The provision of advanced computational services within networks is rapidly becoming both feasible and economical. We present a general approach to the problem of configuring application sessions that require intermediate processing by showing how the session configuration problem can be transformed to a conventional shortest path problem. We show, through a series of examples, that the method can be applied to a wide variety of different situations.

The purpose of this document is to present the overview of tte plugin management architecture and the description of the software developed for the scalable, high performance active network node project in Washington University, St. Louis. The plugin management is a user space daemon program that runs at the code(plugin) server and at the active network component of a router or a switch port processor. The running programs cooperate to load plugins from the code server to the active network component. This software is intended to be used among multiple ...Read More

Synthesizer is a pattern language for designing digital synthesizers using modular synthesis in software to generate sound. Software developed according to this pattern language emulates the abilities of an analog synthesizer. Modular synthesis is one of the oldest sound synthesis techniques. It was used in the earliest analog synthesizers, like the Moog [1] and ARP [2]. These machines introduced the oscillator-filter-amplifier paradigm, where sound generated by an oscillator is passed through a series of filters and amplifers before being sent to a speaker. These first machines had physical modules through ...Read More

Active networks allow customized processing of data traffic within the network which can be used by applications to improve the quality of their sessions. To simplify development of active applications in a heterogeneous environment, we propose active network pipes as a programming abstraction to specify transmission and processing requirements. We describe a routing algorithm that maps application session requirements onto network resources and determines an optimal route through the network transiting all required processing sites. Additionally, we propose a network software architecture to implement the functionality required to support active ...Read More

The FPX provides simple and fast mechanisms to process cells or packets. By performing all computations in FPGA hardware, cells and packets can be processing at the full line speed of the card [currently 2.4 Gbits/sec]. A sample application, called 'Hello World' has been developed that illustrates how easily an application can be implemented on the FPX. This application uses the FPGA hardware to search for a string on a particular flow and selectively replace contents of the payload. The resulting circuit operates at 119 MHz on a Xilinx XCV ...Read More

For systems with a large number of FPGAs, where a design is instantiated across multiple FPGAs in a chassis, an efficient mechanism of programming the FPGA devices is needed. The mechanism described herein allows multiple FPGAs to be programmed across a backplane. Only a single configuration PROM is required to store the configuration for the multiple instances of the design. When the system boots, all FPGAs are programmed in parallel. This design is applicable to any system which contains a multiple board system which has instances of identical FPGA implementations ...Read More

This paper explores the range of constructs and issues facing the designer of mobile code systems which allow for the unit of mobility to be finer-grained than that of execution. Mobile UNITY, a notation and proof logic for mobile computing, provides for this research a clean abstract setting, i.e., unconstrained by compilation and performance considerations traditionally associated with programming language design. Within the context of Mobile UNITY, we take the extreme view that every line of code and every variable declaration is potentially mobile, i.e., it may be duplicated and/or ...Read More

The IP multicast model allows scalable and efficient multi-party communication, particularly for groups of large size. However, deployment of IP multicast requires substantial infrastructure modifications and is hampered by a host of unresolved open problems such as reliability, flow and congestion control, security and access control. Motivated by these problems, we have designed and implemented ALMI, an application level group communication middleware, which does not rely on network infrastructure support and thus, allows accelerated deployment and simplified configuration at the cost of a relatively small increase in traffic load. ALMI ...Read More

LIME is a middleware supporting the development of applications that exhibit physical mobility of hosts, logical mobility of agents, or both. LIME adopts a coordination perspective inspired by work on the Linda model. The context for computation, represented in Linda by a globally accessible, persistent tuple space, is represented in LIME by transient sharing of the tuple spaces carried by each individual mobile unit. Linda tuple spaces are also extended with a notion of location and with the ability to react to a given state. The hypothesis underlying our work ...Read More

In this paper we consider the automatic discovery of design (programming) patterns. While patterns have surfaced as an effective mechanism for authoring and understanding compelx software, popular languages lack facilities for direct specification of patterns or verification of pattern usage in program specifications. Static analysis for patterns is provably undecidable; we focus on discovery and verification of patterns by analyzing dynamic sequences of method calls on object. We show a proof-of-concept of our approach by presenting the results of analyzing a Java program for Iterator patterns.

