TW201909038A - Methods and systems for electromagnetic near-field coherent sensing - Google Patents

Abstract

The present disclosure may be embodied as methods and/or systems for non-contact measuring of an on-body and/or inside-body motion of an individual. A sensing signal is provided within a near-field coupling range of a motion to be measured. In this way, a measurement signal may be generated as the sensing signal modulated by the motion. The sensing signal may be an ID-modulated signal. In some embodiments, the sensing signal is a backscattered RFID link provided a wireless tag. A downlink signal may be provided to power the wireless tag. The sensing signal may be a harmonic of the downlink signal. The measurement signal is detected. The motion is measured based on the measurement signal. The measurement signal may be detected as far-field radiation after transmission through a source of the motion. The measurement signal may be detected as reflected from a source of the motion as antenna reflection.

Description

Method and system for electromagnetic near-field coherent sensing

The present invention relates to the detection of motion and, in particular, the detection of vital signs without physical contact.

Monitoring of vital signs, such as, for example, heart rate, blood pressure, respiratory rate, and respiratory function, is a key procedure for patient management and pathology recording. Current practices based on body electrodes, optical absorption, pressure or strain gauges, and ultrasonic or radio frequency (RF) backscattering have respective limitations on sensing capabilities and sampling rates. The measurement procedure can also be due to direct skin contact or motion constraints that render the subject uncomfortable or disrupt the individual's circadian rhythm. This inconvenience significantly increases the workload of the caregiver and hinders continuous long-term monitoring.

Due to the strong reflection of the body surface and the geometric mean of the reflected signals, conventional techniques for radio frequency ("RF") based far-field vital sign detection can easily pick up respiratory motion, but such techniques are difficult Distinguish small mechanical vibration details such as heartbeats with low RF frequencies and wrist pulse waveforms. Although heart rate can be captured after careful filtering, the estimation of blood pressure and simultaneous monitoring of multiple free-moving people are still not possible.

It is still long-awaited for low-invasive techniques for detecting signs of life.

The present invention provides a method of RF near field coherence sensing (NCS) without direct skin contact. The method of the present invention can be used to directly modulate the instantaneous mechanical motion on and/or within the body to a multiplexed radio that can contain a unique identification (ID). Two exemplary embodiments are described herein to provide flexibility in deployment and operation: passive tags and active tags near the heart and wrist. To reduce deployment and maintenance costs, passive RFID tags can be integrated into garments at the chest and wrist regions, where two multiplexed far-field backscatter waveforms can be collected at the reader to capture heart rate, respiration rate, and breathing Breath effort and blood pressure. To improve reading range and immunity to indoor sports, active tags can be placed, for example, in a front pocket and a wrist cuff to measure antenna reflections due to NCS, and then fully sample and transmit life in digital format. Signals to eliminate indoor multipath interference. The vital sign monitoring system disclosed by the present invention can be used for multiple individuals at the same time and can bring cost-effective automation in the care facility. In addition, direct skin contact and elimination of motion constraints enhance patient comfort, which enables long-term monitoring of improved pathology analysis.

Respiratory, heartbeat, and wrist pulses can be monitored to derive respiration rate, heart rate, and blood pressure by using one of the NCS embodiments of an exemplary harmonic RFID system for vital sign monitoring. In some embodiments, spectral equalization can be applied to the NCS signal to recover an accurate heartbeat time interval in the time domain.

Sensing antenna impedance matching effects are shown for NCS, with a tissue matching antenna being shown to provide improved NCS signal quality and waveform detail. Analyze the performance of the vital signs waveform and discuss the design strategy.

In another embodiment, the high frequency component of a heartbeat signal can be used to mitigate interference from body movements to heart rate estimates. Based on ECG, NCS is shown to be sufficiently accurate for immediate heart rate and heart rate variability for normal body movements.

In another aspect, the invention can be embodied in a method for non-contact measurement of one of the body's movements on the body and/or inside the body. A first radio frequency ("RF") sensing signal is provided within a near field coupling range of one of the first motions to be measured. In this way, a first measurement signal can be generated as the first sensing signal modulated by the first motion. The first sensing signal can be an ID modulation signal. The first sensing signal can be an active radio link. The first sensed signal can be a backscattered RFID link. The first sensing signal can be provided by a wireless tag. A downlink signal can be provided to charge the wireless tag. The first sensed signal can be one of the harmonics of the downlink signal.

Detecting the first measurement signal. The first motion is measured based on the first measurement signal. The first measurement signal can be detected as far-field radiation transmitted through a source of the first motion. The first measurement signal can be detected as being reflected from one of the first motion sources to an antenna reflection. Measuring the first motion can further include filtering the first measurement signal to obtain a first motion signal through the timing and the waveform.

In some embodiments, the method further includes providing a second RF sensing signal within a near field coupling range of one of the second motions to be measured. In this way, a second measurement signal can be generated as the second sensing signal modulated by the second motion. Detecting the second measurement signal. The second motion is measured based on the second measurement signal. The first motion and the second motion may be measured based on the synchronization measurement.

In another aspect, a system for measuring motion of a body is provided. The system includes a first signal source for generating a first sensed signal. A first antenna is in electrical communication with the first signal source. The first sensing signal can be an ID modulated wave. The first sensing signal can be an active radio link. The first sensed signal can be a backscattered RFID link. The first antenna is configured to be disposed within a near field coupling range of one of the first motions to be measured. The first antenna can be configured to be placed within a coupling range of a heart motion, a pulse, a respiratory motion, an intestinal motion, an eye motion, and the like. In this way, a first measurement signal is generated as the first sensing signal modulated by the first motion. The system includes a first receiver for detecting a first measurement signal. The first receiver can be configured to detect the first measurement signal as a transmission signal. The first receiver can be configured to detect the first measurement signal as a reflected signal.

The first signal source and the first antenna can be configured as a wireless tag. A tag reader is configured to transmit a downlink signal to the wireless tag. The receiver can be part of the tag reader. The wireless tag can be configured to be charged by the downlink signal. The first sensed signal can have a frequency that is one of the harmonics of one of the downlink signals. For example, the first sensed signal can be the second harmonic of one of the downlink signals. The wireless tag can modulate the downlink signal using an orthogonal ID such that the first sensing signal is a CDMA signal.

In some embodiments, the system can further include a second signal source for generating a second sensed signal. A second antenna is in electrical communication with the second signal source. The second antenna is configured to be disposed in a near field coupling range of one of the second motions to be measured to generate a second measurement signal as the second sensing by the second motion modulation signal. The receiver can be further configured to detect the second measurement signal.

In some embodiments, a filter is in communication with the receiver. The filter can be configured to demodulate and filter the first and/or second measurement signals to obtain a corresponding first and/or second motion signal. The filter can be, for example, programmed to sample, demodulate, and filter the first and/or second measurement signals to derive a motion processor. In some embodiments, a processor is programmed to measure a derivative value based on the detected coupled signal and the second coupled signal.

Statement on federal sponsorship research

This invention was made with government support under Contract No. DE-AR0000528 awarded by the Department of Energy. The government has certain rights to the invention. Cross-reference to the same application

The present application claims priority to U.S. Provisional Application No. 62/ 521, 352, filed on Jun. 16, s. The disclosures of these are incorporated herein by reference.

Signs of life are important not only for pathological research, but also for wearable devices, which are widely used to infer behavior, mood, and health. Although many such devices are well established and widely used, current devices have the disadvantage of limiting their sensing accuracy or long-term convenience. The present invention can be embodied in a method for Near Field Coherent Sensing ("NCS") that modulates the movement of a body to a radio frequency ("RF") signal, which can be a multiplexed RF signal. The movement of a body can include, for example, movements related to vital signs, such as heartbeat, pulse, breathing, and the like. Embodiments of the method of the present invention can directly modulate the mechanical motion on or within the body of an individual to an RF signal in the near field range. The motion can be modulated to a multiplexed harmonic RF identification ("RFID") backscatter signal using a unique digital identification ("ID").

The "near field" of an antenna is a region in which the dominant characteristic of the inductive property is higher than the radiative characteristic and the relationship between the electric field (E field) and the magnetic field (H field) is not well defined. In an embodiment of the invention, "near field" may refer to an area of proximity of an antenna, wherein the angular field of view distribution depends on the distance from the antenna. In an embodiment, the near field extends to a region within one wavelength (λ) of the antenna. In other embodiments, the near field extends to an area within λ/2, λ/3, λ/4, or λ/2π of the antenna, where the operating wavelength of the λ-based antenna. Other embodiments will be apparent to those skilled in the art from this <RTIgt;

Referring to Figure 21, the present invention can be embodied in a method 100 for non-contact measurement of body motion (e.g., physical or internal body motion) of a body. An individual can be, for example, a human or a non-human animal. The detected motion may be, for example, a heart motion, a pulse, a respiratory motion, a bowel motion, an eye motion, or other body motion as will be appreciated in light of the present invention. Embodiments of the method 100 of the present invention can directly modulate the mechanical motion on or within the body of an individual to a multiplexed radio signal integrated with a unique digital ID. A first radio frequency ("RF") sensing signal is provided 103 in a near field coupling range of one of the first motions to be measured to generate a measurement signal. The first sensing signal provided 103 can be an ID modulation signal. In some embodiments, the first sensing signal is an active radio link. In some embodiments, the first sensed signal is a backscattered RFID link. For example, an antenna can transmit a beacon or ID modulation sensing signal in an active radio link or a backscattered RFID (Radio Identification) link. The first sensing signal is modulated by the first motion, thereby generating a first measurement signal. The method 100 includes detecting 106 a first measurement signal. In some embodiments, the detection 106 can be done at a far field, such as detecting a first measurement signal transmitted through the body of the individual. In some embodiments, the detection 106 has a reflected signal, such as using a near field antenna.

The first motion is measured 109 based on the first measurement signal. As mentioned above, in the NCS, more energy is directed into the body tissue than in the prior art, thus implicitly amplifying backscatter signals from the viscera. Moreover, the shorter wavelengths in the body tissue present a small mechanical motion as a relatively large phase change. The shorter wavelengths within the individual's body naturally increase the signal-to-noise ratio ("SNR"). The differential nature of the signals within the body isolates large surface movements. This can also increase the sensitivity to achieve a measurement of a weak motion signal, such as, for example, a wrist pulse. Since the internal mechanical motion modulation gives a differential signal similar to one of the interferometers, the common signal caused by the external movement can be easily lowered by filtering (see, for example, FIG. 2A). With one of the antennas in the near-field coupling range of mechanical motion inside the body, the propagating or reflected waves can be easily detected in a homogenous manner and will contain instantaneous geometric mean information of the mechanical motion. Motion can be measured by filtering 112 the first measurement signal to obtain a motion signal. In the case of an ID modulated wave, multiple mechanical motions can be simultaneously read in a synchronized manner. Multiplex techniques can be used in passive backscatter or active radio transmission to facilitate simultaneous sensing at multiple points and/or for multiple people. NCS offers new opportunities for comfort, convenience and low cost of vital signs monitoring.

