Mark W. Maier

Integrated modeling: a unified approach to system engineering

Abstract This article describes an approach to system engineering and design called integrated modeling. Integrated modeling describes the process of system development as the iteration of models or “views” from the abstract to concrete and combines and links existing methods, including real-time structured analysis and design, metrics, performance modeling, and quality function deployment. The models represent and integrate four major system views: function, physical structure, end-to-end quantitative performance, and implementation or deployment. In principle, integrated modeling is applicable to general electromechanical and sociotechnical systems. However, this discussion is restricted to electronic sensor and control systems with hardware and software elements. The article presents models for each system view, describes methods for linking the views with a top-down flow of system objectives, and applies the method to a simplified radar warning receiver.

QUANTITATIVE ENGINEERING ANALYSIS WITH QFD

An important difficulty in bringing quality practices to early product design, where quality leverage is highest, is the mismatch between accepted engineering analysis methods and accepted quality tools. Quality Function Deployment (QFD) can serve as th..

Sedsat-1 Lessons Learned

ABSTRACT A number of universities have initiated small scientific satellite projects in which students take the primary roles of investigators, managers, designers, and fabricators. The University of Alabama in Huntsvilles SEDSAT-1 project is among them. SEDSAT-1 is manifested on the July, 1998 Delta II launch as a secondary payload (the JPL DS-1 mission is the primary). Although not yet launched, the SEDSAT-1 project has yielded numerous lessons learned for small satellite programs. Some of these, such as the critical importance of tightly defined and controlled mission scope, are well known. But others are more specific to university student satellites, such as the tradeoffs in cost and educational value of proto-flight versus qualification unit approaches. This paper discusses the primary lessons learned and methods suggested for mitigating the identified risks and problems. Included are the outlines of project templates, both in system architecture and organizational structure, to help guide future projects.

Synthetic aperture radar processing using a nonuniform pulse repetition interval

Due to recent development of the theory related to non- uniform Pulse Repetition Intervals (PRI) for pulse doppler radars, it has become feasible to combine the multiple pulse integration of Synthetic Aperture Radar (SAR) with the temporal agility of non-uniform PRIs. This combination can bring to the traditional SAR system the benefits of both strong resistance to countermeasures and interception, and high sensitivity. A significant problem which must be addressed in synthetic aperture radars is the possible smearing of the azimuth dimension of the image due to doppler sidelobes which are too high. This issue is even more serious with a non-uniform PRI waveform. In a traditional SAR the sidelobes are controlled with signal processing techniques such as amplitude weighting. However, these simpler techniques are not adequate for non-uniform PRI waveforms. One method for increasing the sidelobe rejection capability of these waveforms requires the use of a processor which consists of time varying linear digital filters. One way to design such a filter is through the use of interpolation based techniques.