The design of ad hoc mobile applications often requires the availability of a consistent view of the application state among the participating hosts. Essential to constructing a consistent view is the ability to know what hosts are within proximity of each other, i.e., form a group in support of the particular application. In this paper we propose an algorithm that allows hosts within communication range to maintain a consistent view of the group membership despite movement and frequent disconnections. The novel features of this algorithm are its reliance on location ...Read More

Mobility entails the study of systems in which components change location, in a voluntary or involuntary manner, and move across a space that may be defined to be either logical or physical. Coordination is concerned with what happens when two or more components come in contact with each other. In this paper we put forth a working definition of coordinatoin, we construct argumetns that demonstrate that coordination is central to understanding mobility, we explore the intellectual richness of the notion of coordination, and we consider the practical implications of coordination-centered ...Read More

Current reliable multicast protocols do not have scalable congestion control mechanisms and this deficiency leads to concerns that multicast deployment may endanger stability of the network. In this paper, we present a sender-based approach for multicast congestion control targeted towards reliable bulk data transfer. We assume that there are a few bottleneck links in a large scale multicast group at any time period and these bottlenecks persist long enough to be identified and adapted to. Our work focus on dynamically identifying the worst congested path in the multicast tree and ...Read More

We present a new algorithm and framework for dynamic routing of bandwidth guaranteed flows. The problem is motivated by the need to dynamically set up bandwidth guaranteed paths in carrier and ISP networks. Traditional routing algorithms such as minimum hop routing or widest path routing do not take advantage of any knowledge about the traffic distribution or ingress-egress pairs, and therefore can often lead to severe network underutilization. Our work is inspired by the recently proposed "minimum interference routing" algorithm (MIRA) of Kodialam and Lakshman, but it improves on their ...Read More

Distributed object computing (DOC) middleware shields developers from many tedious and error-prone aspects of programming distribued applications. Without proper support from the middleware, however, it can be hard to evolve distributed applications after they are deployed. Therefore, DOC middleware should support meta-programming mechanisms, such as smart proxies and interceptors, that improve the adaptability of distributed applications by allowing their behavior to be modified without drastically changing existing software. This paper presents three contributions to the study of metaprogramming mechanisms for DOC middleware. First, it illustrates, compares, and contrasts several meta-programming ...Read More

Demands for flexible processing has moved general-purpose processing into the data path of networks. With the development of System-On-a-Chip technology, it is possible to put several processors with memory and I/O components on a single ASIC. We present a model of such a system with a simple performance metric and show how the number of processors and cache sizes can be optimized for a given workload. Based on a telecommunications benchmark we show the results of such an optimization and discuss how specialied hardware and appropriate scheduling can further improve ...Read More

We use the SRB (Storage Request Broker) middleware to design and implement a storage archival system which will be used to archive Neuroscience data. As part of the design process, we developed and used an experimenter's workbench to measure SRB performance. These experiments improved our understanding of both the functionality and the performance of the SRB. This technical report describes the scripts in the experimenter's workbench, the archiving scripts, and performance measurements.

This paper proves that the revenue equivalence theorem ceases to hold for auctions without common knowledge about the agents' prior beliefs. That is, different auction forms yield different expected revenue. To prove this, an auction game is converted to a Bayesian decision problem with an infinite hierarchy of beliefs. A general solution for such Bayesian decision problems is proposed. The solution is a generalization of the standard Bayesian solution and coincides with it for finite belief trees and for trees representing common knowledge. It is shown how the solution generalizes ...Read More

To be an effective platform for performance-sensitive real-time and embedded applications, off-the-shelf OO middleware like CORBA, DCOM, and Java RMI must preserve communication-layer quality of service (QoS) properties to applications end-to-end. However, conventional OO middleware interoperability protocols, such as CORBA's GIOP/IIOP or DCOM's MS-RPC, are not well suited for applications that cannot tolerate the message footprint size, latency, and jitter associated with general-purpose messaging and transport protocols. It is essential, therefore, to develop standard plugable protocols frameworks that allow custom messaging and transport protocols to be configured flexibly and used ...Read More

In this paper, we take the extreme view that every line of code is potentially mobile, i.e., may be duplicated and/or moved from one program context to another on the same host or across the network. Our motivation is to gain a better understanding of the range of constructs and issues facing the designer of a mobile code system, in a setting that is abstract and unconstrained by compilation and performance considerations traditionally associated with programming language design. Incidental to our study is an evaluatoin of the expressive power of ...Read More