It should be noted that the NCS technique disclosed herein directly measures mechanical motion within or on the body, rather than indirectly by electrical signals such as sensing electrical motion or mechanical signals induced by mechanical motion. Therefore, the NCS technology of the present invention can provide more information than a conventional electrocardiogram. For example, by using a second label on the wrist or neck region where a pulse can be felt, a waveform difference from a heart label can be used to obtain an accurate estimate of one of the blood pressures, and this can be done for multiple people in the room. Without ambiguity. Using the techniques of the present invention, the clinical field can be managed in a completely new way: the owner wearing the tag(s) according to the present invention can monitor their ID, location, heart rate, respiration rate, blood pressure, and the like. In addition, long-term monitoring of a body can be accomplished since no skin contact is required (e.g., compared to the ECG pad(s)).

In some embodiments, the method 100 can further include providing 115 a second RF sensing signal within a near field coupling range of one of the second motions to be measured. In this manner, the second motion is coupled to the second RF sense signal to produce a second measurement signal. The second measurement signal is detected 118 and the second motion is measured 121 based on the second measurement signal. The second motion can be measured by filtering 124 the second measurement signal to obtain a second motion signal. The value can be determined based on the motion of the synchronized measurement and the second motion. For example, in the case where the first motor is a heartbeat (measured near the chest) and the second motor is a pulse (measured near the wrist), the derivative value may be based on one of the heartbeat and pulse determinations 127.

In another aspect, the invention can be embodied in a system 10 for measuring the motion of a body (see, for example, Figure 40). System 10 includes a first signal source 12 for generating a first sensed signal. A first antenna 14 is in electrical communication with the first signal source 12. The first antenna 14 is configured to be placed within a near field coupling range of one of the first motions to be measured. For example, the first antenna 14 can be configured to be placed within a coupling range of a heart motion, a pulse, a respiratory motion, a bowel motion, an eye motion, and the like. In this manner, a first measurement signal is generated from the first sensed signal modulated by the first motion. The first sensing signal can be an ID modulated wave. For example, the EM wave can be an active radio link or a backscattered RFID link.

The system includes a receiver 16 for detecting a first measurement signal (a first sense signal coupled to the first motion (modulated by the first motion)). Receiver 16 can be configured to detect the first measurement signal as a transmission signal, ie, far field radiation. The receiver can be configured to detect the first measurement signal as a reflected signal, ie, the antenna reflection. The system can include a filter in communication with the receiver, wherein the filter is configured to demodulate and filter the first measurement signal to obtain a motion signal. The filter may be, for example, a processor (such as a digital signal processor ("DSP")) that is programmed to sample, demodulate, and/or filter the first measurement signal to derive a motion signal.

In some embodiments, a system 10 can include a second signal source 22 for generating a second sensed signal. In these embodiments, a second antenna 24 is in electrical communication with the second signal source 22. The second antenna 24 is configured to be placed within a near field coupling range of one of the second motions to be measured. In this way, a second measurement signal can be generated as the second sensing signal modulated by the second motion. In a particular example, the first motion is a heartbeat and the second motion is a pulse. In this example, the first antenna can be configured to be placed close to the chest of one body, and the second antenna is configured to be placed close to the wrist of the individual. Receiver 16 is further configured to detect a second measurement signal. System 10 can include a processor 30 for measuring a derivative value based on the detected first measurement signal and the second measurement signal. In a particular instance of a heartbeat and pulse, the derivative value can be, for example, one of the individual's blood pressures.

In some embodiments, a wireless tag (such as a passive (i.e., without a local power source such as a battery) RFID tag) can be integrated into the garment in the vicinity of the area in which the vital signs are to be measured. These RFID tags can provide an NCS implementation with low deployment and maintenance costs. These RFID tags provide an ID modulated signal in which one of the unique IDs of each tag helps distinguish between its signal and interference from other tags and surrounding signals. The label backscatter signal is then processed using spectral equalization to amplify the high frequency components to restore not only the waveform detail initially immersed in the low frequency components, but also the sharp peaks of the precise heartbeat time interval with improved peak detection certainty. The derived heartbeat time interval exhibits improved stability compared to synchronous ECG. Principle of operation

Embodiments of the NCS utilize a near field coupling of an EM field to a mechanical motion within the body or on the body surface. Electromagnetic simulation was performed using CST Microwave Studio to illustrate the operating principles of NCS. As shown in FIGS. 1A and 1B, an EM simulation of a male body (1A) and a left lower arm (1B) is constructed. Based on Zubal Phantom's EM simulation model, Zubal Phantom has a stereo pixel resolution of 3.6 × 3.6 × 3.6 mm ³ and is established by magnetic resonance imaging ("MRI") and computed tomography ("CT"). Each voxel is indicated by the 3D coordinates of the body tissue along with the index. After the CST merges into the body model, the Visual Basic for Applications ("VBA") macro language is used to map EM properties using the CST biobank. A finite integration technique ("FIT") is then used to include the RFID tag antenna near the chest area. As can be seen from the simulation, a large amount of RF energy is coupled to the inside of the body due to near field coupling. Due to the high dielectric constant of the human tissue, the wavelength is correspondingly shortened, which further increases the NCS sensitivity.

The NCS method of the present invention uses both the amplitude and phase of the electromagnetic field. Since the phase is very sensitive to the distance between the RF source and the receiver, the external chest movement when a person breathes can be evaluated by the phase. Respiratory rate can be easily retrieved and phase changes can be used to further explain breathing operations. Compared to phase information, the amplitude of the electromagnetic field is less sensitive to small distance changes, which means that breathing or other external body movements will change phase rather than amplitude, providing good isolation of other signals within the body to be properly sensed. In NCS, the interference measurement class structure converts visceral/tissue movement into amplitude modulation of the RF signal.

For our simulation (Fig. 2A), when the human phantom is facing the receiver (Fig. 4A), the antenna on the chest transmits an RF carrier, wherein the antenna characteristics are defined by the local near field region. Depending on the antenna directivity, the portion of the RF energy will be emitted directly towards the receiver, while the other portion will be coupled to the inside of the body due to the near field effect. Intuitively, it is contemplated that the backscattered RF signal from the heart is modulated by mechanical movement of the heart tissue and then interfered with direct emission, resulting in a change in amplitude. From the analogy of the interference meter, the movement inside the body is a "differential mode" modulation, and the movement of the body surface is a "common mode" modulation.

The recording motion can be reflected not only from the far field but also from the antenna shown by the scattering parameter S ₁₁ in Fig. 3B. Using antenna reflection, an NCS signal can be recorded directly using a mobile device and is therefore more immune to body movements and indoor multipath problems in a crowded room. Due to tissue within the near field region of an antenna using NCS movement operation, it will affect the changes in the geometry of the antenna reflector S _11, where the antenna may be considered part of the sensor. The vital signs will be modulated on the S ₁₁ parameter of the antenna and thus captured by the reflected signal.

In the case where the transmission antenna is close to the skin, an NCS device can modulate the motion signal (life sign) through a radio signal. However, conventional microwave transmitters consume significant power in local oscillators and power amplifiers, and such transmitters may require a battery for one of the mobile devices. In addition, synchronization between the body transmitter and the far field receiver will also complicate the system design. In some embodiments, the NCS can be implemented using a passive harmonic RF identification (RFID) tag in which the vital signal is modulated on the harmonic backscatter along with the tag ID. In addition to ultra-low cost, the simple and robust packaging of passive tags also enables direct fabric integration with ready-to-wear garments. An example of an RFID sensor tag wafer integrated with an embroidered antenna on a fabric is shown in Figure 9A. The benefits of harmonic backscattering over conventional RFID are outlined in Figures 8A and 8B. Due to the high transmission power of conventional RFID readers and phase noise edges, self-leakage, antenna reflection and backscatter from unintended surrounding objects contribute to noise and severely degrade the SNR of the backscattered tag signal. However, harmonic backscattering can use a large frequency separation to isolate the downlink (reader to tag) from the uplink (tag to reader), which increases both SNR and sensitivity. The tag remains a passive backscatterer that can easily follow current RF protocols. A schematic diagram of one of the harmonic tags is shown in Figure 5A (photograph of one of the printed circuit board (PCB) prototypes is shown in Figure 9B). The harmonic tag receives a downlink RF signal from the reader under f, which passes through tag antenna 1 (ANT 1) and is split into two parts. One portion provides DC power to the tag circuit by energy harvesting, and another portion is fed to passive harmonic generation at 2f to re-transmit from antenna 2 (ANT 2), which acts as an NCS transmitter. The RF switch in front of the harmonic generator can modulate the digital information by turning on the key control (OOK), similar to conventional RFID operation. The digital information can include the tag ID and additional information from the sensor on the tag.

A schematic diagram of one exemplary harmonic RFID reader as one of the coherent transceivers is shown in FIG. 5B. The same digital clock (dashed line) is fed into two frequency synthesizers at f and 2f for homomorphic demodulation at 2f. The digital module implements the CDMA protocol. The downlink command from the reader to the tag is modulated by a digital to analog converter ("DAC") and then upconverted by the mixer to a carrier at f. The harmonic label is backscattered to the reader at 2f, and f is downconverted to the baseband by a coherent local oscillator at 2f and sampled by a quadrature analog-to-digital converter ("ADC"). The hardware of one of the harmonic readers is tested using a software defined radio (SDR). Analysis of NCS signals

The phase pair is more sensitive to the position of the tag entity relative to the reader. Thus, when a tagged antenna 2 is placed on the chest of a body, respiratory information can be derived from the phase in the orthogonal scheme, as shown in Figure 6A using raw and low pass filtered waveforms. Based on the backscatter phase information, the position of multiple tags can be calculated by millimeter resolution, which can further derive the breathing operation. Although the phase change caused by chest movement is much stronger than the internal movement of the heartbeat and wrist pulse, it is one of the "common components" of the NCS of tissue motion within the body (as further described above). During an experiment performed using an embodiment of the present system, the NCS heartbeat signal is immune to movement caused by the individual's breathing. The use of multiple frequencies, improved signal processing, and reflective structures (Figs. 12A and 12B) can further mitigate severe multipath interference. First, the heart rate in Fig. 6B is taken from the instantaneous period (solid line) and from the count within 10 seconds (dashed line). Star markers were measured from a commercial blood pressure monitor (OMRON BP760N). It should be noted that the breathing and heartbeat information is derived independently from quadrature demodulation without the need for special filtering or pattern recognition as is known in conventional microwave backscattering.

