Patrick L. Smith

A branching algorithm for discriminating and tracking multiple objects

A recursive branching algorithm for multiple-object discrimination and tracking consists of a bank of parallel filters of the Kalman form, each of which estimates a trajectory associated with a certain selected measurement sequence. The measurement sequences processed by the algorithm are restricted to a tractable number by combining similar trajectory estimates, by excluding unlikely measurement/state associations, and by deleting unlikely trajectory estimates. The measurement sequence selection is accomplished by threshold tests based on the innovations sequence and state estimates of each filter. Numerical experiments performed using the algorithm illustrate how the accuracy of the a priori state estimates and trajectory model influences the selectivity of the algorithm.

Budget-constrained portfolio trades using multiobjective optimization

A decision support process using many-objective optimization, high-performance computing, and advanced visualization is applied to gain comprehensive insight into multistakeholder portfolio budgeting trades. The key tradeoffs among nondominated portfolio budget solutions are systematically identified and examined with respect to different stakeholder viewpoints. The approach is illustrated using a portfolio of 14 U.S. Air Force satellite development programs using budget data taken from the 2010 Future Year Development Plan [http://www.saffm.hq.af.mil/budget/, last accessed April 20, 2011]. Practical lessons learned in applying the approach are discussed. ©2012 Wiley Periodicals, Inc. Syst Eng 15

Future Space System Support to U.S. Military Operations in an Ice-Free Arctic: Broadband Satellite Communications Considerations

Significant increases in shipping traffic and resource exploration/exploitation activities are occurring in the Arctic region. The U.S. Navy, Coast Guard and other military services have begun to plan for increased operations in the region, which could start ramping up as early as about 2013 if recent climate model predictions for Arctic melting prove accurate. The polar region is a very different arena for space system operations compared to lower latitudes, and supporting increased U.S. military activities in the Arctic may require new satellite systems and user terminals. U.S. weather, navigation and surveillance satellites in polar orbits already cover the Arctic region, but current military satellite communications capabilities in the region are quite limited and existing passive satellite imagery (for ice monitoring and other types of surveillance) is hampered by persistent cloud cover and seasonal darkness. Given the high costs and long lead times (up to 10 years or more) to develop new space systems, the U.S. military could face a shortage of communications and possibly other space-based support capabilities during a period when nations are intensely jockeying for influence and resource claims in the region. Canada and Russia are developing new satellite programs to support their countries’ increased Arctic activities. The potential need for new space-based support capabilities in the Arctic region is certainly well within the U.S. military’s current planning horizon. This paper focuses on the military’s future requirements and options for broadband satellite communications in the Arctic region. Getting a head start on defining the requirements analysis and procurement options will help to expedite the space-system acquisition process when, perhaps in the very near future, the timeline for Arctic melting and increased military activities in the region becomes better established.

Charged-large-array-flexible mirror

A charged-large-array-flexible (CLAFLEX) mirror is a new concept for the deformable mirror. This mirror consists of a continuous piezoelectric transducer sheet whose thickness is controlled by one or more electron beams. A CLAFLEX mirror can be controlled using image-intensifier or video devices and digital image-processing techniques. A 100 × 100 array of deformable mirror elements can be controlled with state-of-the-art components, making short-wavelength high-power laser-beam steering, focusing, and phase conjugation practical. Two different CLAFLEX mirror control schemes are described. One uses a scanning electron beam; the other uses multiple parallel electron beams. Both designs use phase-lock control circuits that are synchronized by first phase locking a small central region of the mirror and then expanding this region (using an iris diaphragm) until the entire CLAFLEX mirror is phase-locked on the same wave front. A CLAFLEX mirror can be steered and focused optically by biasing the interference intensity distribution using an external light source. A wedge-shaped intensity bias, for example, will tilt the CLAFLEX mirror.

Phase control of HF chemical lasers for coherent optical recombination

A servo system for phase-locking two HF chemical lasers has been designed and simulated. A steady-state phase error is achieved that is adequate for coherent optical recombination. The results are based on the measured frequency drift of a small HF chemical laser and the measured frequency response of a piezoelectric transducer (PZT) mirror driver. A major innovation is the use of rate feedback with a laser Doppler sensor to extend the useful frequency response of the PZT driver.

Backward-forward smoothing interpretation of the a posteriori process noise estimate

The a posteriori discrete-time process noise estimate is interpreted as the combination of two independent estimates: the a priori estimate (which is white noise) and the optimum estimate based on the observations. Equivalence to previous formulations is demonstrated.

Estimation of solar limb brightening at radio wavelengths from solar eclipse measurements

The center-to-limb brightness distribution of the sun at radio wavelengths gives information about the structure of the solar atmosphere. Radio telescopes have poor spatial resolution, however, and cannot provide detailed angular measurements of the brightness distribution. During a solar eclipse, the sharp edge of the moon can be used to greatly increase the angular resolution of the radio telescope in one dimension. This paper describes the technique used to estimate the brightness distribution at 3.3 mm wavelength of the solar limb from measurements made during a total eclipse. The outer edge of the sun is divided into thin concentric bands and parameteric estimation methods are used to estimate the average flux desity of each band. This approach has several advantages for inverting radiometric data, especially in scanning problems.

Feasibility of Space-Based Monitoring for Governance of Solar Radiation Management Activities

Substantive research has begun into proposed schemes to synthetically increase the earth’s albedo (reflectivity) as a potential improvised measure to mitigate impacts of global warming if emission reductions are not sufficient or if the climate response is more extreme than anticipated. The authors of this paper do not take a position on whether Solar Radiation Management (SRM) should be used as a strategy to respond to climate change. However, future international agreements regarding development, testing, and implementation of SRM schemes will not be enforceable without effective means of monitoring and verification, especially since the relatively low cost of injecting reflective particles such as sulfur into the upper atmosphere will allow individual nations perhaps even private corporations or other groups to experiment on their own. This paper discusses monitoring requirements and the feasibility of space-based remote-sensing systems for detecting and monitoring particle injection into the upper atmosphere. Our preliminary findings suggest that detecting clandestine unilateral small-scale precursor particle-injection with satellite instruments may not be practical. This conclusion suggests that future treaty negotiations will need to consider alternative means of monitoring such activities.

Fitting Multistage Models to Input/Output Data

Publisher Summary This chapter discusses the fitting of multistage models to input /output data. In the present study, it is assumed that the objective of modeling a dynamic system is to predict or control the output of the system by observing or manipulating the inputs. The model is a digital computer program, which, when supplied with the measurements of the past and present input and output, computes the predicted future output of the system. The random nature of the problem is considered in developing the model, but the model is a deterministic system. System characterization and system identification are the principal aspects of modeling. System characterization is concerned with defining a class of mathematical models and system identification is concerned with the determination of the specific model belonging to this preselected class, which best fits the observations. The class of models examined in the present study is linear stationary multistage processes. The usefulness and convenience of linear models are well known and many techniques have been proposed to fit linear models to input/output data.