Edward T. Chow

Biological information signal processor

The computation requirements for mapping and sequencing the human genome might soon exceed the capability of any existing supercomputer. The systolic array processor presented in this paper, called biological information signal processor (BISP), has the capability to satisfy the current and anticipated future computational requirements for performing sequence comparisons based on the T.F. Smith and M.S. Waterman algorithm (1981) as extended by M.S. Waterman and M. Eggert (1987). The BISP can conduct the most time consuming sequence comparison functions, establishing both global and local relationships between two sequences. A modified Smith and Waterman algorithm is presented in this paper for efficient VLSI implementation. Methods are developed to reduce the BISP systolic array I/O bandwidth problem by reporting only the statistical significant results. Estimated performance of the BISP is compared with several different computer architectures.<<ETX>>

A systolic array processor for biological information signal processing

The Biological Information Signal Processing (BISP) is a system for high speed sequence comparisons designed to support the computation requirements for mapping and sequencing the human and other genomes. The heart of a BISP system is a versatile processor chip that can conduct the most time consuming sequence comparison functions, establishing both global and local relationships between two DNA or protein sequences. Because of the application’s strong computation and communication requirements, a programmable systolic array architecture was developed. A BISP system can include a large number of processing elements; the initial BISP demonstration system consists of 768 BISP elements, capable of delivering more than 6.25 x 10^9 integer operations per second. The system can be expanded to include over 4,000 elements, This paper describes the comparison algorithm and outlines the BISP chip and system designs. Estimated performance of the BISP system is compared with several different computer architectures.

A high-speed message-driven communication architecture

The performance of a message-passing multiple instruction multiple data (MIMD) concurrent computer depends in large part on the communication processing overhead. A high-speed communication architecture is proposed for a hypercube-type supercomputer to attain the specific goals of message-driven processing. These goals include: direct hardware execution of messages, queueing of messages (using various paradigms), adaptive message routing, and special local registers for fast context switching. A hyperswitch network capable of high-speed switching is used to perform message routing. Available communication paths are pruned by the routing algorithm in response to congestion within the network to evaluate the best-first available path between any two nodes. Simulation results show that this architecture reduces message reception latency by two to three orders of magnitude when compared to commercial hypercube systems. This communication architecture extends the application of parallel systems to supercomputer problems that place heavy demand on the communication network for high bandwidth, low latency, and non-local communication. Detailed simulation studies for the proposed communication architecture are presented.

IsoWAN: A NASA science and engineering information and services framework

We believe that the next evolutionary step in supporting wide-area application and services delivery to customers is a network framework that provides for collocation of applications and services at distinct sites in the network, an interconnection between these sites that is performance optimized for these applications, and value-added services for applications. We use the term IsoWAN to describe an advanced, isolated network interconnect services framework that will enable applications to be more secure, and able to access and be in use in both local and remote environments. The main functions of an IsoWAN are virtual localization of application services, an application service interface, coordinated delivery of applications and associated data to the customer, and supporting collaborative application development for customers. An initial pilot network between three NASA Centers; Ames Research Center, the Jet Propulsion Laboratory, and Marshall Space Flight Center, has been built and its properties are discussed.

An Intelligent Network for Federated Testing of NetCentric Systems

We will describe a dynamic federated autonomic networking system1 for the testing of netcentric systems across organizations. Using a suite of policy-based management software tools, our system can provide netcentric missions with self-configuring, self-debugging, self-healing, and self-protecting capabilities across end-to-end coalition networks. The novelty of our system originates from: (1) A multi-party dynamic policy negotiation algorithm and protocol for autonomous cross-organization negotiations on network resources; (2) A RDF/XML based language to enable automated understanding of network and application configurations across organizations; (3) A teachable, natural language policy capture system; (4) An advanced user interface tool for capturing and representing domain expert knowledge; and (5) A reasoning system for distributing and applying domain knowledge. This system has been developed for large-scale netcentric systems testing applications where the rapid configuration and management of dynamically changing netcentric test exercises is critical. Using a scenario from a real test exercise involving test ranges around the country, we have demonstrated in our testbed that we could reduce the debugging time for a network configuration problem from 6 hours, involving dozens of operators, to less than 2 seconds with minimal human involvement.

Role of Policy Based Enterprise Management in GIG NetOps

The Department of Defense (DoD) global information grid (GIG) is a globally networked information infrastructure comprised of heterogeneous information transport networks, computing/data centers, enterprise services and applications, as well as end-user systems and devices. The net-centric operations (NetOps) mission areas span the operation and defense of the GIG. Real-time management and operations of the highly complex, dynamic, and diverse GIG requires a significant level of automation. Policy based enterprise management (PBEM) is an approach for introducing standards based automation of GIG NetOps incrementally. This paper describes PBEM concepts and use case scenarios developed for the Defense Information Systems Agency (DISA) NetOps that are extendable to GIG NetOps. A number of policy based enterprise management (PBEM) related activities are underway in DISA and DoD. One of the goals of these efforts is to collaborate and leverage lessons learned, as well potentially conduct joint demonstrations, pilots and field trials.