Due to the internal vital signs extracted from the NCS, an interferometer-like structure significantly increases the sensitivity to achieve the collection of motion waveforms, similar to a one-heart impact map ("BCG"). Also record the information from the chest and wrist tags for 3 minutes. Test with a PCB tag and the reader antenna is ~1.5 m to 2 m from the test subject. The harmonic signal converted by the tag is about -20 dBm at 1.9 GHz (2f). To analyze the waveform changes, the cycles are superimposed to obtain the average of the heartbeat and wrist pulse waveforms as shown in Figures 6C and 6D and the whisker bias. The waveform is normalized to the 90th percentile of the recorded data. Dynamic Time Warping ("DTW") is applied to classify waveforms to derive detailed features. Figures 6E and 6F show the DTW distances of the heart and pulse waveforms, and the inset shows variations. Figures 6G and 6H show a comparison of the extracted template waveform with the maximum distance waveform and the median distance waveform. The median distance waveform is still very similar to the template and retains most of the main features, such as the kickback peak in the wrist pulse. Detailed motion waveform analysis can be used, for example, as an electrocardiogram candidate for arrhythmia and aortic valve disease.

The CDMA protocol not only monitors multiple people at the same time, but also monitors multiple points on the same person at the same time. The allowable number of CDMA tags is limited by the baseband data rate and is shown in FIG. Comparison of waveform timings from different body positions provides an estimate of blood pressure ("BP") through pulse transit time ("PTT"), which can be extracted from the characteristic points of the proximal arterial waveform and the distal arterial waveform. The proposed non-contact blood pressure sensing presents a significant advantage over the direct stress based approach, which is based on long-term monitoring of older patients, particularly causing discomfort and damaging the circadian rhythm. Each of the chest label signal (near end waveform) and the left wrist label signal (remote waveform) is recorded for three minutes, as shown in Figure 7A. The PTT can be easily extracted from the main peaks of the two waveforms. The illustration shows the detailed waveform for a particular period. Figure 7B shows the probability density of PTT during 3 minute recording. The distribution of PTT is affected by sample jitter and waveform distortion. One PTT sample can be obtained for each heartbeat, and a moving average or other signal processing method can be readily applied to minimize PTT variations. Figures 7C and 7D show the blood pressure calculated from the PTT and the comparison point (star mark) from the commercial blood pressure monitor (OMRON BP760N). The solid line is the systolic and diastolic pressure of each heartbeat. The dashed line is from the moving average of the sampling points of about 10 seconds. The data in Figure 7C was collected when the subject was sitting on a chair for about 30 minutes, while the data in Figure 7D was collected after a moderate activity. Detailed method

Electromagnetic simulation using CST Microwave Studio. Zubal Phantom is used to construct dielectric models. Tissue geometry information is calibrated using data from computed tomography (CT) and magnetic resonance imaging (MRI). The resolution of the voxel is 3.6 mm × 3.6 mm × 3.6 mm. The CST biobank is used to map the microwave properties of various tissues. First, the Zubal Phantom data is preprocessed layer by layer into the organizational structure of the geometric coordinates and the organizational index. The CST then imports the files and automatically creates each voxel using three-dimensional coordinates and organizational properties to create a dielectric model controlled by the descriptive language of the CST built-in Visual Basic for Applications (VBA) macro language. The program is similar to 3D printing, but is actually only done in CST software. Dynamic simulation of heartbeat and wrist pulse is achieved by geometric changes in which the geometric properties of the heart and wrist vessels vary according to the preset size used as the live condition.

The custom PCB is modified from a Wireless Identification and Sensing Platform (WISP) by prototyping a passive harmonic backscatter tag with a custom PCB. The harmonic generator on the tag is designed with a nonlinear transmission line (NLTL) that contains a trapezoidal structure of the inductor and the varactor. NLTL provides high conversion efficiency with low input power, which is necessary for passive backscatter label design. Harmonic RFID readers and antenna reflection systems are built on the platform of the National Instruments Ettus Software Defined Radio (SDR) B210. To achieve the same tuning wave demodulation, the local oscillator (LO) of the receiver needs to be derived directly from the second harmonic frequency of the transmitter LO. Write real-time control and demodulation software with LabVIEW. Using a homodyne modulation scheme, the operating frequency is f = 950 MHz (at the second harmonic of 2f = 1.9 GHz). The downlink analog baseband is 10 kHz and the uplink analog baseband after harmonic conversion is 20 kHz. Both digital to analog conversion and analog to digital conversion operate at 106 samples per second (Sps). The original digital signal is then filtered, digitally downconverted to a D.C. band, and decoded using a CDMA algorithm to distinguish information from each tag. The signal from each tag is then sampled by a sampling rate of 500 Sps. The respiratory signal is processed by a low pass filter having one of the cutoff frequencies of 0.8 Hz. The heartbeat and pulse signals are processed by a bandpass filter between 0.9 Hz and 15 Hz. The current operating range of passive tags is ~1.5 m, which is limited by the WISP platform. The range can be extended towards 10 m according to the operation of the conventional RFID system in the same frequency band.

In another simulation, the antenna is deployed near the heart and the left wrist where the pulse can be felt. The source should be in the near field of the antenna, but there is no need for direct skin contact through the antenna. Figure 2A shows the simplified lower arm structure and the electric field in the near-field region of the CST Microwave Studio. The antenna is configured to couple more energy into the tissue to achieve a larger signal to noise ratio ("SNR"). The antenna reflection parameter S ₁₁ is shown in Figure 2B. The center frequency is approximately 1.85 GHz. Due to the high tissue permittivity, the antenna bandwidth is wider. The simulation results in Figure 2A show the electric fields coupled into the skin, fat and muscle layers and nearby blood vessels.

To simulate the mechanical motion coupled to the EM field, a small vibration is introduced into the geometric scale of the heart and wrist vessels, and the cross-section of the vessel in the tissue model changes quasi-statically in the time stamps of t1, t2, and t3. To indicate pulse changes. The normalized vibration amplitude is shown using thick solid lines in Figures 3A and 3B. Figures 4A and 4B show the far field patterns simulated in CST Microwave Studio. The shaded slope indicates the phase and the shape represents the amplitude profile of the co-polarized electric field. The far field signal is recorded and shown in Figure 3A (where the sampling point is 1 m above the chest (Figure 4A) and above the wrist (Figure 4B)). It can be seen that the demodulated heart signal (depicted using a dashed line in Figure 3A) and the wrist pulse signal (depicted using a one-dot line) are well matched to known vibrations.

The far field can be seen as the result of interference from two near-field components: the direct propagating wave from the antenna (black arrow in Figure 2A) and the scattered signal from the internal tissue (white arrow). When a heartbeat causes a blood vessel to vibrate, the phase of the scattered signal is modulated due to direct propagation of the interference. Since the deduced vascular motion signal is derived from the differential interference, the antenna motion caused by breathing or other body movement can be considered as a common mode that can be rejected. Alternatively, the back scattered signal may be coupled to the same antenna for mediation with the modulation, this resulting in the indicated as 2B of FIG. 2A and FIG gray arrow S _11.

2A illustrates how a portion of the EM field energy from the antenna is directly radiated to the far field (as indicated by the black arrow) while other portions of the EM energy are coupled into the multi-layer tissue until the target is sensed (here, the arterial blood vessels). Due to the difference in dielectric constant, the mechanical motion of the arterial blood vessels from the pulse will modulate the backscatter signal, indicated by the white arrow. This signal is also transmitted to the far field along with the direct radiated signal. From the point of view of the EM field, the two signals originate from the same source but pass through different paths, ie the two signals are coherent and the amplitude of the combined signal will change due to the phase difference. It can thus be understood how the mechanical movement of the blood vessel will result in amplitude modulation. This principle of operation is similar to an interference meter, and near-field modulation can be considered as a "differential component" in the interference meter analogy. On the other hand, the movement of the surface or the entire body (such as breathing and body movement) will simultaneously change the phase of the two signal paths, which can be considered as the "common mode" in the interferometer structure. Thus, NCS not only utilizes the sensitivity of the interferometer structure to enhance performance, but also isolates two different modulations: one modulation from the interior of the body with near-field coupling and another modulation from surface motion with direct emission. Passive backscatter

In an exemplary embodiment, the NCS can be implemented by a passive harmonic RF identification (RFID) tag in which the vital sign signal is modulated on the harmonic backscatter along with the tag ID. In addition to being ultra-low cost, the simple and robust package of this passive tag embodiment also enables direct fabric integration with the ready-to-wear garment. An exemplary harmonic RFID reader and antenna reflection system was built using the Ettus ResearchTM Software Defined Radio ("SDR") B210 platform.

A schematic diagram of one exemplary harmonic tag is shown in Figure 5A, and Figure 9B is a photograph of one of the PCB prototypes of the tag. For easy protocol access, all exemplary NCS operations are performed on the PCB prototype of the harmonic tag based on the Wireless Identification and Sensing Platform ("WISP") in Figure 8B. The prototype RFID sensor tag wafer is integrated with one of the embroidered antennas on the fabric (shown in Figure 9A). Fabric RFID tags are used to demonstrate the feasibility of garment integration of the disclosed methods. Verification of normal RFID data changes for tag ID and temperature sensing in accordance with the Electronic Production Code ("EPC") protocol.

Write real-time demodulation software with LabVIEW. Using a homodyne modulation scheme, the operating frequency is f = 950 MHz (at the second harmonic of 2f = 1.9 GHz). The downlink analog baseband is 10 kHz and the uplink analog baseband after harmonic conversion is 20 kHz. Both digital to analog conversion and analog to digital conversion operate at 10 ⁶ samples per second (Sps). The proposed passive tag operating range is approximately 1.5 m, which is limited by the WISP platform. The range can be extended towards 10 m according to the operation of the conventional RFID system in the same frequency band.

The principle of harmonic RFID backscattering is shown in Figure 8B. In one embodiment, a reader transmits a downlink signal at a fundamental frequency f that will supply power to any harmonic tag within the range of one of the readers. The harmonic tag receives from the reader a downlink RF signal at f, which passes through the tag antenna 1 and is divided into two parts: one for energy harvesting to provide DC power to the tag circuit and the other to Feeding into passive harmonic generation at 2f to retransmit from antenna 2 as an uplink carrier to eliminate reader self-interference and reflection from nearby objects. In this way, the antenna 2 is used as an NCS transmitter. A low pass filter ("LPF") at the transmitter (Tx) and a high pass filter ("HPF") at the receiver (Rx) further isolate the two carriers at f and 2f. Using the same tuned wave backscatter, Rx has a very low noise floor. Therefore, the Rx sensitivity can be increased, which enables the system to distinguish between weak vital signals modulated on the uplink. One of the RF switches in front of the harmonic generator (Fig. 5A) can modulate the digital information by turning on the key control ("OOK"), similar to conventional RFID operations. The digital information can include the tag ID and additional information from the sensor on the tag.

Multi-tag access can be achieved by using, for example, one of the Code Division Multiple Access ("CDMA") protocols. The use of this protocol provides better inter-tag synchronization, higher channel efficiency, and higher power efficiency in harmonic backscattering. The allowable number of CDMA tags is limited by the baseband data rate and is shown in FIG. The solid line is a CDMA code in which the length of the chip code is linearly proportional to the number of tags. However, as the number becomes larger, the semi-orthogonal code length can be approximated as a logarithm, shown as a dashed line. To recover waveform detail, the sampling rate should be above 500 Hz, which can be easily implemented in an exemplary embodiment of a harmonic backscatter system with CDMA.

