SBIR-STTR Award

We propose to investigate and design an interactive and practical environment called NVision for modeling, simulation, dynamic reconfiguration, evaluation and visualization of software and system architectures for performance and reliability. NVision will be capable of modeling, simulating, documenting and visualizing the behavior of distributed dynamic systems. NVision will provide interfaces to popular tracing tools for commonly used platforms in order to collect, analyze and import performance data. This will be used to construct software performance models using a popularly used framework and populate model data. In addition to a variety of performance-metric collection and reporting, the simulation engine will support interactive time-based debugging and system inspection, to facilitate insight into the operation of complex protocols and behavior. Standard and programmable plug-in modules will aggregate and propagate key reliability and performance parameters through hierarchy elements for easy visualization and navigation. The standard model can be augmented using a powerful plug-in framework allowing custom definition of (a) components using visual catalogs, (b) component properties using customizable property sheets, and (c) component behavior using an application programming interface in terms of resource-centric operations. NVision will provide a user-friendly environment for independent, hierarchical design, and interactive binding of hardware and software aspects of the system, aiding performance-sizing design decisions. NVision will also offer standard and customizable visualization of results using charts and reports. Benefit In U.S. defense systems, fly-by-wire avionics, process control, nuclear power control, automated manufacturing facilities, air traffic control and medical systems, life-threatening situations can arise due to software performance problems or components failures. The design of such performance-critical large-scale defense and commercial systems is a complex process, requiring assembly of software and hardware components with disparate and custom behavior. Errors caught late in the process are relatively expensive to address. The NVision environment will facilitate the disciplined integration of COTS components in the design of such dynamic real-time and dependable systems. NVision will result in significant cost reduction with the design, analysis and simulation of complex component behavior and interactions early in the process. Keywords Openness and Extensibility, Schedulability analysis, Timing and reliability analysis, Visual design tool, , Hierarchical visualization, reconfigurability, COTS component integration, Component repository

We propose to enhance and apply the nVision toolset to a chosen subset of the Navy’s DD(X) C2I subsystem. The focus of the Phase 2 project is to develop technology, integrate and demonstrate nVision within the DD(X) project environment, thus directly demonstrating the benefits and scalability of the approach in the design, implementation and maintenance phases of the DD(X) program. We will leverage the innovative nVision analysis, simulation, diagramming and user interface capabilities that were previously developed. For the base Phase 2 effort, we shall use a two step approach: first, we will demonstrate (via analysis and simulation) the evaluation of one chosen Technical Performance Measure (TPM) in the current DD(X) design release, using reconstructed, enhanced and sanitized data files; second, we will demonstrate the repeatability of the process to one or more chosen TPMs. The nVision toolset will be enhanced to use and augment the same project files that are being created and used by the DD(X) design team, performing data import and export out of the nVision environment. The Phase 2 Option task will develop necessary technology and demonstrate the applicability of the nVision toolset to run-time implementation artifacts, thus mitigating the risk of design-time and implementation-time performance mismatches. Phase 2 Option will utilize and demonstrate roundtrip data import from the run-time DD(X) C2I workflow generators, such as those developed as part of the DARPA ARMS program.

Keywords:
Performance Modeling, Timing Analysis, Discrete Event Simulation, Real-Time Architecture, Uml Data Extraction, Software Producibility Tool, Real-Time