NetScript is a language system for dynamically programming and deploying protocol software in an active network. NetScript programs are packet flow processors composed out of dataflow components. They can be dynamically dispatched to remote nodes, composed with installed software, and executed to enhance node functions. NetScript has proven useful for implementing a variety of practical network systems including Internet protocol (IP) routers, firewalls, protocol analyzers, traffic shapers, load balancers, various queuing disciplines, intrusion detection systems, Web transport protocols, and IP telephony systems. The paper illustrates the principal concepts of NetScript with an example application, an active IP router that is dynamically extended to provide firewall protection.

This paper describes novel protection technologies, developed by the MarketNet project at Columbia University, that shifts power from attackers to defenders, giving the defenders control over the exposure to attacks and over detectability and accountability of attackers. MarketNet uses market-based techniques to regulate access to resources. Access to a resource must be paid-for with currency issued by its domain. Domains can control the power of attackers by limiting the budgets allocated to them, and control the exposure of resources by setting their prices, effectively providing a quantifiable access control mechanism. Domains can monitor currency flows and use uniform resource-independent statistical algorithms to correlate and detect access anomalies indicating potential attacks. Currency is marked with unique identifiers that permit domains to establish verifiable accountability in accessing their resources. Domains control and fine tune their exposure to attacks; adjust this exposure in response to emerging risks; detect intrusion attacks through automated, uniform statistical analysis of currency flows; and establish coordinated response to attacks. MarketNet mechanisms unify and kernelize global information systems protection by containing all protection logic in a small core of software components. The paper presents the architecture and operation of MarketNet along with the design and implementation of the main architectural components. The paper illustrates the application of MarketNet to the protection of the simple network management protocol (SNMP) and compares it with the security features offered by SNMPv3.

Several QoS provisioning mechanisms such as differentiated services (Diffserv) and integrated services (Intserv) have been recently devised and applied to bring quality of service (QoS) to the Internet. This paper studies end-end QoS performances of two QoS-demanding applications using different transport protocols. Both applications are tested in a real network environment, with end-end QoS provisioning by Intserv. They use QoSockets, a new extension of QoS specification and management to the Berkeley sockets. Their performance in terms of throughput, delay, jitter, and loss are measured under a number of test cases combining several factors: (1) single or multiple flows, with or without resource reservations; (2) normal, heavy, or overloaded scenarios; (3) uni- or bi-directional streams; and (4) TCP or UDP protocols. The experimental results show that the performance of two applications with the Intserv resource reservations are significantly improved, but not always guaranteed. It is also shown that UDP applications are able to get the requested QoS while TCP applications may not because of the nature of its bi-directional traffic flow. The paper provides detailed interpretation of the results and provides generic conclusions on application QoS.

The distributed resource controller (DRC) technology described in this paper provides a novel approach to interfacing applications with emerging network mechanisms to deliver quality of service (QoS) and controlling network resource utilization. DRC aims to unify network services (e.g., Diffserv, Intserv, and ATM) and application QoS provisioning by introducing a middleware system and a set of generic interfaces. DRC middleware is the core part that translates application requests for QoS delivery into respective access to underlying network systems in a manner that optimizes resource utilization and shelters applications from such a complexity. Applications can request QoS to the DRC middleware in two ways: either on-line using the DRC QoS API or off-line via the DRC utility. Application QoS requirements are specified in terms of two categories of parameters: traffic profile (quantitative) and user expectation (qualitative). DRC currently uses the CORBA-based object-oriented technology to develop a CORBA-based resource controller (CRC). The working CRC prototype built to manage the integrated services is also presented.

This paper describes novel market-based technologies that uniquely establish quantifiable and adjustable limits on the power of attackers, enable verifiable accountability for malicious attacks, and admit systematic and uniform monitoring and detection of attacks. These technologies, incorporated in the MarketNet system, establish a financial economy to regulate the trade and use of access rights in information systems. Resources are instrumented to use currency for access control and monitoring, establishing accountability in their use. Domains control access to their resources through resource prices and budgets available to clients. Domains control and fine tune their exposure to attacks; adjust this exposure in response to emerging risks; detect intrusion attacks through automated, uniform statistical analysis of currency flows; and tune their exposure to resource unavailability by purchasing protection through financial-like instruments.

Configuration management presently requires complex labor-intensive processes by experts. A single configuration task-installing/reconfiguring a system, or provisioning a service-typically involves a large number of activities fragmented among multiple network elements, each with its own proprietary configuration management instrumentation and tools. A change may cause configuration inconsistencies resulting in failures or inefficiencies; undoing changes to recover an operational state is often very difficult or even practically impossible. Therefore, configuration management is very costly, error prone, and often results in unpredictable failures and costly recovery. NESTOR seeks to replace labor-intensive configuration management with one that is automated and software-intensive. Configuration management is automated by policy scripts that access and manipulate respective network elements via a resource directory server (RDS). RDS provides a uniform object-relationship model of network resources and represents consistency in terms of constraints; it supports atomicity and recovery of configuration change transactions, and mechanisms to assure consistency through changes. RDS pushes configuration changes to network elements using a layer of adapters that translate operations on its object-relationship model to actions on respective elements. NESTOR has been implemented in two complementary versions and is now being applied to automate several configuration management scenarios of increasing complexity, with encouraging results.