The phase of the backscattered signal is modulated by an additional path and thus related mechanical movements (here, arterial and venous pulses). At the far field, the two parts interfere with each other as an interferometer, ie the wrist pulse will modulate the amplitude of the EM wave at the receiver. At the same time, an external movement, such as, for example, a swinging hand, modulates the phase of the two parts and will result in a common modulation of the two signals. In short, the hand movement will be in the common mode and the wrist pulse will be in differential mode. Figure 1A shows a sensing antenna deployed near the chest area to monitor heartbeat. Due to the near field effect, the EM field is coupled into the body and the backscatter signal is modulated by the heartbeat. The direct transmission and backscatter signals are received by the receiver at the far field. The two parts are from the same source and have different paths, and they will interfere with each other, and their center hops can be demodulated from amplitude. The chest movement caused by breathing changes the phase of the two signals, which is the common mode of the NCS. Similar to the hand movements in the wrist tag, the breathing and heartbeat signals are well isolated due to the common and differential modes that are independent of the latter filtering.

To implement the NCS as a wearable device in an indoor environment, an exemplary harmonic RFID system having one of the code division multiple access (CDMA) protocols is used. An RF schematic is shown in Figure 8B. The reader transmits a downlink (reader to tag) signal below frequency f. The harmonic tag receives the downlink signal and backscatters the second harmonic at 2f into an uplink (tag to reader) signal. The transmitter (Tx) and receiver (Rx) spectrums are separated to increase the signal-to-noise ratio (SNR) and receiver sensitivity. The CDMA protocol provides better tag synchronization, higher sample rates, lower jitter and lower power consumption. A schematic diagram of a harmonic tag is shown in Figure 22A. Antenna 1 (ANT 1) receives a downlink RF signal, wherein the tag collects RF energy, supplies power to the logic circuit, and demodulates the downlink information. A portion of the downlink RF signal is coupled to a non-linear transmission line (NLTL) to generate a second harmonic signal that is modulated by the RF switch and transmitted back to the reader through antenna 2 (ANT 2). The backscattered second harmonic signal performs the NCS function and is then received by the Rx antenna of the reader. A PCB prototype for the following experiment is shown in Figure 22B, which has a size of about 57 x 20 mm. The label can be further integrated into a passive wafer and packaged directly using a garment, wherein the antenna can be implemented by the conductive ink of Figure 30 and by the embroidery of Figure 9A.

Figure 10 shows the tag positioning error when applying a phase-based harmonic backscatter positioning method in conjunction with a CDMA protocol. The simulation results show the cumulative probability function of the positioning errors at various SNRs. Since self-interference and direct reflection have been eliminated in harmonic backscatter systems, the noise floor can be very low to easily achieve an SNR of 20 dB.

The experimental antenna reflection system is shown in Figure 12A and the measured heartbeat waveform is shown in Figure 12B. The schematic in Figure 12A can be implemented by an active tag mounted on a garment near the heart, where the tag antenna performs an NCS function. For direct comparison, a harmonic generator and a harmonic transceiver are selected to be similar to the passive tag system of Figures 5A and 5B to establish an antenna reflection unit. Other designs can be used including, for example, the use of a high isolation circulator in a single frequency. Both the signal splitter and the harmonic generator used in the experimental system are bidirectional. A signal from Tx at f is coupled to the harmonic generator and then coupled to the antenna at both f and 2f. The antenna reflections are fed again through the harmonic generator and splitter, but only about 2f of the signal is selected by the bandpass filter ("BPF") for feeding back to Rx for coherent demodulation. The life signal of more than 2f is sampled and transmitted to the remote device in a fully digital format, wherein inter-symbol interference caused by multipath and occupancy motion can be easily eliminated by standard techniques with low data rates. Figure 12B shows the demodulated heartbeat signal (applying a 0.9 Hz to 15 Hz filter) normalized to the 90th percentile of the data. Using the antenna reflection scheme, the vital signal detected by the NCS is more immune to severe multipath interference caused by the surrounding movement, such as in a crowded room.

As one of the independent verification procedures for NCS operation without direct mechanical contact, an asymmetric axis motor (ASLONG JGB37-520) with a known rate of rotation is used (Fig. 13). The passive tag antenna is placed close to the asymmetric rotation axis, and the mechanical rotation is converted into a sine wave through phase modulation and demodulation.

In another experiment, a first label (label 1) according to one embodiment of the invention is placed outside of a known motor having a known rotation, and a second label (label 2) is placed in the subject. Test one of the individual's shirts outside the heart area. The collected waveform from the motor (label 1) is shown in Figure 15 and used as verification or calibration. The collected waveform from the chest (label 2) is shown in FIG. The heart motion waveform of Figure 17 is obtained after applying a simple bandpass filter between 1.4 Hz and 15 Hz. In Fig. 18, one of the specific waveforms is placed for about three seconds, wherein the main motion features are extracted by including the characteristic points A to F. For example, point C indicates a primary contraction motion and point E indicates a shock wave kickback after closing the aortic valve.

In another experiment, the label 1 was placed near the chest area and the label 2 was near the wrist of the same person, and the resulting filtered waveform is shown in FIG. A segment of approximately three seconds is placed in Figure 20, wherein the time delay of the main peak (point C) of the two waveforms can be used to give an estimate of blood pressure using known techniques. There are 35 sampling points between peak 1 and peak 2, and the time interval is about 0.074 seconds and has a sampling rate of 470 Hz, resulting in a diastolic pressure of about 88 mmHg. Averaging the determined blood pressure estimates (e.g., obtained at each heartbeat) improves the reliability of the estimate. Accurate extraction of heartbeat time interval

Continuous monitoring of vital signs (including heart rate, blood pressure, respiratory rate, and respiratory function) is critical for eHealth. Previous methods were limited by accuracy, convenience, and sensing capabilities. Some methods, such as electrocardiogram (ECG), are difficult to use for long-term applications due to the requirements of direct skin contact, which makes the wearer uncomfortable, limits body movement, and destroys the circadian rhythm. At the same time, timing characteristics between heartbeats are important for health monitoring and human emotion research.

For existing systems, electrocardiogram (ECG) is the most popular method for heart rate monitoring, which uses body electrodes to concentrate body potential through small skin currents induced by heartbeat. To achieve good signal quality, the electrodes require direct skin contact by discomfort of conductive glue and hair removal. To further reduce noise, a large electrode with non-polarizable Ag/AgCl is required. Furthermore, when frequent hair removal is impractical in animal testing, insertion of a subcutaneous electrode not only causes additional measurement difficulties but also causes infection concerns. Infection can change the constant conditions in the body environment, which can cause serious deviations in the measurement of vital signs. Optical volume scanning (PPG) is widely used for pulse rate in both hospital and wearable devices. PPG utilizes the periodic change in optical absorption of the oxyhemoglobin level to modulate the intensity of the semiconductor laser and requires heavy processing of its signal to capture a clear vascular pulse waveform when detailed information is often lost. Furthermore, accurate measurements of heart rate and heart rate variability (HRV) are still challenging, such as limited by laser penetration depth, and even severe relative deviations from the beam relative to the laser beam. Acoustic methods, such as stethoscope-based heart sound maps (PCG) and ultrasound-based echo maps, have similar problems. The converter size undermines comfort and limits continuous monitoring of the wearable method. The distortion of the sound waves in the tissue, although compensated by the body fat-dependent post-processing, also degrades the signal quality. Other signs of life such as respiration rate/operation and blood pressure are usually measured by pressure or strain, but discomfort from wristbands or cuffs hinders long-term use.

A conventional RF method illuminates an RF beam to the chest region and backscatters far field electromagnetic (EM) waves by the body surface. The breathing and heartbeat signals are modulated on the EM waves detected by the receiver. Sensing at the far field leads to some disadvantages: (1) The difference in dielectric constant between air and human tissue results in strong reflection on the surface of the skin, which means that the respiratory signal is much stronger than the signal from inside the body. The strong breathing signal can exceed the heartbeat signal, and the heartbeat signal's signal-to-noise ratio (SNR) and waveform detail can be severely degraded due to a finite amount of energy and a small geometric mean. (2) Conventional methods often lack multi-channel or multi-point sensing, which limits their practical applicability. (3) From the viewpoint of the RF converter, the transmission and reception bands overlap each other, and self-interference can degrade system performance. Due to the above problems, it is difficult for the conventional RF vital sign sensing system to accurately measure the heartbeat waveform or time interval.

In contrast, the NCS method disclosed in the present invention modulates mechanical motion on the surface of the body and under the skin tissue in the near-field of the RF to a multiplexed harmonic RFID (RF) reversal with unique digital identification (ID). Scattered signal. In NCS, the respiration signal is well isolated from the heartbeat signal with higher RF energy coupled to the beating heart. The Tx and Rx signals are widely separated by harmonic backscattering, which improves SNR and provides multiplex of tag ID. With the improved signal quality of the heartbeat and further spectral equalization of the high frequency components to reduce sample jitter, the accurate heartbeat time interval can be reliably measured. Exemplary embodiment

To experimentally implement NCS for the heartbeat, the sensing antenna is placed close to the chest area. Conventional RF interrogators with full-featured transceivers will require active power from the battery or power line, which limits long-term monitoring capabilities due to size and convenience considerations. A passive and maintenance-free wearable device is preferred. The UHF RFID system is a good candidate. An RFID scheme is shown in Figure 8A. The reader transmits a downlink signal below frequency f, where the transmitter (Tx) signal also passes through the circulator to the reader antenna. The downlink signal received by the tag follows the Electronic Product Code (EPC) protocol. The tag does not require any active power because it uses the energy in the downlink RF signal to power it. The charge pump on the tag collects the small RF energy received by the tag antenna, and then the logic circuit modulates the backscattered uplink signal. The uplink signal is then received by the reader receiver (Rx) antenna and passed through the circulator. However, several issues will limit the system performance for NCS purposes. (1) There is strong self-interference directly from the downlink to the uplink. As shown in Figure 8A, strong Tx to Rx leakage still exists due to circulators, direct Tx antenna reflections, and insufficient isolation from ambient reflections from nearby objects. Due to the low modulating rate of the passive tag, the Rx signal is subject to large noise from the Tx phase noise edge, which limits SNR and reader sensitivity. (2) Conventional RFID systems typically employ a circulator as a Tx/Rx duplexer that has a finite bandwidth and is therefore susceptible to indoor multipath interference. (3) The conventional EPC protocol utilizes Time Division Multiple Access (TDMA) to handle tag collisions, where the tag selects a random delay time in the Aloha scheme. This random delay introduces additional aperture jitter to the sensed signal, which further degrades the SNR and causes signal distortion.

To solve all of these problems, a harmonic RFID system can be used. As shown in Figure 8B, the Tx signal is at f and is transmitted through a low pass filter (LPF) to the Tx antenna. The harmonic tag is powered by the downlink RF signal to the harmonic tag and backscatters the 2f harmonic signal in the uplink through the Rx antenna and the high pass filter (HPF). The Tx and Rx spectra are well separated and the Rx signal is hardly affected by the Tx phase noise edge. The reader RF front end uses LPF and HPF as duplexers, and the bandwidth can be wider for indoor multi-path immunity. At the same time, the harmonic tags can also operate on the CDMA protocol, with the reader synchronizing the harmonic tags to provide multi-tag access with minimal aperture jitter.