Advances in wireless communication and network computing technologies make possible new kinds of applications involving transient interactions among physical components that move across a wide range of spaces, from the confines of a room to the airspace across an ocean, and require no fixed networking infrastructure to communicate with one another. Such components may come together to form ad hoc networks for the purpose of exchanging information or in order to engage in cooperative task-oriented behaviors. Ad hoc networks are assembled, reshaped and taken apart as components move in and ...Read More

The provision of a reliable communication infrastructure for mobile agents is still an open research issue. The challenge to reliability we address in this work does not come from the possibility of faults, but rather from the mere presence of mobility, which slightly complicates the problem of ensuring the delivery of information even in a fault-free network. For instance, the asynchronous nature of message passing and agent migration may cause situations where messages forever chase a mobile agent that moves frequently from one host to another. Current solutions rely on ...Read More

As computing components get smaller and people become accustomed to having computational power at their disposal at any time, mobile computing is developing as an important research area. One of the fundamental problems in mobility is maintaining connectivity through message passing as the user moves through the network. An approach to this is to have a single home node constantly track the current location of the mobile unit and forward messages to this location. One problem with this approach is that during the update to the home agent after movement, ...Read More

In this paper, we propose a new framework called Active Network Management and Control (ANMAC) for the management and control of high speed networks. The software architecture in ANMAC allows routers to execute dynamically loadable kernel plug-in modules which perform diagnostic functions for network management. ANMAC uses mobile probe packets to perform efficient resource reservation (using our novel reservation scheme), facilitate feedback-based congestion control, and to provide "distributed debugging" of complex anomalous network behavior. ANMAC also provides security measures against IP spoofing, and other security attacks. The network manager has ...Read More

With wide deployment of high-speed networks such as vBNS today, video-conference applications over WANs have become increasingly feasible. MMX has proven to be a good desktop video-conference devide for local ATM networks. Now, ALX has been designed to extend MMX's video conferencing capability to IP-over-ATM WANs such as vBNS. In this report, we discuss a floor control protocol for ALX video-conference applications. We first show how an "ideal" protocol should behave to meet our requirements. Then we compare three protocols based on distributed algorithms, and a protocol based on a ...Read More

The term distributed computing conjures the image of a fixed network structure whose nodes support the execution of processes that communicate with each other via messages traveling along links. Peer-to-peer communication is feasible but client-server relationships dominate. More recently, servers have been augmented with brokerage capabilities to facilitate discovery of available services. Stability is the ideal mode of operation; changes are relatively slow; even in the case of failure, nodes and links are expected eventually to come back up. By contrast, mobility represents a total meltdown of all the stability ...Read More

Pattern matching techniques have been useful in solving many problems associated with computer science, including data compression (Chrochemore and Lecroq, 1996), data encryption (RSA Laboratories, 1993), and computer vision (Grimson and Huttenlocher, 1990). In recent years, developments in molecular biology have led to large scale sequencing of genomic DNA. Since this data is being produced in such rapid fasion, tools to analyze DNA segments are desired. The goal here is to discuss various techniques and tools for solving various pattern matching questions in computational biology, including optimal sequence alignment, multiple ...Read More

The Human Genome Project (HGP) has led to the deposit of human genomic sequence in the form of sequenced clones into various databases such as the DNA Data Bank of Japan (DDBJ) (Tateno and Gojobori, 1997), the European Molecular Biology Laboratory (EMBL) Nucleotide Sequence Database (Stoesser, et. al., 1999), and GenBank (Benson, et. al., 1998). Many of these sequenced clones occur in regions where sequencing has taken place either within the same sequencing center or other centers throughout the world. The assembly of extended segments of genomic sequence by looking ...Read More

In automated negotiation systems consisting of self-interested agents, contracts have traditionally been binding. Leveled commitment contracts - i.e. contracts where each party can decommit by paying a predetermined penalty - were recently shown to improve Pareto efficiency even if agents rationally decommit in Nash equilibrium using inflated thresholds on how good their outside offers must be before they decommit. This paper operationalizes the four leveled commitment contracting protocols by presenting algorithms for using them. Algorithms are presented for computing the Nash equilibrium decomitting thresholds and decommitting probabilities given the contract ...Read More

Combinatorial auctions, i.e. auctions where bidders can bid on combinations of items, tend to lead to more efficient allocations than traditional auctions in multi-item auctions where the agents' valuations of the items are not additive. However, determining the winners so as to maximize revenue is NP-complete. First, existing approaches for tackling this problem are reviewed: exhaustive enumeration, dynamic programming, approximation algorithms, and restricting the alloable combinations. Then we present our search algorithm for optimal winner determination. Experiments are shown on several bid distributions. The algorithm allows combinatorial auctions to scale ...Read More