Distributed multimedia applications are sensitive to the quality of service (QoS) delivered by underlying communication networks. The main question this work addresses is how to adapt multimedia applications to the QoS delivered by the network and vice versa. We introduce QoSockets, an extension to the sockets mechanism to enable QoS reservation and management. QoSockets automatically generates the instrumentation to monitor QoS. It scrutinizes interactions among applications and transport protocols and collects in QoS management information bases (MIB) statistics on the QoS delivered. The main advantages of QoSockets are the following: (1) support of single API for transport-layer QoS negotiation, connection establishment, and data transmission; and of single API for OS QoS negotiation; (2) support of a single QoS negotiation protocol; (3) generality across application QoS needs; (4) automatic management of application QoS needs. QoSockets are available for Solaris and Linux and support RSVP, ATM adaptation, ST-II, TCP/UDP, and Unix native protocols.

This paper introduces a novel approach to distributed computing based on delegation-agents, and describes its applications to decentralize network management. Delegation agents are programs that can be dispatched to remote processes, dynamically linked and executed under local or remote control. Unlike scripted agents, delegation agent programs may be written in arbitrary languages, interpreted or compiled. They can thus be more broadly applied to handle such tasks as real-time monitoring, analysis and control of network resources. Distributed management by delegation (MbD) uses this to manage remote elements and domains. MbD provides a paradigm for distributed, flexible, scalable and robust network management that overcomes the key limitations of current centralized management schemes.

This paper proposes scalable conflict-free media-access protocols with a novel architecture called parallel queueing for WDM optical passive star networks. The processing and hardware resource requirements are low and the control signalling demand is significantly reduced by using a small number of parallel queues in each station. A new scheme called wormhole scheduling is introduced to compensate the relatively slow laser-filter tuning speed. The protocol using wormhole scheduling decreases the overhead of tuning (thus increases the throughput) and relieves the electronic processing bottleneck. The key ideas are: (1) to associate several parallel queues with a data channel, rather that one deep queue (as in DT-WDMA) of N static queues, N is the number of stations; and (2) to schedule multiple successive packets which are destined to the same receiver at one time. An analytical model is presented and verified through simulation. Numerical results from analysis and simulation indicate that with only 5-6 parallel queues, the protocols are capable of achieving high throughput. Thus the protocols can be scaled up for networks with a large number of stations and could lead to potential implementation.>

An overview is given of a switching architecture for high-speed networks called Isochronets, which time-divide network bandwidth among routing trees. Traffic moves down a routing tree to the root during its time band. Network functions such as routing and flow control are entirely governed by band timers and require no processing of frame header bits. Frame motions need not be delayed for switch processing, allowing Isochronets to scale over a large spectrum of transmission speeds and support all-optical implementations. The network functions as a media-access layer that can support multiple framing protocols simultaneously, handled by higher layers at the periphery. Internetworking is reduced to a simple media-layer bridging. Isochronets provide flexible quality of service control and multicasting though allocation of bands to routing trees. They can be tuned to span a spectrum of performance behaviors outperforming both circuit and packet switching.>

A new approach to performing resource allocation in autonomous distributed computer systems is explored. As opposed to previous work which is based on interprocessor cooperation, the distributed system is modeled as a competitive society of microeconomic agents. The model is applied to the problem of managing distributed, replicated data objects. In this economy, jobs and transactions submitting read/write operations on data objects pay the processors to perform the operations. The processors use the revenue they earn to lease copies of data objects on which they can make a profit. The data objects themselves are active agents and set the prices for leasing copies. The economy is evaluated by a simulation study. The experiments show that the economy can substantially improve performance by varying the placement and number of copies of each data object.>

Rule interpreters usually start with an initial database and perform the inference procedure in cycles, ending with a final database. In a real time environment it is possible to receive updates to the initial database after the inference procedure has started or even after it has ended. We present an algorithm for incremental maintenance of the deductive database in the presence of such updates. Interestingly, the same algorithm is useful for parallel and distributed rule processing in the following sense. \\'hen the processors evaluating a program operate asynchronously. then they may have different views of the database. The incremental maintenance procedure we present can be used to synchronize these views.