A schematic diagram of a harmonic tag is shown in Figure 22A. Antenna 1 receives a downlink signal from a reader. A portion of the signal is delivered to the energy harvesting module to power the tag control circuitry, and the tag receiver demodulates the downlink command for proper air protocol and corresponding logic operations. The other portion of the RF energy is coupled to a non-linear transmission line (NLTL) to generate a second harmonic signal in the uplink that is modulated by the RF switch to radiate through the antenna 2. Figure 22B shows a PCB prototype of an exemplary embodiment in which the digital logic portion is modified from the WISP platform.

The Ettus B200 Software Defined Radio (SDR) is programmed into a harmonic RFID reader. The Tx local oscillator (LO) is set at 1 GHz and the Rx LO is set at 2 GHz, which have the same clock source to maintain the same tuning wave transceiver scheme.

An exemplary vital sign monitoring setting is shown in FIG. A first label senses the heartbeat and breathing in the chest area, and a second label is on the left wrist to sense the pulse. The chest strap and wrist strap are only used for convenient label deployment on garments and do not require skin contact or elastic band tension. The tag receives the downlink signal from the SDR reader and backscatters the uplink signal to the reader antenna. The demodulated NCS signal is processed by LabVIEW and displayed on the screen, with a white curve tracking the heartbeat and a red curve tracking the wrist pulse. The waveform is processed by a bandpass filter (BPF) having a frequency of 0.8 Hz to 15 Hz. The delay from the heart signal to the pulse signal is the pulse transit time (PPT), which can be used to estimate blood pressure. 24A and 24B show demodulated NCS signals for heartbeat and respiration collected from a chest RFID tag. The upper curve in Figure 24A is the normalized amplitude portion of the NCS signal. It can be seen that the main modulation is the heartbeat, but the coupling is slightly coupled. The isolation between the amplitude and phase modulation in the quadrature receiver gives a clear separation between the respiratory waveform and the heartbeat waveform, as depicted in the interferometer analogy. After the digital BPF (0.8 Hz to 15 Hz), the heartbeat signal is clearly captured as the lower curve. Figure 24B is a normalized phase portion of the NCS signal. The original phase waveform is shown as an upper curve where the main modulation system breathes, but is also slightly coupled to the heartbeat. The raw phase data is processed by the digital BPF in a different range (0.1 Hz to 1.2 Hz) to capture the filtered breathing signal in the magenta curve. Instead of relying solely on band filtering in previous work to separate the respiratory and heartbeat signals, the NCS can distinguish between external and internal mechanical motion by an orthogonal scheme. Accurate extraction of heartbeat time interval in an exemplary embodiment

To measure the heartbeat time interval, the waveform and associated peak-to-peak time can be accurately detected. The timing accuracy therefore depends on the temporal resolution and sharpness of the feature points. The sharp edges or peaks in the time domain are better reflected by the high frequency components in the spectrum. However, the low frequency components from the breathing and main ventricular motion naturally have very large amplitudes and can therefore exceed spikes for precise time interval detection. As shown in FIG. 25A, the solid curve is an NCS heartbeat signal and the dotted curve is an ECG signal, and the ECG signal is synchronized to the main link in the SDR reader. The sampling rate of NCS is 5000 samples per second (Sps), and the sampling rate of ECG is 512 Sps, which is initially limited by electrode noise. The ECG signal is upconverted to 5000 Sps for simultaneous and graphical display. To measure the heartbeat time interval by ECG, the peak of the QRS complex in each cycle can be used. Since the peak is already sharp, its timing can be easily located. However, the NCS signal directly measures mechanical motion and is therefore smoother without significant peaks, which reduces peak detection accuracy. In order to capture precise timing, spike feature points are required. Therefore, the original NCS signal must be further processed to improve peak detection certainty. Figures 25B and 25C show the spectrum of the synchronized NCS and ECG signals. The two spectra are normalized to respective peak intensity values of approximately 1 Hz. Compared to the NCS signal, the ECG has a stronger high frequency (2 Hz to 8 Hz) component and therefore has a sharper peak. However, due to the direct motion modulation, harmonic backscatter design and higher sampling rate in the NCS, the noise floor is much lower than the ECG noise floor. However, due to the simultaneous measurements of the same heartbeat source, the frequency components of the two spectra are reasonably aligned and share a similar distribution.

The NCS signal has a rich high frequency component of approximately 4.5 Hz to 6 Hz, but the intensity amplitude is at least 10 dB lower than the dominant heartbeat signal of approximately 1 Hz. In contrast, ECGs also have these components but have higher strength and noise. Therefore, a direct spectral equalization can be applied: a 4 Hz to 7 Hz bandpass finite impulse response (FIR) filter is applied to the original signal, and then the output of the filter is amplified by 13 dB to add back to the original signal. To align the two signals in time domain, the filter can be designed as a zero phase filter to eliminate phase delay. The FIR structure is advantageously used to maintain a linear phase frequency response. The processed signal is shown in Figure 26A with more feature points and spikes than the original signal of each heartbeat cycle. The highest peak feature point for each heartbeat cycle is selected as a timing indicator, shown as a triangle in Figure 26A.

The heartbeat interval is defined as the time between a cycle and a feature point in the next cycle. The heartbeat time interval is shown in Figure 26B. The heartbeat time interval is calculated for the original NCS (dashed line), ECG (dotted line), and equalized NCS (solid line) signals. For the original NCS and ECG signals, the feature points are selected as the highest positive peak. Since the heartbeat time interval is calculated based on the time between heartbeats, the horizontal axis is a heartbeat index rather than time. The vertical axis is the heartbeat interval and its reciprocal can be considered as the heart rate. All three curves show the same overall downward trend or equivalently increase the heart rate, since the measurement system is performed after physical exercise. Although no live is provided here, the following observations can be made. Compared to the equalized NCS heartbeat interval, the original NCS has a large change due to inaccurate peak detection; the ECG signal has a large variation due to the low sampling rate of 512 Sps. The more stable heartbeat interval variation in the equalized NCS within several heartbeats has physiological significance. Body antenna impedance effect

The associated antenna design strategy is used to analyze the effect of the antenna on NCS performance. When better antenna impedance matching is achieved, the energy coupling efficiency and SNR can be improved to detect ECG waveform details. Antenna impedance matching in NCS

To demonstrate the operating principles of NCS, a human body electromagnetic (EM) simulation model was built in CST Microwave Studio, as shown in Figure 27A. The geometrical and tissue properties of the organs of the body were extracted from Zubal Phantom. The sensing antenna is attached close to the chest area but does not require contact with the skin. The transmitted EM wave is partially transmitted to the far field and displayed as a gray arrow. Another part of the RF energy is coupled to the body due to the near-field effect, shown as a dashed arrow. The phase of the backscattered signal is modulated by the heartbeat due to the movement of the dialectical boundary. The modulated backscatter RF signal, shown as a black arrow, interferes with direct transmission (gray arrows) and is received by the RF receiver. Since the two signals are from the same source but have different paths, the operation is similar to an interferometer structure. The internal heartbeat gives a differential modulation of the two signals and can be demodulated from the far-field RF magnitude. At the same time, external chest movement caused by breathing or body movement will change the phase delay of the two signals (grey and black arrows) (this may be referred to as common mode modulation) and may be demodulated from the far field RF phase. The magnitude and phase information at the receiver can be easily and accurately separated by a quadrature mixer.

One part of the NCS is a sensing antenna that is close to external and internal body motion. Due to the high dielectric constant of the human tissue in the near field region of the antenna, its S-parameters will vary significantly. From a design point of view, the antenna can be seen as a matching component between the impedance of the RF circuit (typically about 50 W) and the impedance of the connected region of free space and body, which will change only in free space compared to the antenna. Impedance, frequency response and radiation pattern. For the CST simulation in Figure 27B, when the 2 GHz dipole antenna is operating in free space, S _{11 of} the antenna is shown as a solid line. However, when the antenna is attached near the chest region, S ₁₁ will be offset to the dotted line, which means that the frequency around 2 GHz is no longer good at high transmission due to high reflection. There are two possible strategies to solve simple: to shift the operating frequency band having a low S _11, the re-design or geometric properties of the original antenna to fit the 2 GHz band. The dashed line is a redesigned antenna that matches the body that is present. The reflection at 2 GHz has been greatly improved from -3 dB to -18 dB.

To further investigate the effectiveness of the antenna matching effect, Figure 28 shows the power flow at 2 GHz when an antenna is placed on the chest in Figure 28B. In Fig. 28A, the original antenna 1 has a dotted line, and the modified antenna 2 of Fig. 28B has a broken line. The cross section of the human body is depicted as the left (L) lung, the right (R) lung, and the heart. At the same drive signal strength and color profile scale, it can be observed that the energy coupled from the non-matching antenna in Figure 28A to the body is much less than the energy of the matching condition in Figure 28B. Strong energy coupling also increases the total intensity of the backscattered signal. Therefore, both SNR and sensing sensitivity will be improved with the same noise floor. Experiment and analysis

In an exemplary embodiment, a harmonic backscatter RFID system is utilized to implement NCS on a convenient and high performance sensing platform. The NCS antenna is part of the harmonic tag and a schematic diagram of the harmonic tag is shown in Figure 22A. The tag is designed to collect downlink RF energy from the reader to power the passive device of the tag circuit. A portion of the RF energy received by antenna A (Ant. A) is collected by a charge pump to operate the receiver and microcontroller unit (MCU). The other part of the RF energy is coupled into a nonlinear transmission line (NLTL) for second harmonic generation to backscatter through the antenna B (Ant. B) to the reader. The RF switch uses on-off keying (OOK) to tune the harmonic signal to the uplink baseband. A code division multiple access (CDMA) protocol is implemented on the tag MCU to achieve better synchronization and performance in a multi-tag case. 22B shows a printed circuit board (PCB) prototype for an exemplary harmonic tag with an NCS sensing antenna mounted to an antenna B connector. The uplink RF signal from antenna B is also modulated by the breath and heartbeat and then received and demodulated by the reader. The harmonic reader is implemented by a software defined radio (SDR, Ettus B200). The local oscillator (LO) of the reader transmitter (Tx) is set at the fundamental frequency f, and the receiver (Rx) LO is set at the second harmonic frequency 2f. The two synthesizers of Tx and Rx LO are driven by the same frequency reference, so the reader is configured as a harmonic coherent transceiver. One of the main benefits of using harmonic backscattering (rather than the conventional RFID system) is the broad frequency separation of the downlink and the uplink, so the weak backscattered signal from the tag is not subject to the Tx The high phase noise edge of the leak. Therefore, sensitivity and SNR can be much higher than conventional RFID solutions.