This paper presents eMediator, a next generation electronic commerce server that demonstrates some ways in which AI, algorithmic support, and game theoretic incentive engineering can jointly improve the efficiency of ecommerce. First, its configurable auction house includes a variety of generalized combinatorial auctions, price setting mechanism, novel bid types, mobile agents, and user support for choosing an auction type. Second, its leveled commitment contract optimizer determines the optimal contract price and decommitting penalties for a variety of leveled commitment contracting protocols, taking into account that rational agents will decommit insincerely ...Read More

This paper analyzes automated distributive negotiation where agents have firm deadlines that are private information. The agents are allowed to make and accept offers in any order in continuous time. We show that the only sequential equilibrum outcome is the one where the agents wait until the first deadline, at which point that agent concedes everything to the other. This holds for pure and mixed strategies. So, interestingly, rational agents can never agree to a nontrivial split because offers signal enough weakness of bargaining power (early deadline) so that the ...Read More

In computational markets utilizing algorithms that establish a general equilibrium, competitive behavior is usually assumed: each agent makes its demand (supply) decisions so as to maximize its utility (profit) assuming that it has no impact on market prices. However, there is a potential gain from strategic behavior via speculating about others because an agent does affect the market prices, which affect the supply/demand decisions of others, which again affect the market prices that the agent faces. Determining the optimal strategy when the speculator has perfect knowledge about the other agents ...Read More

In automated negotiation systems consisting of self-interested agents, contracts have traditionally been binding. Leveled commitment contracts - i.e. contracts where each party can decommit by paying a predetermined penalty - were recently shown to improve expected social welfare even if agents decommit insincerely in Nash equilibrium. Such contracts differ based on whether agents have to declare their decommitting decisions sequentially or simultaneously, and whether or not agents have to pay the penalties if both decommit. For a given contract, these protocols lead to different decommitting thresholds and probabilities. However, this ...Read More

Current IP routers are stateless: they forward individual packets based on the destination address contained in the packet header, but maintain no information about the application or flow to which a packet belongs. This stateless service model works well for best effort datagram delivery, but is grossly inadequate for applications that require quality of service guarantees, such as audio, video, or IP telephony. Maintaining state for each flow is expensive because the number of concurrent flows at a router can be in the hundreds of thousands. Thus, stateful solutions such ...Read More

Internet routers forward packets based on the destination address of a packet. A packet's address is matched against the destination prefixes stored in the router's forwarding table, and the packet is sent to the output interface determined by the longest matching prefix. While some existing schemes work well for IPv4 addresses, we believe that none of the current schemes scales well to IPv6, especially when fast updates are required. As the Internet evolves into a global communication medium, requiring multiple addresses per user, the switch to longer addresses (e.g. IPv6) ...Read More

This report summarizes progress on Washington University's Terabit Burst Switching Project, supported by DARPA and Rome Air Force Laboratory. This project seeks to demonstrate the feasibility of Burst Switching, a new data communication service which can more effectively exploit the large bandwidths becoming available in WDM transmission systems, than conventional communication technologies like ATM and IP-based packet switching. Burst switching systems dynamically assign data bursts to channels in optical data links, using routing information carried in parallel control channels. The project will lead to the construction of a demonstration switch ...Read More

Current research in Active Networking is focused on developing software architectures and defining funtionality of Execution Environments. While active network systems show superior functionality compared to traditional networks, they only operate at substantially lower link speeds. To increase the acceptance of Active Network in environments where link speeds of several Gb/s are common, we propose a hardware architecture that performs high-speed packet handling while providing the same flexibility as a common software system. The design exploits the independence between data streams for parallel processing. To measure the impact of different ...Read More

This paper presents a benchmark, CommBench, for use in evaluating and designing telecommunications network processors. The benchmark applications focus on small, computationally intense program kernels typical of the network processor environment. The benchmark is composed of eight programs, four of them oriented towards packet header processing and four oriented towards data stream procesing. The benchmark is defined and various characteristics of the benchmark are presented. These include instruction frequencies, computational complexity, and cache performance. These measured characteristics are compared to the SPEC benchmark which has traditionally been used in evaluating ...Read More

Active networking is a general approach to incorporating general-purpose computational capabilities within the communications infrastructure of data networks. This paper proposes a design of a scalable, high performance active router. This is used as a vehicle for studying the key design issues that must be resolved to allow active networking to become a mainstream technology.

