# US 11,444,702 B2 · Transmitter identification based on machine learning > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,444,702 B2 - Document ID: US11444702B2 - Status: Granted - Kind: B2 - Application number: 16/836,122 - Filed: 2020-03-31 - Published: 2021-09-30 - Granted: 2022-09-13 - Earliest claimed: 2020-03-31 - Assignees: International Business Machines Corporation - Applicants: International Business Machines Corporation - Inventors: Bodhisatwa Sadhu; Supriyo Chakraborty - Web page: https://patentgrep.com/patent/US11444702B2 ## CPC - G06N 3/0675: {using electro-optical, acousto-optical or opto-electronic means} - H04B 17/104: {of other parameters, e.g. DC offset, delay or propagation times} - G06N 3/045: Combinations of networks - G06N 3/0464: Convolutional networks [CNN, ConvNet] - G06N 3/08: Learning methods - G06N 3/09: Supervised learning - H04B 7/15521: {combining by calculations packets received from different stations before transmitting the combined packets as part of network coding (network coding aspects for detection or prevention of errors in the information received H04L1/0076; network traffic management with optimizing of information sizing, e.g. header compression, by using assembly and disassembly of packets H04W28/065)} - H04B 17/12: of transmit antennas, e.g. of the amplitude or phase - G06N 3/044: Recurrent networks, e.g. Hopfield networks - G06N 20/00: Machine learning - H04B 17/10: of transmitters - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - H04B: TRANSMISSION ## Abstract Systems and methods for identifying a device are described. In an example, a processor can receive a first signal having a first waveform encoding data. The processor can extract the data from the first signal. The processor can determine a transformation being used to encode the data in the first waveform. The processor can generate a second signal using the determined transformation to encode the data in a second waveform. The processor can determine a difference between the first waveform and the second waveform. The processor can identify a device as a candidate device that sent the first signal, based on the determined difference. ## Claims ### Claim 1 A method comprising: receiving a first signal having a first waveform encoding data representing a bit vector; demodulating the first signal to extract the bit vector from the first signal; determining a transformation being used to encode the data in the first waveform; generating a second signal using the determined transformation to encode the extracted bit vector in a second waveform, wherein the generated second signal is an attempt to reconstruct the first signal; determining a difference between the first waveform and the second waveform; and identifying a device as a candidate device that sent the first signal, based on the determined difference. ### Claim 2 (depends on 1) The method claim 1, wherein the difference indicates at least one non-ideality associated with hardware components of the device. ### Claim 3 (depends on 1) The method of claim 1, wherein identifying the device comprises, running a machine learning model using the determined difference to output a probability indicating a likelihood that the first signal was transmitted by the device. ### Claim 4 (depends on 3) The method of claim 3, in response to an absence of an identification of the device in the output from the machine learning model, assigning a new signature to the device using the determined difference. ### Claim 5 (depends on 1) The method of claim 1, further comprising training a machine learning model based on attention machine learning by initializing attention parameters using the difference. ### Claim 6 (depends on 1) The method of claim 1, wherein the difference indicates at least one of: non-ideality associated with hardware components of the device; non-ideality associated with hardware components of a device receiving the first signal; and at least one error caused by the transmission of the first signal over a wireless channel. ### Claim 7 (depends on 1) The method of claim 1, further comprising applying a delay to the first signal, and determining the difference comprises determining a difference between the delayed version of the first signal and the second signal. ### Claim 8 A system comprising: a receiver configured to receive a first signal from a transmitting device, the first signal having a first waveform encoding data representing a bit vector; a device configured to be in communication with the receiver, the device being configured to: demodulate the first signal to extract the bit vector from the first signal; determine a transformation being used to encode the data in the first waveform; generate a second signal using the determined transformation to encode the extracted bit vector in a second waveform, wherein the generated second signal is an attempt to reconstruct the first signal; determine a difference between the first waveform and the second waveform; and identify the transmitting device based on the determined difference. ### Claim 9 (depends on 8) The system of claim 8, wherein the difference indicates at least one non-ideality associated with hardware components of the transmitting device. ### Claim 10 (depends on 8) The system of claim 8, wherein the device is configured to run a machine learning model using the determined difference to output a probability indicating a likelihood that the first signal was transmitted by the transmitting device. ### Claim 11 (depends on 10) The system of claim 10, in response to an absence of an identification of the transmitting device in the output from the machine learning model, assigning a new signature to the transmitting device using the determined difference. ### Claim 12 (depends on 8) The system of claim 8, wherein the device is configured to train a machine learning model based on attention machine learning by initializing attention parameters using the difference. ### Claim 13 (depends on 8) The system of claim 8, wherein the difference indicates at least one of: non-ideality associated with hardware components of the transmitting device; non-ideality associated with hardware components of the receiver; and at least one error caused by the transmission of the first signal over a wireless channel. ### Claim 14 (depends on 8) The system of claim 8, wherein the device is configured to apply a delay to the first signal, wherein the determination of the difference comprises a determination of a difference between the delayed version of the first signal and the second signal. ### Claim 15 A computer program product for identifying a device, the computer program product comprising a non-transitory computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processing element of a device to cause the device to: receive a first signal having a first waveform encoding data representing a bit vector; demodulate the first signal to extract the bit vector from the first signal; determine a transformation being used to encode the data in the first waveform; generate a second signal using the determined transformation to encode the extracted bit vector in a second waveform, wherein the generated second signal is an attempt to reconstruct the first signal; determine a difference between the first waveform and the second waveform; and identify a device as a candidate device that sent the first signal, based on the determined difference. ### Claim 16 (depends on 15) The computer program product claim 15, wherein the difference indicates at least one non-ideality associated with hardware components of the device. ### Claim 17 (depends on 15) The computer program product claim 15, wherein the program instructions is executable by the processing element of the device to cause the device to identify the device by running a machine learning model using the determined difference to output a probability indicating a likelihood that the first signal was transmitted by the device. ### Claim 18 (depends on 15) The computer program product claim 15, wherein the program instructions is executable by the processing element of the device to cause the device to train a machine learning model based on attention machine learning by initializing attention parameters using the difference. ### Claim 19 (depends on 15) The computer program product claim 15, wherein the difference indicates at least one of: non-ideality associated with hardware components of the device; non-ideality associated with hardware components of a device receiving the first signal; and at least one error caused by the transmission of the first signal over a wireless channel. ### Claim 20 (depends on 15) The computer program product claim 15, wherein the program instructions is executable by the processing element of the device to cause the device to apply a delay to the first signal, and the determination of the difference comprises determining a difference between the delayed version of the first signal and the second signal.