The NCS measurement setup is shown in FIG. One label is in the chest area for heartbeat and breathing, and the other label is on the left wrist to obtain a pulse. Chest straps and wrist straps are used for convenient label deployment on garments without the need for skin contact or elastic band tension. The tag receives the downlink signal from the SDR reader and backscatters the uplink signal to the reader. The demodulated NCS signal is processed by a 0.8 Hz to 15 Hz bandpass filter (BPF) in LabVIEW and displayed on the screen, with a white curve tracking the heartbeat and a gray curve tracking the wrist pulse. The delay from the heart signal to the pulse signal is the pulse transit time (PPT), which can be used to estimate blood pressure. The reader antenna is currently 1.5 to 3 meters away from the person, which is limited by RF power acquisition and passive harmonic conversion losses. This range can be easily improved by custom design of the reader and the label.

A monopole antenna is used as the sensing antenna for an exemplary embodiment of the NCS, and S ₁₁ of the antenna is shown in Figure 29A as a solid line having a center frequency of 2.1 GHz when the antenna is operating in free space. When the NCS antenna is placed near the chest region, S ₁₁ responds to the dotted line with a center frequency of 1.9 GHz in Fig. 32A. The reflection of about 2.1 GHz is now very high. To achieve high energy coupling and high SNR in the NCS, the downlink signal can be changed to 950 MHz, so the sense antenna is well matched at 1.9 GHz. As shown in Figure 29B, the solid line is obtained by the uplink signal at 1.9 GHz (shown as the mark of frequency 1 in Figure 29A). The dotted line with the uplink at 2 GHz (frequency 2 in Figure 29A) and the dotted line at 2.1 GHz (frequency 3 in Figure 29A) have very weak NCS due to less energy coupled into the body. signal. It should be noted that the RF radiation efficiencies at frequencies 2 and 3 are still quite high, estimated at 90% and 80% at S ₁₁ of -10 dB and -7 dB, respectively, and therefore most of the degradation in the NCS signal can be attributed. The reduction in tissue coupling. Perform all measurements under the same system setup and extraction procedures (except for different frequencies). At the same system noise floor, higher NCS signal strength increases SNR to restore sharper waveform detail. Repeatable details can be observed along the decreasing slope in good matching conditions of the solid line, which is barely visible in non-matching conditions with uplinks at 2.0 GHz and 2.1 GHz. Reduce body movement interference

In this paragraph, an exemplary NCS vital sign monitoring system has a mitigation method based on a high frequency heartbeat component to combat interference from body movement. A low error probability is experimentally demonstrated in immediate heart rate measurement by using synchronous ECG calibration.

As mentioned above, electrocardiogram (ECG), optical volumetric scanning (PPG), and acoustic methods (such as stethoscope heart sound maps and ultrasonic echocardiograms) are current techniques for measuring heart rate and its variability, with ECG Jia is established as a clinical standard. However, these methods have received a lot of attention in wearable sensing systems. In order to achieve good signal quality in ECG, the electrodes need to be in direct skin contact with unsuitable conductive glue and hair removal. Although wearable ECG garments have been tried, it is still difficult to use for everyday wear. Current PPG devices are widely used on smart watches, wristbands, patches, and wrist or finger grips, but they require a close contact to avoid distortion and relative motion from ambient light. The PPG signal relies on heavy processing to obtain a fairly sharp pulse waveform, and detailed information in both the low frequency range and the high frequency range is often lost. The sensing depth of the oxygen content is limited such that the PPG is most commonly applied to areas of the body having a high blood concentration in the shallow blood flow or transmission mode in the reflective mode, which is difficult to achieve with direct cardiac motion measurement. Stethoscope heart sound maps are susceptible to internal noise (breathing and speech) and external interference (sound and vibration) and are only practical in a controlled laboratory environment. The ultrasonic echocardiogram is difficult to implement as a wearable device because its bulk converter requires the use of an impedance matching gel to directly contact the skin surface. In addition, strain gauges are often required when it is desired to simultaneously measure respiration rate/operation and blood pressure, but the discomfort from wristband and cuff tension prevents long-term use and can also disrupt sleep or circadian rhythms. Several RF-based methods have also been proposed in which the RF beam is radiated to the chest region as one of the far fields for backscattering by the human body. Both the breathing and heartbeat signals are modulated on the RF carrier and then received by the receiver antenna. If the backscattered signal is further modulated by an RFID, it is believed that impending skin contact is important for reliably capturing the tiny skin movements of breathing and heartbeat. Interference from any other non-specific backscattering can be disadvantageous if the backscattered signal is not modulated by one of the digital IDs (identifications) in the unlabeled case. With or without a personal tag, body movements will result in severe inaccuracies in RF-based heart rate estimates.

In this paragraph, an exemplary NCS vital sign monitoring system has a mitigation method based on a high frequency heartbeat component to combat interference from body movement. A low error probability is experimentally demonstrated in immediate heart rate measurement by using synchronous ECG calibration. Experiment and analysis

One of the embodiments of the test NCS technique is set using the experiment shown in FIG. A software-defined radio (SDR, Ettus B200) is programmed to be used as a harmonic reader. A passive harmonic tag is deployed at the chest and left wrist area. Chest straps and wrist straps are used for convenient label deployment on garments without the need for skin contact or elastic band tension. The NCS (measurement) signal is received by the reader Rx antenna and demodulated by the SDR harmonic reader. The resulting demodulated heartbeat and wrist pulse are displayed on the screen as a light curve and a dark curve, respectively. The delay of the wrist pulse relative to the heartbeat is the pulse transit time ("PTT"), which can be used to estimate blood pressure. Figure 31 shows the demodulated signal from the chest tag. The heartbeat signal is the amplitude of the measured signal from bandpass filtering (BPF: 0.8 Hz to 15 Hz). The respiratory signal is derived from the phase of the measurement signal after the BPF from 0.1 Hz to 1.2 Hz.

Although the NCS can isolate external mechanical movements and internal movements (such as breathing from the heartbeat and hand movements from the wrist pulse), the measured signal can still be measured when the characteristic spectral component of the body motion is close to the characteristic spectral component of the vital sign. Interfered with large body movements. This additional coupling is similar to signal contamination by a mixture of amplitude modulation (AM) and frequency modulation (FM), especially when one of the AM and FM sidebands is very large. The original NCS measurement signal is shown in Figure 32A when a large motion is involved. During the first 40 seconds, the test subject sat down to establish a baseline of clear heartbeat signals for reference, as shown in the inset. During the period between 40 s and 70 s, the personnel violently waved their hands. During the period between 70 s and 100 s, the person moves his body and stands up at the 97th second, then continues to move his body left and right. In the 120th second, the person resumes sitting for 15s. In about 135 seconds, the person stood up again until he sat down again in the 145th second. The spectrum of the 20 second window with and without movement is shown in Figure 32B. The spectrum is normalized to the intensity of the main peak (without movement) of about 1.5 Hz and amplified to the lower intensity portion. Filter out components below 0.8 Hz and above 15 Hz. The light gray curve is the spectrum during a time period from 10 s to 30 s without body movement, and the dark gray curve is between 85 s and 105 s when the heartbeat signal is disturbed by large body movements. The spectrum without body movement not only displays the main peak of about 1.5 Hz, but also clearly shows the higher frequency harmonics. Body movements typically have strong characteristic frequencies at lower frequencies. The main peak of the heartbeat is severely distorted by the body movement with a signal-to-interference ratio ("SIR") of only above 0 dB. However, components at higher frequencies are less contaminated by body movements. The SIRs of peaks A, B, and C are 3.4 dB, 5.9 dB, and 7.6 dB, respectively.

Since the high frequency components of the heartbeat signal have a better SIR, they can be used for heartbeat counting with less impact from body movement. The NCS signal passes through one of the BPFs from 4 Hz to 5.5 Hz to capture the B-peak of the third harmonic, and when the person stands up at the 97th second, a synchronized electrocardiogram is used in Figure 33A (solid line) for 95 s to 99 s ( ECG) (dashed line) display. The choice of the third harmonic is due to its reasonable SIR and a large number of values. The high frequency component is also multiplied by the heart rate to reduce the counting error. As shown in the previous two heartbeats in Figure 38, there are six largest and smallest peaks in the NCS, so twelve total peaks are used for heart rate estimation. However, during the period from 96 s to 98 s, the personnel stood up and the original NCS signal changed significantly, while the ECG signal was also distorted in the case of direct gel-coating pads, however its QRS characteristics were attributed to its high frequency. The characteristics remain fairly clear. There are seven heartbeats according to the ECG signal and seventeen peaks from the third harmonic of the NCS. The NCS misses a peak, but the heartbeat count error is only 5.6% within the three heartbeats and 2.4% for the range in Figure 33A.

To analyze the immediate heart rate, the heartbeat count can be resampled evenly at the time domain. The heartbeat count is based on the maximum and minimum peaks, but the time information of each peak is not evenly distributed as a discrete signal. Therefore, the heartbeat data is fitted to the cubic spline and uniformly distributed with a time resolution of 0.05 s. The heartbeat count versus time is shown in Figure 33B. The real curve and the dashed curve are the NCS and ECG count results, respectively, and the bottom curve (labeled ERROR) is the error. The maximum error occurs after the 40th second when the person is waving his hand. The heartbeat count curve for this period is shown in the inset of Figure 33B. After the entire 160 s count, the ECG results were 243 (mean heart rate of 91.1 beats/min) and NCS was 241 (mean heart rate of 90.4 beats/min). The error is -0.8%. The heart rate curve monitored by the third harmonic of NCS (solid line) and ECG (dashed line) is shown in Figure 33C. The curve is processed by a low pass filter of 0.5 Hz. The average heart rate and error based on the second harmonic, third harmonic, and fourth harmonic peak counts (A, B, and C in Fig. 32B) are shown in Table I below. The coherence coefficients are calculated using the heart rate curve obtained by the ECG and the harmonics of the NCS. The second harmonic of the NCS gives the maximum error due to the low SIR. Table I Heartbeat analysis using different peaks

Compared to other heartbeats to noise or counting methods, the currently used extraction method provides an immediate heart rate without the need to calculate the entire spectrum of one of the long durations of the data needed to obtain a reasonable frequency resolution. The computational load is also small, which can be easily performed on a wearable device having a basic microcontroller. Sleep score

Long-term sleep scoring is very important in clinical settings to monitor patient recovery and monitor children and adults at home. In a cost-effective manner, sleep quality can usually be assessed by upper body movement along with heartbeat and respiratory monitoring. Instead of the conventional polysomnography (PSG) that causes discomfort due to skin contact between the sensor and the electrode, this paragraph presents an RF near-field homology using a single passive radio frequency (RF) identification (RFID) tag in the chest region. An exemplary sleep monitoring system and method for sensing (NCS) without skin contact, wherein heart rate, respiratory rhythm, and motion detection can be simultaneously extracted. The motion classification is based on a support vector machine (SVM) with semi-supervised learning. Sudden body twitching, shaking and turning can be correctly identified in 91.06% of the test cases. Heart rate detection accuracy is also improved after motion artifact correction.

