SBIR-STTR Award

The accurate tracking of resident space objects (RSO)s depends on the rapid estimation of orbits using the knowledge gained from sparsely sampled observations of satellites under the influence of interacting gravitational and drag effects. Examples of scenarios operating within this environment include tasking follow up observations of debris created from collision events, accurately establishing the identity of objects that are located within close proximity, and reacting to controlled on-orbit deployments of additional space objects. New near real time and computationally efficient algorithms that can estimate non-Gaussian RSO error characteristics are available that could characterize RSO error to a much higher fidelity than current methods. For example, it has been shown that typical “banana-shaped” covariance profiles displaying more uncertainty along-track, than cross-track are reproducible with this technique. This type of information combined with orbital estimates provides more actionable space situational awareness (SSA) knowledge. Combined with an innovative space surveillance network (SSN) simulator that uses smart scheduling of assets in a flexible and responsive publish-and-subscribe network environment, these algorithms will be developed and tested for their applicability to improving the speed, accuracy and responsiveness of RSO tracking.

Benefit:
Currently, the SSN uses the NORAD SGP4 orbit models for predicting satellite positions that do not have the associated covariance estimates. PDS will provide a performance assessment of utilizing these innovative orbit estimation and RSO track association algorithms developed under this project by testing their accuracy and responsiveness of RSO tracking against realistic use cases generated with an innovative space surveillance network (SSN) simulator. Once these algorithms are validated under “real world” simulations, PDS will test and validate these algorithms with actual SSN data. PDS intends to work closely with the Air Force in transferring technology for their critical objectives. The primary DoD end-customer for these algorithms is the JFCC-Space through the Joint Space Operations Center (JSpOC), which detects, tracks, and identifies all man-made objects in Earth orbit. Through current program experiences, PDS understands the acquisition process involved in transitioning algorithms from concept to validation, development, testing, (SMC SSA Technology Branch) and deliverance of an operational product to the warfighter (AF Space Command).

Keywords:
Space Surveillance, Track Association, Orbit Estimation, Satellite Position, Identification, Classification

The accurate tracking of resident space objects (RSO)s depends on the rapid estimation of orbits using the knowledge gained from sparsely sampled observations of satellites under the influence of interacting gravitational, solar radiation pressure and atmospheric drag effects. While there are many established sequential estimators that can perform real-time orbit estimation and provide the associated covariance, the RSO tracking problem presents special difficulties. The current estimation technique tends to be applied with limited tracking data for a wide variety of orbit regimes when there is little or no information included in the estimation process on the RSO’s mass, shape, radiative properties, or attitude. In addition, it is likely that the uncertainty distribution for many RSOs is not Gaussian and cannot be represented accurately by a covariance matrix that has been developed with an assumed Gaussian probability density function. The AGSF algorithm developed under Phase I is designed to be scalable, relatively efficient for solutions of this type, and able to handle the nonlinear effects which are common in the estimation of RSO orbit states. In addition, information theoretic metrics in conjunction with AGSF were examined for data association purposes. The AGSF and corresponding observation association methods were evaluated using simulated data to determine their performance and feasibility. Combined with an innovative space surveillance network (SSN) simulator, these algorithms will be developed and tested for their applicability to improving the speed, accuracy and responsiveness of RSO tracking.

Benefit:
Currently, the SSN uses the NORAD SGP4 orbit models for predicting satellite positions that do not have the associated covariance estimates. PDS will provide a performance assessment of utilizing these innovative orbit estimation and RSO track association algorithms developed under this project by testing their accuracy and responsiveness of RSO tracking against realistic use cases generated with an innovative high fidelity space surveillance network (SSN) simulator. Once these algorithms are validated under “real world” simulations, PDS will test and validate these algorithms with actual SSN data. PDS intends to work closely with the Air Force in transferring technology for their critical objectives. The primary DoD end-customer for these algorithms is the JFCC-Space through the Joint Space Operations Center (JSpOC), which detects, tracks, and identifies all man-made objects in Earth orbit. Through current program experiences, PDS understands the acquisition process involved in transitioning algorithms from concept to validation, development, testing, (SMC SSA Technology Branch) and deliverance of an operational product to the warfighter (AF Space Command).

Keywords:
Space Surveillance, Satellite Tracking, Data Association, Covariance, Non-Linear Filters