# US 10,122,479 B2 · Systems, methods, and devices for automatic signal detection with temporal feature extraction within a spectrum > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,122,479 B2 - Document ID: US10122479B2 - Status: Granted - Kind: B2 - Application number: 15/412,982 - Filed: 2017-01-23 - Published: 2018-07-26 - Granted: 2018-11-06 - Earliest claimed: 2017-01-23 - Assignees: DGS Global Systems, Inc. - Applicants: DGS Global Systems, Inc. - Inventors: Ronald C. Dzierwa - Web page: https://patentgrep.com/patent/US10122479B2 ## CPC - H04B 17/30: of propagation channels - G06N 20/00: Machine learning - H04B 17/391: Modelling the propagation channel - H04B: TRANSMISSION - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract Systems, methods and apparatus for automatic signal detection with temporal feature extraction in an RF environment are disclosed. An apparatus learns the RF environment in a predetermined period based on statistical learning techniques, thereby creating learning data. A knowledge map is formed based on the learning data. The apparatus automatically extracts temporal features of the RF environment from the knowledge map. A real-time spectral sweep is scrubbed against the knowledge map. The apparatus is operable to detect a signal in the RF environment, which has a low power level or is a narrowband signal buried in a wideband signal, and which cannot be identified otherwise. ## Claims ### Claim 1 A method for automatic signal detection in a radio-frequency (RF) environment, comprising: learning the RF environment in a period of time to a settled percent of at least 99.95% based on statistical learning techniques, thereby creating learning data including power level measurements of the RF environment; indexing the power level measurements for each frequency interval in a spectrum section in the period of time; forming a knowledge map of the RF environment based on the power level measurements of the RF environment; automatically extracting at least one temporal feature of the RF environment from the knowledge map; scrubbing a real-time spectral sweep against the knowledge map; calculating a first derivative of the power level measurements and a second derivative of the power level measurements; selecting most prominent derivatives of the first derivative and the second derivative; performing a squaring function on the most prominent derivatives; detecting at least one signal in the RF environment based on matched positive and negative gradients; averaging the real-time spectral sweep, removing areas identified by the matched positive and negative gradients, and connecting points between removed areas to determine a baseline; subtracting the baseline from the real-time spectral sweep to reveal the at least one signal; and wherein one or more of the at least one signal is a narrowband signal hidden in a wideband signal. ### Claim 2 (depends on 1) The method of claim 1, wherein the knowledge map comprises an array of normal distributions, wherein each normal distribution corresponds to how often a power level at each frequency has been at a particular level. ### Claim 3 (depends on 1) The method of claim 1, further comprising creating a profile of the RF environment based on the knowledge map, wherein the profile comprises a highest power level at each frequency during the predetermined period of time. ### Claim 4 (depends on 1) The method of claim 1, wherein the at least one signal is transmitted from a remote signal emitting device and has a low power level. ### Claim 5 (depends on 1) The method of claim 1, wherein the narrowband signal has a bandwidth ranging from 1 kHz to 60 kHz and is inside the wideband signal across a spectrum up to about 6 GHz. ### Claim 6 (depends on 1) The method of claim 1, further comprising automatically fine-tuning a threshold of power level on a segmented basis while extracting the at least one temporal feature from the knowledge map. ### Claim 7 (depends on 1) The method of claim 1, wherein a frequency resolution of the knowledge map is based on a Fast Fourier Transform (FFT) size setting. ### Claim 8 (depends on 1) The method of claim 1, further comprising periodically reevaluating the RF environment and updating the knowledge map. ### Claim 9 (depends on 1) The method of claim 1, further comprising automatically recording relevant information in high definition when the at least one signal is detected. ### Claim 10 (depends on 1) The method of claim 1, further comprising sending a notification and/or an alarm to an operator after detecting the at least one signal. ### Claim 11 A system for automatic signal detection in a radio-frequency (RF) environment, comprising: at least one apparatus for detecting signals in the RF environment; wherein the at least one apparatus is operable to sweep and learn the RF environment in a period of time to a settled percent of at least 99.95% based on statistical learning techniques, thereby