Applying Policy Based Enterprise Management towards realization of DISN Real-Time Services

Real-Time Services (RTS) will rely on distributed capabilities to enable end-to-end management of IP based voice and video services across multiple domains. Policy Based Enterprise Management (PBEM) will be a key RTS enabler by coordinating and automating management processes among domains. This will facilitate effective distributed management of RTS services across the enterprise. For example, PBEM will execute Commanderpsilas Intent (CI) by translating it into sets of event triggered conditions and actions that adjust bandwidth, traffic priority and traffic admission controls. This paper discusses how PBEM enables agile network operations by providing real-time management responses to high impact events. It will also address the relevance of Policy Based Management (PBM) in RTS and the end-to-end management capability to be supported by PBEM in support of the RTS Pilot. Pilot scenarios are proposed that include: focused overload, Commanderpsilas Intent (CI) and RTS bandwidth reallocation, failure mitigation, and synchronizing policies across multiple management domains. A framework for a PBEM test-bed that will mature the capabilities for a RTS pilot targeted for FY09-FY11 time frame is provided. Based on the scenarios described in this paper, operational metrics are identified that will allow the benefits of PBEM to be evaluated in a RTS test-bed activity and during the RTS pilot.

A system to provide real-time collaborative situational awareness by web enabling a distributed sensor network

The paper presents two systems called PATS and SAP that when integrated realize Sensor Web Enablement (SWE) of spatially distributed mobile sensors. The Personal Alert and Tracking System (PATS) consists of a networked collection of custom-designed low-power wireless nodes, arranged in ad-hoc network topologies, to provide tracking for wild land firefighters. These mobile nodes form arbitrary network topologies and use a multi-hop packet routing protocol to relay sensor data to the command center. The multi-hop capability enables robust communication in a variety of environments by routing around natural and man-made terrain features. Situational Awareness and Prediction (SAP) works with the PATS sensor network to convert sensor data to information and to provide real-time collaborative situational awareness. The goal is to deliver a resource utilizing intelligent reasoning coupled with rule-based actionable intelligence using diverse knowledge fusion and modal trend forecasting. The SAP makes this data available to information sharing middleware using OGC standards. The paper describes the architecture of both the PATS and SAP systems and how these two systems interoperate with each other. The SAP system works in concert with the Unified Incident Command and Decision Support (UICDS) information sharing middleware to provide data fusion from multiple sources. UICDS can then publish the sensor data using the OGCs Web Mapping Service, Web Feature Service, and Sensor Observation Service standards. The system described in the paper is able to integrate a spatially distributed sensor system, operating without the benefit of the Web infrastructure, with a remote monitoring and control system that is equipped to take advantage of SWE.

Secured Advanced Federated Environment (SAFE): a NASA solution for secure cross-organization collaboration

This paper discusses the challenges and security issues inherent in building complex cross-organizational hierarchy and inter-organizational federation, the Secured Advanced Federated Environment (SAFE) is laying the foundation for a collaborative virtual infrastructure for the NASA community. A key element of SAFE is the Micro Security Domain (MSD) concept, which balances the need to collaborate and the need to enforce enterprise and local security rules. With the SAFE approach, security is an integral component of enterprise software and network design, not an afterthought.

Dynamic communication resource negotiations

Todays advanced network management systems can automate many aspects of the tactical networking operations within a military domain. However, automation of joint and coalition tactical networking across multiple domains remains challenging. Due to potentially conflicting goals and priorities, human agreement is often required before implementation into the network operations. This is further complicated by incompatible network management systems and security policies, rendering it difficult to implement automatic network management, thus requiring manual human intervention to the communication protocols used at various network routers and endpoints. This process of manual human intervention is tedious, error-prone, and slow. In order to facilitate a better solution, we are pursuing a technology which makes network management automated, reliable, and fast. Automating the negotiation of the common network communication parameters between different parties is the subject of this paper. We present the technology that enables inter-force dynamic communication resource negotiations to enable ad-hoc inter-operation in the field between force domains, without pre-planning. It also will enable a dynamic response to changing conditions within the area of operations. Our solution enables the rapid blending of intra-domain policies so that the forces involved are able to inter-operate effectively without overwhelming each others networks with in-appropriate or unwarranted traffic. It will evaluate the policy rules and configuration data for each of the domains, then generate a compatible inter-domain policy and configuration that will update the gateway systems between the two domains.