Devices and System

Abstract

A device which includes a microphone configured to detect a noise from a surrounding area of the device; an electronic circuit coupled to the microphone, wherein the electronic circuit comprises (i) at least one low-pass filter configured to filter the detected noise and (ii) a frequency-to-voltage converter configured to convert a frequency of the filtered detected noise into a voltage; a machine learning unit comprising a processor and a memory, wherein the processor is configured to analyze the detected noise using a machine learning algorithm stored in the memory; and an alarm unit configured to receive, from the machine learning unit, information on the analyzed detected noise and to receive, from the electronic circuit, the voltage.

What is claimed is: 
     
         1 . A device comprising:
 two matched antennas configured to transmit and receive ultra-wideband, UWB, signals,   an electronic circuit, coupled to the matched antennas, the electronic circuit comprising a microprocessor configured to process the received UWB signals and to measure a channel impulse response, CIR, between a transmitter channel and a receiver channel of the electronic circuit, the microprocessor further configured to convert a time between a transmission of the UWB signals and a receipt of the UWB signals into a distance, and the electronic circuit being further configured to detect a target and/or estimate a distance from the device to the target and/or track the target and/or track a distance of the target from the device and/or track a velocity of the target based on the processed UWB signals; and   an alarm unit configured to receive, from the machine learning unit, information on the detected target, the information comprising a measurement of CIR confidence intervals and an adaptive threshold.   
     
     
         2 . The device as claimed in  claim 1 , further comprising a machine learning unit coupled to the electronic circuit, wherein the machine learning unit comprises a processor and a memory, wherein the machine learning unit is configured to detect the target, classify the target, and declare an alarm based on the received UWB signals, the machine learning unit further comprising a machine learning algorithm stored in the memory, wherein the machine learning algorithm is configured to detect of the target, classify the target and the declare the alarm. 
     
     
         3 . The device as claimed in  claim 1  or  2 , wherein at least one of the two matched antennas comprises a composition configured to enable a required bandwidth and optimize impedance-matching for an obstacle in a near field, wherein the bandwidth enabling and the impedance-matching optimization comprises the at least one of the two matched antennas having a radiation pattern comprising a gain towards the obstacle in the near field, and side lobe rejection. 
     
     
         4 . The device as claimed in  claim 3 , wherein the bandwidth enabling and the impedance-matching optimization of at least one of the two matched antennas further comprises at least one of:
 (i) a reduction in influence of the obstacle in the near field;   (ii) a reduction of a reflection of the obstacle;   (iii) a support of a large ultra-wideband, UWB, bandwidth, and   (iv) antenna gain towards the obstacle, and reduced directivity away from the obstacle.   
     
     
         5 . The device as claimed in any one of the preceding claims, wherein an effective field of view of the device is up to 180° and/or wherein the electronic circuit and/or the machine learning unit is configured to measure the CIR based on the received UWB signals and/or wherein a transmitter and receiver is configured to transmit and/or receive one or more pseudo-random noise, PRN, codes. 
     
     
         6 . The device as claimed in any one of preceding claims, further comprising a 3-dimensional system configured to estimate a height of the detected target based on information received from the transmitter channel and/or the receiver channel of the electronic circuit. 
     
     
         7 . The device as claimed in any one of the preceding claims, when dependent on  claim 2 , wherein the machine learning unit is configured to determine a position and/or detected target size and/or direction of movement and/or velocity of the detected target, and/or metal object being carried by the detected target, based on the processed UWB signals. 
     
     
         8 . The device as claimed in any one of the preceding claims, wherein the electronic circuit and/or the machine learning unit is configured to process:
 (i) a plurality of reference CIRs obtained previously in time, wherein the reference CIRs are aligned in time, and   (ii) a test CIR measured at a current point in time.   
     
     
         9 . The device as claimed in  claim 8 , wherein the electronic circuit and/or machine learning unit is configured to compare one or more of the plurality of reference CIRs and the test CIR, and wherein the machine learning unit is configured to identify a target based on a difference between the compared one or more of the plurality of reference CIRs and the test CIR. 
     
     
         10 . The device as claimed in  claim 9 , when dependent on  claim 2 , wherein the machine learning unit is configured to confirm a detection of a target when at least one of:
 (i) the test CIR differs in magnitude and/or phase, wherein the magnitude and phase comprise a real and an imaginary part, from the one or more of the plurality of reference CIRs;   (ii) a test CIR is outside a confidence interval for a certain distance, wherein the adaptive threshold is configured to be changed when more reference CIRs are collected; and   (iii) a cross correlation of one or more of the plurality of reference CIRs and the test CIR is below a certain threshold.   
     
     
         11 . The device as claimed in any one of  claims 8  to  10 , wherein the detected target is configured to be tracked by the device, and wherein a velocity of the detected target is configured to be estimated by the device, by the electronic circuit and/or the machine learning unit, wherein the device is configured to evaluate a distance estimate over time and track a phase change of a plurality of different parts of the test CIR. 
     
     
         12 . The device of any one of  claims 8  to  11 , wherein upon a difference is detected the electronic circuit and/or machine learning unit in the test CIR in comparison to one or more of the plurality of reference CIRs, the target is configured to be detected, and a distance between the device and the target is configured to be determined by a position of the difference in the test CIR and/or if the difference no longer exists, the electronic circuit and/or machine learning unit is configured to determine that the target is blocking the transmitter channel and/or the receiver channel. 
     
     
         13 . A system comprising the device of any one of the preceding claims, and a further device comprising a microphone configured to detect a noise from a surrounding area of the further device, the further device comprising;
 a second electronic circuit coupled to the microphone, wherein the second electronic circuit comprises (i) at least one low-pass filter configured to filter the detected noise and (ii) a frequency-to-voltage converter configured to convert a frequency of the filtered detected noise into a voltage;   a second machine learning unit comprising a second processor and a second memory, wherein the second processor is configured to analyze the detected noise using a second machine learning algorithm stored in the second memory; and   a second alarm unit configured to receive, from the second machine learning unit, information on the analyzed detected noise and to receive, from the second electronic circuit, the voltage.   
     
     
         14 . A device comprising:
 matched antennas designed specifically to penetrate various surfaces and to transmit and receive signals through structures,   an electronic circuit, coupled to the matched antennas, wherein the microprocessor configured to process the received UWB signals and to measure a channel impulse response (CIR) between the transmitter channel and the receiver channel to convert time into distance,   a machine learning unit coupled to the electronic circuit, wherein the machine learning unit comprises a processor and a memory, wherein the processor is configured for detection of target, classification of target, calibration of the system, and alarm declaration based on the received UWB signals using a machine learning algorithm stored in the memory;   and, an alarm unit configured to receive, from the machine learning unit, information on the identified object, and to receive from the electronic circuit measures of CIR, confidence intervals and adaptive threshold.   
     
     
         15 . A device comprising:
 a transmitter configured to transmit ultra-wideband, UWB, signals   a receiver configured to receive UWB signals;   an electronic circuit coupled to the receiver, wherein the electronic circuit comprises a signal processing unit configured to process the received UWB signals;   a machine learning unit coupled to the electronic circuit, wherein the machine learning unit comprises a processor and a memory, wherein the processor is configured to identify an object based on the processed UWB signals using a machine learning algorithm stored in the memory; and   an alarm unit configured to receive, from the machine learning unit, information on the identified object,   wherein the electronic circuit and/or the machine learning unit are configured to measure a channel impulse response, CIR, based on the received UWB signals.