This embodiment attempts to have the ability to simultaneously monitor vital signs and body movements for sleep scoring in a low cost and non-invasive manner, this paragraph presenting one of the passive UHF RFID systems based on Near Field Coherent Sensing (NCS). A sleep scoring system that enhances RF energy coupled into the body and thus increases backscatter signals from heartbeat, respiration, and motion within and on the body. Modulation of the dynamic dielectric boundary can be modulated on a radio signal having a unique digital identification (ID), which can easily be extended to monitor multiple tags and personnel by a single RFID reader. In some embodiments, the subject can wear only a single chest label that can be integrated into the fabric without the need for skin contact or motion constraints, as shown in FIG.

A signal anomaly detection algorithm based on the deformation of vital sign waveforms has been developed for motion in PPG and ECG signals and others for error detection, in which time and spectral characteristics for SVM classification, multi-layer (MLP) perceptrons, Decision tree or other classifier. The motion detection algorithm is based on detecting changes in the NCS waveform during shaking, turning, and sudden body twitch movement. An SVM classifier only uses the signal characteristics at rest to train and can detect anomalous features during motion. Experimental setup for sleep monitoring

The NCS can be performed when the antenna of the NCS is deployed in the near field region of the motion source. To assess sleep quality, harmonic RFID tags are placed in the chest area, which is prototyped on a Radio Identification and Sensing Platform (WISP). The harmonic reader is implemented using the National Instruments Ettus Software Defined Radio (SDR) B200. Analog-to-digital and digital-to-analog conversions are performed at 10 ⁶ samples per second to capture accurate baseband waveforms. The reader transmitter frequency f is 1 GHz and has a 10 kHz analog baseband; the corresponding reader receiver frequency is 2 GHz and has a 20 kHz baseband.

A controlled data collection was performed to simulate the following cases during sleep: 1) Rest: The subject consciously remained stationary. 2) Body movement: a slight body movement of 5 seconds. 3) Body twitching: One of the body and arm of 0.5 seconds to 2 seconds, fast and high energy exercise. 4) Turn around: The subject turns left or right in 1 second to 2 seconds.

The data was collected for 1 hour, and the NCS signal sampling rate was 500 samples per second. The simulated movement is performed at intervals of 1 minute. Indoor multipath effects can be ignored because the delay spread is much lower than the NCS sampling time. Signal analysis

The surface and internal movement of the body (mainly composed of respiratory and heartbeat signals during the test) are modulated on the label backscatter and can be captured by the signal processing of the demodulated variable bit data. Both the amplitude and phase of the NCS data capture breath, heartbeat, and motion information with different weights. The phase is most sensitive to the overall label motion, while the amplitude is less sensitive to small label motion but is a strong function of the antenna characteristics caused by the near field coupled to the internal motion. Figure 35A shows the raw DC filter amplitude and phase data at rest. Sign of life

The breathing rhythm is extracted from the phase by simple filtering. In Figure 35B, a 20th order Butterworth filter is used to pass frequencies between 0.01 Hz and 1 Hz. In Figure 35C, the heartbeat is extracted from the NCS amplitude by removing frequencies below 0.6 Hz that contain spontaneous breathing information. A linear phase filter is used to remove all spectral content above 10 Hz.

The heartbeat signal can be severely disturbed by respiratory signals during a slight wheezing event. Simple filtering can also distort the signal in the case of obstructive breathing. Motion detection

Motion detection for sleep scoring using accurate heartbeat and breathing information is not directly achieved by simple filtering. Exercise affects the waveform characteristics of both heartbeat and respiration, but accurate life signals can be captured by correcting the signals affected by the motion. Perform a heartbeat segmentation to achieve a finer time resolution of the motion. Accurate peak detection is difficult in non-stationary heartbeat waveforms with multiple peaks. The method begins with multi-level one-dimensional wavelet decomposition, using the Daubechies 10-tap wavelet to identify the coefficients that cause the greatest correlation with the heartbeat after reconstruction. The reconstructed waveform of the 8-level (NCS-d8) detail coefficients containing the main components of the heartbeat waveform is shown in Figure 35D. Accurate peak detection can be performed at this stage, with an additional constraint of the minimum peak distance being determined by one of the maximum heart rate of 200 heart beats per minute.

The motion characteristics are identified based on the difference between the motion-affected waveform and the waveform obtained at rest. Moreover, the features are advantageously robust to account for changes in breathing, heartbeat, and signal amplitude over time. The relative heartbeat interval and relative heart rate root mean square (RMS) are based on the assumption that the heartbeat interval is not expected and the RMS varies significantly between heartbeats. The statistical mean, variance, skewness, and kurtosis are calculated to capture the main difference between the motion and the waveform at rest, where one of the five heartbeats is used to inspire the window. The normalized spectral power in the range of 0.6 Hz to 10 Hz within the same window is used as another characteristic motion in addition to the heartbeat harmonics.

The motion classification is based on the above seven features. SVMs with radial basis function kernels have been used to detect motion. Semi-supervised learning is used to train the model, ie, training is performed using only the data collected by spontaneous breathing at rest. Excluding exercise data for training reduces over-fitting and generalization of motion from redundant learning and reduces the inconvenience of performing mobile routines during training. 36A and 36B show NCS heartbeat and corresponding motion detection. The effectiveness of the motion detection algorithm is evaluated by manual annotation results. Error statistics

Table II shows the number of correct affirmation (TP) and false negative (FN) cases for each sport category. False positive (FP) indicates signal transmission motion without actually occurring, which is mainly due to irregular breathing patterns and heartbeat detection errors. A beat-by-beat motion classification gives an accuracy of 97.58%, a sensitivity of 88.28%, and a specificity of 98.10%. Table III shows the confusion matrix with the number of heartbeats in each category. Figure 36C shows an example of an improved heart rate estimate after removal of motion artifacts. The heart rate is estimated using a moving average of one of the 30 window lengths. Table II, motion detection results of the proposed algorithm Table III, Heart rate-by-heartbeat analysis of accurate heart rate

The body's twitching movement can be detected with good precision, while the slow turning can be misclassified as rest. Accuracy can depend on training data and related algorithms. Experimental examples were trained using spontaneous breathing at rest (including regular breathing with occasional deep breathing). Using a regular breathing pattern training increases the sensitivity to motion and causes an increase in FP in the case of irregular breathing. Small animal life sign

Current methods of measuring animal vital signs typically involve complex and invasive preparation procedures and result in major discomfort in the animal being tested such that they typically need to be anesthetized. For example, an electrocardiogram (ECG) for a heartbeat waveform requires a body electrode with good electrical contact and is therefore difficult to apply to mammals with thick fur, reptiles with scales or shells, feathered birds, and scaled fish. Exposed areas of the skin such as the soles of the feet and lips typically have insufficient electrical signals and are sensitive to touch. The muscle electrode can usually be performed using anesthesia. Similarly, body surface conditions also result in complexity in optical volumetric scanning (PPG) settings, which limits its application to animals. Auscultation and ultrasound require close skin contact or impedance matching gels to obtain a clear signal, which requires a high degree of animal handling. A radio frequency (RF) method based on Doppler far-field backscattering of minute skin motion has a non-specific wireless channel and is susceptible to any surrounding motion interference. Breathing is usually the dominant signal and becomes one of the main disturbances to accurate heartbeat detection. The RF method based on the transmission line model again requires good impedance matching of the skin electrodes. Small animals are at the forefront of most of these RF methods due to limited signal sensitivity.

The NCS method of the present invention, which utilizes near-field modulation of antenna characteristics on multiplexed radio signals, does not require skin contact and provides an effective solution for long-term monitoring of vital signs of conscious small animals. A schematic of two exemplary NCS settings suitable for use with small animals is shown in Figures 37A and 37B. The wireless sensing in Figure 37A utilizes a harmonic RFID (Radio Frequency Identification) architecture that makes passive sensing tags inexpensive and maintenance free, but requires a specific multiplexed reader. This version can be adapted to be deployed in a natural habitat with weather resistant passive tags and to collect vital signs from a reader near one of the operator or fixture. The use of harmonic systems reduces self-interference by separating the frequency bands of the transmitter (Tx) and receiver (Rx), improving signal sensitivity and signal-to-noise ratio (SNR), and reducing the required strength of the test RF Tx signal, all These are advantageous for sensing the vital signs of small animals. The NCS signal on the animal's body can be much lower than the prior art (eg, 0.1 mW/cm ² and 0.15 W/kg) to follow the health and safety criteria in the rodent model. In Figure 37A, the Ettus X310 Software Defined Radio (SDR) harmonic reader transmits a downlink signal at about 950 MHz through the reader Tx antenna. The downlink signal is supplied to the passive harmonic RFID tag and then converted to a second harmonic frequency at 2f to serve as the NCS sensing signal in the near field range of the target animal body. As long as the vital signal is within this near-field range of the sensing tag antenna (typically about one-third of the wavelength used) without any skin contact, motion on the animal's body and inside the animal's body can be coupled to the backscatter signal. The reader Rx antenna is received. The high SNR in the NCS allows for accurate measurement of small internal motions, such as the human wrist pulse waveform, which is beneficial to the vital signs of small animals. The main integral tag antenna motion relative to the reader will primarily be reflected as phase modulation, which can be naturally separated from the magnitude modulation due to the dielectric boundary motion relative to the tag antenna in the near field. The signal from the sense tag may contain a unique identification (ID) code for component code multiple access, which improves channel isolation to prevent non-specific interference and enables simultaneous reading of multiple sense tags. The manufacture of passive sensing tags is similar to conventional RFID tags, which provide low cost production and flexible substrate selection in addition to convenient deployment without maintenance.

Alternatively, the setup in Figure 37B replaces the reader to the tag channel with an RF cable, which reduces interference and can be conveniently deployed in an indoor laboratory with heavy operator traffic or with other sources of interference. The reader Tx antenna directly transmits the NCS sensing signal at 2f and is within the near field of the animal being tested. The NCS signal modulated by the vital signs is then received by the reader Rx antenna, which can be deployed according to the application considered.

To compare the NCS vital signs on the heart waveform of small animals, first perform a simultaneous NCS and ECG measurement on a rodent model, which is not only important for clinical trials, but also has a very weak and rapid heartbeat, providing a more challenging test. . An Long-Evans laboratory rat (groove) (coded #110) was anesthetized to perform abdominal hair removal by razor and Veet gel cream for ECG electrode deployment, as shown in Figure 38A. Try using a spring clip and a conductive sticker pad on both claws, but the ECG signal is very weak and there are many noises. The sensing tag antenna having the architecture of Figure 38B is placed near the back of the neck region by a rubber heddle without the need for epilation. A reasonable NCS signal is also used to try to include placement along the tail, other antennas in front of the chest, and along the hind legs. A representative 5 minute record is shown in Figure 38B, and the inset shows waveform detail within half a second of the third minute. The ECG waveforms of the ECG and NCS in Figure 38B present a very similar heartbeat time interval as shown in Figure 38C, but detailed feature timing (e.g., the position of the NCS feature relative to the timing between the S and T feature points of the ECG) may be required. Further characterization. Based on timing comparisons, NCS can replace ECG for heart rate based (HRV) based behavioral studies. The respiratory waveforms were collected synchronously by NCS with additional 0.5 Hz to 2.5 Hz low pass filtering in Figure 38D, allowing for further cardiopulmonary analysis that cannot be achieved by ECG alone.

