protocols). They further argue that, if a message has not been treated expeditiously prior to its bid for network access, assigning a high network-access priority does little to decrease its latency. Results from simulations made of the IEEE 802.5 token ring and the Fiber Data Distributed Interface token ring are presented to reinforce the authors' thesis.>

The authors have developed a real-time messaging system for token ring networks. The system is currently operational in a shipboard environment. The system conforms to the IEEE 802.2 LLC and 802.5 token-ring standards and is consistent with the Navy's emerging SAFENET specification. The authors provide a library of 'C' routines which establish and manage sockets, transmit and receive messages, and report network status. They show a sample user interface for a basic datagram service and another for a reliable virtual circuit service. They present performance data from an implementation on PCs and PC/ATs. Using the proposed software, a single transmitting station can generate up to 1.42 Mb/s, and 100-byte messages can be processed every 4 ms.>

The authors present performance measurements for data transfer services at the CASE (common application service elements), transport, and datalink layers of Motorola's implementation of the Manufacturing Automation Protocol (MAP) for a range of message sizes. They compare the performance results of using a stop-and-wait vs. a sliding window protocol, and find that the sliding window permits a 40% speedup. The observe that end-to-end latency exceeds transmission delay by a factor of 25 for 1000-byte messages, thereby making all network access, transmission, and propagation delays negligible when compared to protocol processing. They show that the throughput of CASE and transport is basically linear with message size whereas the throughput of datalink is not. They identify the bottlenecks which constrain total system throughput at each layer.>

Abstract The Computer Networks Laboratory at the University of Virginia has developed a real‐time messaging service that runs on IBM PCs and PC/ATs when interconnected with a Proteon ProNET‐10 token ring local area network. The system is a prototype for a real‐time communications network to be used aboard ships. The system conforms to the IEEE 802.2 logical link control standard for type I (connectionless, or datagram) service, with an option for acknowledged datagrams. The application environment required substantial network throughput and bounded message delay. Thus, the development philosophy was to emphasize performance initially and to offer only primitive user services. After providing and measuring the performance of a basic datagram service, the intent is to add additional user services one at a time and to retain only those which the user can ‘afford’ in terms of their impact on throughput, delay, and CPU utilization. The current system is programmed in C. The user interface is a set of C procedure calls that initialize tables, reserve buffer space, send and receive messages, and report network status. The system is now operational, and initial performance measurements are complete. Using this system, an individual PC can transmit or receive approximately 200 short (about 100 bytes) messages per second, and the PC/AT operates at nearly 500 short messages per second.

As a research project for NASA's Langley Research Center, we developed a variation on MIL-STD-1553B (the military standard for avionics buses) whose goal was increased fault tolerance. The resulting protocol, called implicit token passing (ITP), replaces an explicit token with brief "soundoff" messages from all nodes participating on the LAN. Since every node participates on every "token cycle," bus silence is an error indication and initiates recovery action. By encoding state information in the headers of the transmitted words, nodes are continuously aware of the global state of the network. This local knowledge of the global network state allows the system to continue operation in spite of nodal failures. A station which fails but then recovers can quickly assess the global network state and then safely rejoin the active nodes. ITP features high throughput and bounded message delay, and achieves high reliability through tolerance of failed nodes and automatic resynchronization when failed nodes are revived. The protocol is ideally suited for a bus topology and fiber optic media.

The IEEE 802.4 token bus defines both synchronous and asynchronous message access classes. Their performance is compared for networks implementing a single message class. Throughput bounds are derived from the access class timer. It is shown that for some configurations, the asynchronous class yields lower observable message delays. Mean observable delay is minimized by allowing each medium access controller unrestricted service.

On the premise that many of the fundamental concepts of computer science can be better taught in a hands-on, dedicated computing environment (i.e., a microcomputer), as opposed to a large multi-purpose system in which the student is insulated from the machine by multiple layers of operating system software, we have developed a microcomputer-based lecture/lab course to teach CPU organization, digital computer architecture, and assembly language programming as a third undergraduate course in computer science. By beginning with simple machine organizations and simple assembly languages, and later on making a transition toward more complex architectures and languages, the transfer of knowledge and experience is positive at every step. The same laboratory also supports a graduate course in microcomputer systems design which teaches hardware technology, component specification, operating system design, hardware/software tradeoffs, and practical applications such as process control. This paper outlines the motivation and justification for the project, and then discusses the actual design of these courses and their supporting laboratory.This project is supported in part by two grants from the National Science Foundation: SER-7915929 for the acquisition of the microcomputer equipment and SER-8000802 for the development of the undergraduate course material.

This paper reports the author's progress toward developing a microcomputer laboratory suitable for teaching some of the basic concepts of computer science. The rationale for this approach is presented, followed by a description of the hardware and software environment thus far developed. Finally we present two course outlines: one for an undergraduate laboratory course in computer systems and another for a graduate course in microcomputer systems design for engineers.