You are here

Advanced Hierarchical Temporal Memory (HTM) Algorithm

Description:

TECHNOLOGY AREA(S): Info Systems 

OBJECTIVE: Develop an innovative HTM algorithm to pre-process raw radar return data. 

DESCRIPTION: This topic seeks to develop innovative applications of the open source HTM algorithm to pre-process raw radar returns in order to reduce debris, clutter, and other unwanted return data which should lead to reduced radar loading and an increase in engagement capability. By pre-processing the returns, erroneous signals would not be forwarded to the computer suite for processing. This should allow the weapon system to process additional targets since it is not spending as much time processing clutter data. 

PHASE I: Develop a proof of concept design. Identify designs and conduct feasibility assessment for the proposed solution. Perform an analysis to demonstrate the concept and an understanding of the new and innovative technology. 

PHASE II: Develop and refine the proposed solution based on the results and findings of the Phase I. Validate the feasibility of the Phase I concept by development and demonstrations that will be tested to ensure performance objectives are met. Validation would include, but not be limited to, system simulations, operation in test-beds, or operation in a demonstration subsystem. The Phase II effort should result in a prototype with substantial commercialization potential. 

PHASE III: Based on Phase II lessons learned, revise the systems model to prove out the new design. Develop and execute a Phase III incremental test & integration plan that will produce a final prototype. 

REFERENCES: 

1: https://numenta.com/machine-intelligence-technology/

2:  https://numenta.org/resources/HTM_CorticalLearningAlgorithms.pdf

KEYWORDS: Hierarchical Temporal Memory, Algorithm, Radar Returns, Radar Loading 

US Flag An Official Website of the United States Government