
Address
Description
The NSA Technology Transfer Program (TTP), located within the Research Directorate, establishes partnerships with industry, academia, and other government agencies. These partnerships help accelerate mission goals, advance science, foster innovation, and promote the growth and commercialization of technology originally created for Agency mission. Successful transfer of NSA technologies:
- Gives traction to the federal government's annual commitment to research and development, not totaling over $145 billion
- Mobilizes mission solutions by engaging NSA personnel with trusted technology partners
- Creates new companies, new job jobs, and new revenue
- Strengthens the economy, which in turn strengthens national security
For a full listing of our technologies, please download our full-color, PDF Patent Portfolio here.
Mission
The NSA Technology Transfer Program (TTP) mission is to:
- Establish partnerships to transfer NSA discoveries into commercial products and practical applications that strengthen national security and help grow the economy, and
- Stimulate the exchange of expertise and capabilities between NSA and industry, academia, and other government agencies to advance NSA mission, technology, and innovation.
Our mission fulfills federal technology transfer legislation, White House Cross-Agency Priority (CAP) goals, and NSA mission objectives.
Technology Disciplines
This technology provides three methods of identifying invalid digital signatures in a group of signatures that have failed a batch verification test. The first two methods offer significant increase in speeds for processing batches of pairing-based digital signatures. The third method improves the efficiency of previous “divide-and-conquer” methods and has applications to both pairing-based and non-pairing-based digital signatures. High-speed bulk processing of digital signatures commonly starts with group processing until a failure occurs, at which point the process slows until bad signatures are segregated from the remainder of the batch. These new methods speed up the step of segregating the bad signatures that caused the batch to fail, allowing faster processing of large batches.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent 7,890,763
Patent Number | 7,890,763 |
Patent Issue Date | Feb 15, 2011 |
US PATENT # 6,580,841 Chirped Fiber Acousto-Optic Bandpass Filter
US PATENT # 6,556,729 Acousto-Optic Bandpass Filter
This group of patented acousto-optic bandpass filters provide three varying methods of enabling one or more wavelength bands to be selected for further transmission. Signal switching is an important function in telecommunications. With the move to fiber optic cable transmissions, the demand for lower costs and higher switching speeds steadily increased, resulting in an increased need for optical switching. These technologies perform the function of bandpass filters at a lower cost, with lower power consumption, lower optical insertion loss, more versatility, and greater reliability than other bandpass filter technologies.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent 6,647,159; Patent 6,580,841; Patent 6,556,729
Patent Number | 6,647,159; 6,580,841; 6,556,729 |
Patent Issue Date | Nov 11, 2003 |
US Patent # 7,761,286 Natural Language Database Searching Using Morphological Query Term Expansion
US Patent # 7,539,611 Method of Identifying and Highlighting Text
This technology is a combination of three inventions: two natural language database search methods and a method of identifying and highlighting text. The Tagger searches for transliterated terms (such as names) within large data sets using lexical, contextual, and morphological information. The Matcher is then applied to determine which candidate terms are known within the database and are therefore of likely increased interest to the reader. This knowledge-based system’s ability to work without relying on any one query language grants non-expert users access to an extensive collection of linguistic information and expertise. A typical application for this technology is name searching since names have unique linguistic and cultural complications and extracting names from data sources can be difficult.
Patent Number | 7,797,152; 7,761,286; 7,539,611 |
Patent Issue Date | Sep 14, 2010 |
This technology enables all-digital power amplification suited for small, battery- operated devices. Utilizing a sleep mode for the switching amplifier prolongs the battery life and increases efficiency in power usage. Reduced electromagnetic interference (EMI), which can degrade performance in surrounding elements, allows the amplifier to be employed without additional shielding. Both of these qualities make the system easier to conceal and reduce the cost of production.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent 7,400,194
Patent Number | 7,400,194 |
Patent Issue Date | Jul 15, 2008 |
This patented method establishes a standard for translation metrics, offering analysis and feedback for grading and improving translation activities. When translating from a source language to a target language, word-for-word substitutions are not always adequate due to nuance, cultural variation, and native meaning. This technology provides user-definable ratings of accuracy, interpretation of intended message, and formatting, enabling effective language translation and interpretation of the source material. By not focusing on specific word choice, the technology is also able to more accurately assess the quality and accuracy translations and interpretations, taking concerns such as sentiment and content into account.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #8,185,373
Patent Number | 8,185,373 |
Patent Issue Date | May 22, 2012 |
This technology detects duplicate audio and voice recordings in a sample collection without relying on transcription or phonetic comparison for detection. This method is able to determine that two recordings of different lengths may be duplicates and two recordings of the same length may not be duplicates. This invention is independent of the recording’s language and content.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #7,571,093
Patent Number | 7,571,093 |
Patent Issue Date | Aug 04, 2009 |
(U) This technology yields an innovative approach to the search and discovery of a wide range of content and relationship information (ideas and concepts, as well as semantic, hierarchical, referential, and logical relationships) within a given corpus. Structured as a “knowledge system”, this process and system architecture will enable users to vary what types of information can be discovered by mapping documents into a reconfigurable data and knowledge model (DKM). The system also provides a single interface for users to discover concepts, documents, keywords, entities, and relationships, as well as to reason through logical content of documents by means of the rules knowledge base (RKB).
(U) Ultimately, the system enables machine-driven processing and automation of document structural and referential features by using a DKM to break documents into relevant features, such as titles, sections, section titles, authors, and references. Furthermore, the technology is intended to enhance automated reconstruction of documents, section-by-section, allowing users to view their full document textually, graphically, or by selecting specific document sections or tags.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #10,042,928
Patent Number | 10,042,928 |
Patent Issue Date | Dec 03, 2014 |
This language-agnostic technology automatically characterizes documents as handwritten, machine-printed, or unknown, enabling quick scoring and storing of images. The technology can be useful in maximizing the accuracy of optical character recognition (OCR) applications, resulting in the capability to more accurately search and sort large document collections.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #7,027,514
Patent Number | 7,072,514 |
Patent Issue Date | Jul 04, 2006 |
This technology automatically describes and categorizes the topic of text using nouns, allowing it to serve as a foundation for semantic interfacing with a knowledgebase. It has the advantage of not needing stop lists or extensive word stemming, which enables the tool to find words regardless of their form. Additionally, this technology provides results that are uniquely accurate by using a set list of nouns to concisely limit the search space and a probability-based measure of distance to categorize the topic.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #7,805,291
Patent Number | 7,805,291 |
Patent Issue Date | Sep 28, 2010 |
This technology eliminates the need to manually search audio files for speech content by automatically locating speech intervals that contain other signals such as music, noise, or even silence. It reports the start and stop times of these intervals relative to the beginning of the file, ignoring non-speech portions of the file. By classifying signal segments as speech or non-speech, this technology improves the performance of speaker recognition systems by reducing the amount of bandwidth and traffic to be analyzed.
Click to view PDFs of these patents on the US Patent and Trademark Office website: Patent #7,127,392
Patent Number | 7,127,392 |
Patent Issue Date | Oct 24, 2006 |