The problem of managing large communication networks using statistical tests, alerts, and correlation among alerts is considered. The authors propose a model of these network management functions as data-manipulation operations. They argue that this approach can improve the flexibility of network management systems by providing a language that is declarative and set-oriented. These are properties of existing data-manipulation languages and it is shown that any data-manipulation language, augmented with several new capabilities, can serve as a language for specifying the aforementioned network management functions. The new capabilities required are specification of events, correlation among events, and change-tracking.>

A flow control mechanism based on an economic model of optimum decentralized decision making is presented. The virtual circuits (VCs) of the network are considered as economic agents competing for link capacity resources. Each VC is equipped with a preference relation defining its individual throughput-delay goals. The work makes three contributions to the study of flow control protocols. First, it presents fully decentralized flow control algorithms that allocate capacity to reflect both, varying priorities among VCs as well as their preferences and needs. Second, it establishes that the flow control algorithms result in Pareto-optimal allocation of capacities. This implies that the algorithms accomplish optimally fair allocation of resources given the set priorities and preferences. Third, experimental evidence is provided that the algorithms converge rapidly.>

A novel approach to allocating and sharing communication and computational resources in a distributed system is described. The approach, which is based on concepts drawn from microeconomics, uses algorithms that are competitive rather than cooperative. The effectiveness of these concepts is demonstrated by describing an economy that improves the performance of a distributed system by implementing load balancing. In this economy, competition sets prices for the resources in the system. Jobs complete for the resources by issuing bids, and the resource allocation decisions are made through auctions held by the processors. The benefits of the method include limited complexity and algorithms that are intrinsically decentralized and modular. Simulation studies show that these economies achieve substantial performance benefits.>

The authors examine the use of a group random-access protocol based on time windows for supporting time-constrained communication applications in a multiple-access network. First they formulate a policy for controlling protocol operation to minimize the percentage of messages with waiting times greater than some given bound. A semi-Markov decision model is then developed for protocol operation, and three of the four optimal control elements of this policy are determined. Although the semiMarkov decision model can also be used to obtain performance results, the procedure is to computationally expensive to be of practical use. Thus, an alternate performance model based on a queuing system with impatient customers is developed. Protocol performance under the optimal elements of the control policy shows significant improvements over cases in which the protocol is not controlled in this manner. Simulation results are presented to corroborate the analytic results.>

The purpose of this paper is to present an asymptotic performance analysis of a class of multihop radio networks using the calculus of residues. The analysis provides a closed-form approximation of global performance measures, whose precision increases exponentially with the size of the system. The complexity of computing these approximations is independent of the problem size. The method is easily applicable to a variety of topologies and access protocols.

This note proposes a statistical perturbation scheme to protect a statistical database against compromise. The proposed scheme can handle the security of numerical as well as nonnumerical sensitive fields. Furthermore, knowledge of some records in a database does not help to compromise unknown records. We use Chebyshev's inequality to analyze the trade-offs among the magnitude of the perturbations, the error incurred by statistical queries, and the size of the query set to which they apply. We show that if the statistician is given absolute error guarantees, then a compromise is possible, but the cost is made exponential in the size of the database.

For many time-constrained communication applications, such as packetized voice, a critical performance measure is the percentage of messages which are transmitted within a given amount of time after their arrival at a sending station. We examine the use of a group random access protocol based on time windows for achieving time-constrained communication in a multiple access environment. First, we formulate a policy for controlling protocol operation in order to minimize the percentage of messages with waiting times greater than some given bound. A semi-Markov decision model is then developed for protocol operation and three of the four optimal control elements of this policy are then determined.

A real-time protocol is usually concerned with the transportation of a real-time data stream over a packet switched network. Among the major issues distinguishing real-time protocols from ordinary transport protocols is the problem of trading delay for loss. That is, if some loss of packets may be acceptable, and usually inevitable, the objective of the protocol is to minimize the delay of packets, subject to constraints on the acceptable loss. This is unlike usual transport protocols, which are designed to guarantee no loss at the expense of increased delays.

The position-location problem is that of computing the coordinates of a set of objects in space (usually a plane) from a sparse set of distance measurements. Because the problem is analogous to that of constructing a pin-Jointed structure from rigid bars (of given respective lengths), it is intimately linked to problems of structural rigidity. In addition to its practical significance, the problem leads to a number of surprising results and intriguing theoretical problems in geometry, combinatorics, and algorithm design. This paper presents some of the theoretical algorithmic aspects of the position-location problem; its major objective is to attract researchers to complexity problems of structural rigidity. Among the major results presented is the discovery of a large class of geometrical decision problems, all of which are randomly decidable (i.e., decidable by a probabilistic polynomial-time algorithm), but many of which seem to be Intractable.

A method for studying the asymptotic behavior of discrete transformations is developed using numerical quadrature theory. This method allows a more convenient examination of the correlation properties of common unitary transforms for large block sizes. As a practical result of this method it is shown that the discrete cosine transform is asymptotically optimal for all finite-order Markov signals.