Machine learning training system for identification or classification of wireless signals
Inventors
Kleider, John • Baiense, Joao • Morgan, Chris
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11349743-B2
Publication Date
2022-05-31
Expiration Date
Abstract
A signal generator outputs a reference signal corresponding to at least one wireless signal according to the predefined signal encoding to a channel emulator processor. The channel emulator processor is programmed to use at least one synthesized channel parameter and the reference signal to produce and store a perturbed signal as data for training machine learning and artificial intelligence systems. The synthesized channel parameter is synthesized using a channel synthesizer processor programmed to: ingest map elevation data, reference a transmitter and a receiver to the map elevation data, and perform ray tracing of a representative signal between the transmitter and the receiver, while applying at least one predetermined perturbation property to synthesize at least one channel parameter.
Core Innovation
The disclosed invention provides an apparatus and method to train a machine learning system to identify and/or classify at least one wireless signal of the type having a predefined signal encoding scheme or predefined modulation encoding scheme. A signal generator outputs a reference signal corresponding to the at least one wireless signal according to the predefined encoding. The training is based on reference signals combined with synthesized channel behavior derived from environmental map data.
A channel synthesizer processor ingests map elevation data and references a transmitter and a receiver to the map elevation data. The channel synthesizer processor performs ray tracing of a representative signal between the transmitter and the receiver while applying at least one predetermined perturbation property to synthesize at least one synthesized channel parameter. The channel parameter synthesis is tied to the representative signal propagation using the ray tracing and the applied perturbation properties.
A channel emulator processor then uses the at least one synthesized channel parameter and the reference signal to produce and store a perturbed signal as data for training the machine learning system to identify and/or classify signals. In addition, the synthesized channel parameter output is supplied to a channel emulation processor programmed to receive an input signal and perturb the input signal using the at least one synthesized channel parameter. The resulting stored perturbed signal data is produced for use as training data.
Claims Coverage
The document includes three independent claims. Each independent claim centers on reference-signal generation, ray-tracing-based channel parameter synthesis from map elevation data with predetermined perturbation properties, and channel emulation to produce perturbed signals for machine-learning training or to perturb an input signal.
Reference signal generation from a predefined encoding
A signal generator generates a reference signal corresponding to the at least one wireless signal according to the predefined signal encoding or predefined modulation encoding scheme.
Ray tracing with elevation-map synthesis using predetermined perturbation properties
A channel synthesizer processor ingests map elevation data, references a transmitter and a receiver to the map elevation data, and performs ray tracing of a representative signal between the transmitter and the receiver while applying at least one predetermined perturbation property to synthesize at least one channel parameter.
Channel emulation that stores perturbed signals for machine-learning training
A channel emulator processor uses the at least one synthesized channel parameter and the reference signal to produce and store a perturbed signal as data for training the machine learning system.
Channel emulation processor perturbing an input signal using synthesized channel parameters
A channel emulation processor is programmed to receive an input signal and to perturb the input signal using the at least one synthesized channel parameter.
Across the independent claims, the inventive coverage is the combination of producing a reference signal from a predefined encoding, synthesizing channel parameters by ray tracing between transmitter and receiver referenced to map elevation data while applying predetermined perturbation properties, and emulating and perturbing signals using the synthesized channel parameters, with perturbed signal storage for machine-learning training in the apparatus and method claims.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?