Kaonyx Labs is a commercial SIGINT technologies provider. Their core technology was developed through NSA and CIA initiatives which combined nonlinear chaotic analysis with proprietary advanced signal processing methods to revolutionize the transmission, storage, analysis and management of signals. Kaonyx's tools and technology allow all signal types (regardless of source) to be treated within a consistent framework, whether a simple audio track, multi-dimensional sensor array data, or multi-layered video data.
Over the past decade, the nonlinear dynamics research team at University of New Hampshire headed by Dr. Kevin Short (founder of Kaonyx Labs) has developed and commercialized a number of applications relating to fundamental new advances in control of chaotic systems and signal processing. Initial applications included predicting and removing seismic sensor noise to detect nuclear tests, and breaking secure chaotic communication schemes under grants from US government agencies. These early projects demonstrated that more secure chaotic communication schemes could be developed, which led to the discovery of a new class of waveforms, named cupolets, which can be produced with few bits of information and yet are highly complicated in nature and can be used to develop extremely compressed representations of signals.
- Isolation, identification and decomposition of multi-layered image, audio or sensor signals.
- Significant reduction in storage requirements for simple and complex signals.
- Ability to scan and search stored multi-layered signals without decompressing and reconstituting.
- Optimization of channel bandwidth through “on-the-fly” alterable data-rates from stored and compressed files.
This relatively new class of waveforms, cupolets, is derived by controlling nonlinear chaotic systems that normally exhibit almost-random behavior, so that they stabilize onto complicated periodic motions that have many practical applications. The name is derived as chaotic unstable periodic orbit-lets and the mathematical breakthrough means that a single chaotic system, if implemented in simple circuitry, can be given 16-bit controls and will produce a vast number of periodic signals, such as might be used in communications, spread spectrum, or signal representations. It has been shown that these signals are similar to those of musical instruments, human voice, image and video signals, and even turbulence signals when viewed in a series of short timeframes.
The fundamental discovery of cupolets supported with focused research and development efforts led to successful commercialization of the core technology, including:
- The chaotic compression technology was applied to audio signals, producing high quality stereo music files 4x more compressed than typical MP3 files and led to the creation of Chaoticom Inc., (later renamed Groove Mobile). The company developed the first real-time, full-track music downloads to cell phones and portable devices throughout Europe and North America.
- The advanced signal processing techniques were applied to the problem of signal restoration, where the signal analysis was accurate enough to detect the presence of “indicator signals” in recordings that were related to imperfections in the original recording equipment and had a negative impact on the intelligibility or clarity of the recording. The most notable result was the restoration of a 1949 wire recording that resulted in a Grammy Award for Kaonyx Labs’ founder Dr. Short and others on the restoration team. Additionally, the restoration extracted spoken “storytelling” tracks that were originally unintelligible, but became perfectly clear after processing.
In order to take advantage of the cupolet creation technology, it was necessary to develop highly accurate, advanced signal processing techniques. The techniques that have proven to be most useful are based on a family of methods that use the spectral phase evolution of a signal to achieve super-resolution of frequencies. These Complex Spectral Phase Evolution (CSPE) methods represent a new approach to a wide range of signal processing problems. The work is an outgrowth of techniques developed in early collaboration with colleagues at the National Security Agency. The following suite of techniques can be applied to signals from any source and can be used for detection, analysis, compression, reconstruction and compressive sensing:
- Short-time stable (quasi-stable) Sinusoids.
- Linearly Modulating Signals
- Modulating Groups of Harmonics
- Correction of Near-DC aliasing for extremely low frequency signals
- Detection of Signals above the Nyquist rate
- Multi-channel/Multi-sensor processing in a Unified Domain
- New Hilbert Transform with improved accuracy for AM, FM and mixed signals
- Compression and Reconstruction of signals - speech, audio, image, video
- Signal Restoration – degraded environments, scratch and pop repair, pitch and modulation correction
- Transient representation and compression
- New representation of phase information in Harmonic Group Coordinates
- Analysis and modeling of turbulence data
Kaonyx Labs 3 Cormorant Circle Durham, NH 03824-3339 (603) 868-3487 Manta.com estimates one employee