|Acquisition Program: || Objective: ||Design, develop, and test an optimization co-processor to enable real-time resource allocation
|| Description: ||The long range challenge is to efficiently allocate complex computing tasks among small but powerful wireless mobile computing devices. Recent advances in computer technology are creating opportunities for more intelligent decisions in the field. While there is a long history of solving complex problems using the power of modern computing machinery, there is growing need to solve tasks involving information and decision processing in the field, close to where events are actually happening. Meanwhile, there is a growing number of small but surprisingly powerful computing devices becoming available. While these units can not individually compete with the power of a large central computer, they can communicate and work together.
An important goal is to put information processing closer both to the source of information as well as the people who need to use the information. Although wireless mobile computers are becoming increasingly powerful, they are small and even when aggregated they are not up the task of optimizing resource allocation, while performing the associated decision processing tasks.
By drawing on the collective capability of the multiple computers that are within communications range of each other, the challenge is that there is more than one sensor that may be collecting data and other computers may have their own sensor data to work on. But we can exploit the property that much of the data collected is redundant with previously collected data. When one computer encounters a burst of potentially important data, it might be possible to pre-empt a nearby computer, sending it data (and possibly algorithms) that contain potentially important information needing to be processed now!
Challenges for this topic include 1) developing architecture to enable real-time resource allocation that includes optimization, such as simplex, change point detection, together with adaptive filtering, to help in the identification of potentially important data 2) designing, developing, and testing an optimization co-processor.
The focus of this effort is to design, develop, and test a hardware co-processor to enable real-time resource allocation.
The OSD is interested in innovative R&D that involves technical risk. Proposed work should have technical and scientific merit. Creative solutions are encouraged.
|| ||PHASE I: Complete a feasibility study, research plan, and initial design that establishes the proof of principle of the approach for a real-time resource allocation co-processor that includes optimization, change point detection, and adaptive filtering, to optimize real-time resource allocation, while identifying critical technology issues that must be overcome to achieve success. Prepare a revised research plan for Phase 2 that addresses critical issues.
|| ||PHASE II: Refine design, develop, and test a prototype system that is capable of optimizing computational resources through the use of optimization, change point detection, and adaptive filtering. The prototype should lead to a demonstration of the capability. Test the prototype in at least two environments with two different contexts.
|| ||PHASE III: Produce a system capable of deployment in an operational setting of interest. Test the system in an operational setting in a stand-alone mode or as a component of larger system. The work should focus on capability required to achieve transition to program of record of one or more of the military services. The system should provide metrics for performance assessment.
|| References: ||1. J. Chen and A. Gupta, Parametric Statistical Change Point Analysis, Birkhauser, Boston, MA, 2000.
2. D. Pados and G. Karystinos, “An Iterative Algorithm for the Computation of the MVDR Filter,” IEEE Transactions on Signal Processing, Vol. 49, pp.290-300, February 2001.
3. J. A. Shapiro and W. B. Powell, “A Metastrategy for Large-scale Resource Management Based on Informational Decomposition, Informs Journal on Computing, Vol. 18, pp.43-60, 2006.
|Keywords: ||optimization, change point detection, adaptive filtering, real-time resource allocation|