The last few years have seen the development of a model for Integrated Services Internet, which extends the traditional Internet by adding multiple service classes in addition to the traditional best effort service class, and a signaling protocol called RSVP for applications to reserve resources. While this framework has been standardized in the IETF WGs and the RSVP protocol has been defined, there has been no movement towards a commercial implementation of this framework, principally due to its perceived complexity and lack of scalability. This paper analyzes RSVP, discusses some ...Read More

Present day routers typically employ monolithic operating systems which are not easily upgraded and extensible. WIth the rapid rate of protocol development it is becoming increasingly important to dynamically upgrade router software in an incremental fashion. We have designed and implemented a high performance, modular, extended integrated services router software architecture in the NetBSD operating system kernel. This architecture allows code modules, called plugins, to be dynamically added and configured at run time. One of the novel features of our design is the ability to bind different plugins to individual ...Read More

We study an on-line problem that is motivated by the networking problem of dynamically adjusting delays of acknowledgments in the Transmission Control Protocol (TCP). The theoretical problem we study is the following. There is a sequence of n packet arrival times A = and a look-ahead coefficient L. The goal is to partition A into k subsequences sigma1, sigma2, ...,sigmak (where a subsequence end is defined by an acknowledgment) that minimizes a linear combination of the cost for the number of acknowledgments sent and the cost for the additional latency ...Read More

We discuss congestion control mechanisms in multicast transport protocols and we propose TCP-M - a TCP-friendly Multicast transport protocol. TCP-M uses IP multicast to deliver data packets and acknowledgements to provide reliability. Ack implosion at the source is prevented by fusing acknowledgements at some intermediate routers. TCP-M reacts to network congestion exactly like TCP by having the sender emulate a TCP sender.

We describe mechanisms to enhance the reliability and performance of Mobile IP. In Mobile IP today home agents and foreign agents are single points of failure and potential performance bottlenecks. For example, a home agent crash can lead to communication failure if the mobile is away from home. In this paper we describe new mechanisms to allow redundant home and foreign agents. Redundant agents can take over from each other in case of failure, and also split load amongst themselves. Our mechanisms are simple, transparent to existing mobile nodes, and ...Read More

Goldberg, Goldman, and Scott demonstrated how the problem of recognizing a landmark from a one-dimensional visual image can be mapped to that of learning a one-dimensional geometric pattern and gave a PAC algorithm to learn that class. In this paper, we present an efficient on-line agnostic learning algorithm for learning the class of constant-dimension geometric patterns. Our algorithm can tolerate both classification and attribute noise. By working in higher dimensional spaces we can represent more features from the visual image in the geometric pattern. Our mapping of the data to ...Read More

We introduce the UAV learning model in which some of the attributes in the examples are unspecified. In our model, an example x is classified positive (resp., negative) if all possible assignments for the unspecified attributes result in a positive (resp., negative) classification. Otherwise the classificatoin given to x is "?" (for unknown). Given an example x in which some attributes are unspecified, the oracle UAV-MQ responds with the classification of x. Given a hypothesis h, the oracle UAV-EQ returns an example x (that could have unspecified attributes) for which ...Read More

We study non-preemptive, online admission control in the hard deadline model: each job must be either serviced prior to its deadline, or be rejected. Our setting consists of a single resource that services an online sequence of jobs; each job has a length indicating the length of time for which it needs the resource, and a delay indicating the maximum time it can wait for the service to be started. The goal is to maximize total resource utilization. The jobs are non-preemptive and exclusive, meaning once a job begins, it ...Read More

As the primary tool for detecting breast carcinoma, mammography provides visual images from which a trained radiologist can identify suspicious areas that suggest the presence of cancer. We describe an approach to image processing that reduces an image to a small number of values based on its structural characteristics using wavelets and neural networks. To illustrate its utility, we apply this methodology to the automatic screening of mammograms for mass lesions. Our results approach performance levels of trained human mammographers.

With recent advances in wireless communication technology, mobile computing is an increasingly important area of research. A mobile system is one where independently executing components may migrate through some space during the course of the computation, and where the pattern of connectivity among the components changes as they move in and out of proximity. Mobile UNITY is a notation and proof logic for specifying and reasoning about mobile systems. In this paper it is argued that Mobile UNITY contributes to the modular development of system specifications because of the declarative ...Read More