creating learning data including power level measurements of the RF environment; wherein the at least one apparatus is operable to index the power level measurements for each frequency interval in a spectrum section in the period of time; wherein the at least one apparatus is operable to form a knowledge map based on the power level measurements of the RF environment; wherein the at least one apparatus is operable to automatically extract at least one temporal feature of the RF environment from the knowledge map; wherein the at least one apparatus is operable to scrub a real-time spectral sweep against the knowledge map; wherein the at least one apparatus is operable to calculate a first derivative of the power level measurements and a second derivative of the power level measurements; wherein the at least one apparatus is operable to select most prominent derivatives of the first derivative and the second derivative; wherein the at least one apparatus is operable to perform a squaring function on the most prominent derivatives; wherein the at least one apparatus is operable to detect at least one signal in the RF environment based on matched positive and negative gradients; wherein the at least one apparatus is operable to average the real-time spectral sweep, remove areas identified by the matched positive and negative gradients, and connect points between removed areas to determine a baseline; wherein the at least one apparatus is operable to subtract the baseline from the real-time spectral sweep to reveal the at least one signal; and wherein one or more of the at least one signal is a narrowband signal hidden in a wideband signal. ### Claim 12 (depends on 11) The system of claim 11, further comprising a remote device in network-based communication with the at least one apparatus, wherein the knowledge map and detecting results are displayed on a remote device in real time. ### Claim 13 (depends on 11) The system of claim 11, wherein the knowledge map comprises an array of normal distributions, wherein each normal distribution corresponds to how often a power level at each frequency has been at a particular level. ### Claim 14 (depends on 11) The system of claim 11, wherein the apparatus is operable to create a profile of the RF environment based on the knowledge map, wherein the profile comprises a highest power level at each frequency during the period of time. ### Claim 15 (depends on 11) The system of claim 11, wherein the apparatus is operable to send a notification and/or an alarm to an operator after detecting the at least one signal. ### Claim 16 An apparatus for detecting at least one signal in a radio-frequency (RF) environment, comprising: at least one processor coupled with at least one memory, and at least one sensor; wherein the apparatus is operable to sweep and learn the RF environment in a period of time based on statistical learning techniques, thereby creating learning data including power level measurements of the RF environment; wherein the apparatus is operable to index the power level measurements for each frequency interval in a spectrum section in the period of time; wherein the apparatus is operable to form a knowledge map of the RF environment based on the power level measurements of the RF environment; wherein the apparatus is operable to automatically extract at least one temporal feature of the RF environment from the knowledge map; wherein the apparatus is operable to scrub a real-time spectral sweep against the knowledge map; wherein the apparatus is operable to calculate a first derivative of the power level measurements and a second derivative of the power level measurements; wherein the apparatus is operable to select most prominent derivatives of the first derivative and the second derivative; wherein the apparatus is operable to perform a squaring function on the most prominent derivatives; wherein the apparatus is operable to detect at least one signal in the RF environment based on matched positive and negative gradients; wherein the apparatus is operable to average the real-time spectral sweep, remove areas identified by the matched positive and negative gradients, and connect points between removed areas to determine a baseline; wherein the apparatus is operable to subtract the baseline from the real-time spectral sweep to reveal the at least one signal; and wherein one or more of the at least one signal is a narrowband signal hidden in a wideband signal. ### Claim 17 (depends on 16) The apparatus of claim 16, wherein at least one knowledge map is stored in the apparatus. ### Claim 18 (depends on 16) The apparatus of claim 16, wherein the apparatus is operable to obtain a different knowledge map by communicating with another apparatus. ### Claim 19 (depends on 16) The apparatus of claim 16, wherein the apparatus is automatic and unmanned. ### Claim 20 (depends on 16) The apparatus of claim 16, wherein the apparatus is water resistant.