The NCS measurement changes the characteristics of the tag antenna by the near-field geometry, while the ECG measures the body potential difference induced by the tiny skin current, which is further induced by cardiac electrical stimulation and blood flow. From this point of view, the NCS has a waveform similar to one of the heart impact maps (BCG) and is one of the more direct heartbeat motion measurements than the ECG. NCS is also less subject to changes from skin conditions and preparation steps.

After confirming the cardiopulmonary signals on the anesthetized rats, the possible non-invasive NCS settings in Figure 39 are then demonstrated, which are very difficult (if possible) for ECGs and other conventional techniques for conscious small animal systems. A pet gold hamster (Mesocricetus) named "Timo" was monitored in its cage sleeping area as shown in Figure 39A. Both the wireless and wired NCS architectures of Figures 37A and 37B are applied outside the cage. For the wireless version, the passive signal is powered by the downlink signal, and the portion of the downlink signal is converted to a second harmonic as an NCS sense signal coupled into the body of the hamster. The reader Rx antenna is approximately 1.5 m from the tag. For the wired version, the sense antenna is mounted on the right side of the sleep zone, which directly transmits the NCS sense signal. In Figure 39B, the vital signs of their breathing and heartbeat are obtained without the hamster's attention. The amplitude of the vital signs is normalized to the maximum of the entire recording period. The illustration shows the details of the heartbeat waveform about the eighth second. The waveform features are not only similar to the various heartbeats during recording, but are also similar to the waveform features in anesthetized rats in Figure 38B. After applying a moving average with a window size of 20, the minimum negative value of the waveform is extracted for the heartbeat time interval, as shown in Figure 39C. The hamster heartbeat interval is about 20% longer than the rat heartbeat interval. To demonstrate the applicability of NCS to birds, a pet ornamental parakeet (Budgerigar, also known as Acacia parrot) named "Banana" was measured, as shown in Figures 39D-39D. Surveillance of vital signs in conscious birds can enable behavioral research and new capabilities in bird health checks. Figure 39D illustrates a wired experiment setup. The harmonic Tx antenna transmits an NCS sensing signal that is coupled into the body of the long-tailed parrot. The harmonic Rx antenna is also integrated into the perch and is capable of acquiring detailed features of both the heartbeat and the breath as shown in Figure 3E with the extracted heartbeat time interval in Figure 3F. Antenna deployment is very convenient for wired NCS setup. After a simple observation of the behavior of the long-tailed parrot, several locations on the perch that the bird usually stands on are identified. Select one of the most frequent locations and install the antenna. During the NCS measurement, the circadian rhythm of the pet long-tailed parrot is not disturbed.

The next demonstration is performed on a Russian pet turtle (four-clawed tortoise, also known as Horsfield's tortoise or Central Asian tortoise) named "Blimp", as shown in Figures 39G-39I. The Russian tortoise is the first vertebrate species in the lunar orbit and is important for the study of long space travel due to its sub-hibernating ability, which makes the continuous long-term record of vital signs even more scientific. . Sea turtles have a body structure that is separated by a small air gap and one of the hard shell and soft body tissue. It is possible to obtain vital signs by ECG and ultrasound only when using a neck or muscle probe, which leads to the main discomfort of the animal. The physiological characteristics and external temperature of the shell, together with the lack of appropriate diagnostic methods, make turtle emergency care very difficult. The heart of birds and rodents has four lumens (similar to humans) containing two atriums and two ventricles, and thus the NCS waveforms in Figures 38B, 39B, and 39E have similar characteristics. In contrast, tortoises have a three-chambered heart with two atriums and one ventricle, and the NCS electrocardiogram has different characteristics, yet the heart rate can still be accurately captured. The NCS settings are shown in Figure 39G. One of the similar antenna pairs in Figure 39D is placed under the cedar wood piece outside the glass cage. Due to the body structure, the breathing and heartbeat signals are embedded in the raw data of the NCS amplitude in Figure 39H because there is no shell surface movement during breathing. The heartbeat signal can be clear (as shown in the light shaded areas in the illustration), but during the breathing cycle it will be covered by a relatively large volume of strong respiratory signals (dark shaded areas) of the lungs. To separate the two overlapping signals for accuracy estimation, a continuous wavelet transform (CWT) is employed to extract the peak characteristics of the two waveforms in Figure 39I. Attempts have also been made to place the antenna directly on the vertebral canal and plastron, which also clearly records the heartbeat and respiratory signals of the NCS.

Last but not least, the RF signal can operate in a short range of water, especially near-field coupling. Although ECG telemetry is possible, the instrumentation procedure is only suitable for larger fish with a one-week recovery period before an unbiased measurement can be performed. Fish scholars have long explored other non-invasive solutions to capture the physiological information of behavioral research and evolutionary biology in small fish. The vital signs of a pet colorful boa (salmon, also known as Siamese squid) named "Glee" are shown in Figures 39J to 39L. The demodulated variable NCS phase signal is interpreted as the pectoral fin movement in Figure 39K, which has a periodic waveform containing one of the frequencies confirmed by the asynchronous video recording. It is believed that the demodulated NCS magnitude signal in Figure 39L represents a periodic waveform derived from a heartbeat. Hardware configuration

As shown in Figures 37A and 37B, the harmonic reader is implemented by a software defined radio (SDR). Use a combination of Ettus USRP X310 and UBX 160MHz RF daughter boards. During our experiments, the lower Ettus USRP B200/210 can also be configured as a harmonic reader for this application, but the X310 has a high data sampling rate for high resolution of waveform details. To operate as one of the tuned wave transceivers of Figures 37A and 37B, the Tx chain and the Rx chain share the same RF clock source. The synthesizer in the Rx chain configures the local oscillator (LO) frequency to be twice the OL frequency in the Tx chain. The Tx baseband signal is generated by a field programmable gate array (FPGA), and the Rx baseband is fed into the FPGA for demodulation for data recording and display.

Both the wireless and wired versions shown in Figures 37A and 37B are applicable to NCS. In the wireless version of Figure 37A, passive harmonic tags are designed with the modification of the Wireless Identification and Sensing Platform (WISP) and the Nonlinear Transmission Line (NLTL). NLTL is a trapezoidal structure of inductors and varactors, so varactor symbols are used to represent harmonic labels and harmonic generators. Harmonic tags operating at the acquired RF power can be easily deployed in large numbers on many sensing targets or at multiple points of the same target. At the same time, for single point monitoring, the wired version of Figure 37B is easier to apply to indoor animal laboratories. The harmonic generator here can still be like the previous NLTL to provide high conversion efficiency. The low pass filter at the input of the NLTL isolates the direct harmonic reflection from the NLTL to the reader, and the high pass filter at the output of the NLTL attenuates the fundamental frequency signal to the Tx antenna. However, any active or passive frequency multiplier with an appropriate frequency response can replace the NLTL in the wired version without the strict power constraints as in passive tags. During the experiment, try to use the custom diode and commercial passive frequency multiplier (CRYSTEK CPPD-0.85-2) as a harmonic generator, both of which provide satisfactory performance. The wired NCS system can also be extended to multiple points using reader CDMA technology, but the system cost will increase proportionally. Demonstrate the benefits of harmonic systems in our previous work. As a brief overview, the harmonic system provides very good isolation between the Tx chain and the Rx chain, so the Rx noise floor can be very low to improve both SNR and reader Rx sensitivity. In turn, the Tx power can also be very low to still maintain the SNR required for vital sign sensing, which eliminates any further health concerns regarding RF power directed to living tissue. Software configuration

Control the SDR with a computer with a LabVIEW interface. The sampling rates of the DAC (digital to analog converter) and ADC (analog to digital converter) are configured to be at 10 MSps. The frequency of the baseband output from the DAC is 1 MHz. When the Tx LO frequency is 950 MHz, the signal output from Tx is 951 MHz. After harmonic conversion, the center frequency of the Rx signal will be 1902 MHz. The Rx LO at 1900 MHz is set to twice the Tx LO. Therefore, the Rx baseband frequency is 2 MHz to be sampled by the ADC. The digital baseband signal is downconverted and the sampling in LabVIEW is reduced to an NCS sampling rate of 5 kSps.

While the invention has been described with respect to the specific embodiments of the embodiments of the invention Accordingly, the invention is considered to be limited only by the scope of the appended claims.

100‧‧‧ method

103‧‧‧ Provided

106‧‧‧Detection

109‧‧‧Measure

112‧‧‧Filter

115‧‧‧ Provided

118‧‧‧Detection

121‧‧‧Measure

124‧‧‧Filter

127‧‧‧

In order to more fully understand the nature and purpose of the present invention, reference should be made to the accompanying drawings, in which: FIG. 1A depicts one of the CST Microwave Studio simulations of vital signs transmitted by radio transmission by Near Field Coherent Sensing (NCS). A model in which the field model is used to represent the actual portion of the co-polarized electric field in a body simulation model of heartbeat sensing. 1B depicts a CST Microwave Studio simulation model of vital signs transmitted by radio transmission through the NCS, where the field model shows the actual portion of the co-polarized electric field in one of the wrist sensing models for pulse sensing. 2A depicts an NCS simulation of a conceptual skin and tissue structure with one of the adjacent vessels, wherein a dipole antenna that emits a 1.85 GHz, 0 dBm signal is placed over the skin without direct skin contact. The tissue above the blood vessels contains skin, fat and muscle. Geometric changes in the pulse vessel will modulate the near field and change the far field backscatter pattern. 2B shows an antenna reflection parameter S ₁₁ for the quasi-static blood vessel of FIG. 2A with cross-sectional marks t1, t2, and t3. 3A shows a graph in which the simulated vibration amplitudes and the samples at the far field points of FIGS. 1A and 1B are compared. 3B shows a graph in which the analog comparison of FIGS. 1A and 1B and the vibration amplitude of the sampled reflected by the antenna ₁₁ of the scattering parameter S represented. Figure 4A shows a simulated RF radiation pattern for a heartbeat where the far field sample point is 1 m in front of the chest. Figure 4B shows an analog RF radiation pattern for a wrist pulse where the far field sample point is 1 m above the wrist. 5A is a diagram of one of passive harmonic RFID tags for near field coherent sensing, in accordance with an embodiment of the present invention. Figure 5B is a diagram of one of the harmonic RFID readers in accordance with one embodiment of the present invention. Figure 6A shows the waveform after demodulating one of the original breathing signals and low pass filtering from an exemplary harmonic RFID system. 6B shows demodulation from an exemplary harmonic RFID system to one of the original heartbeat signals and one of the average heartbeats within the ten second moving window. The marker shows the measurement from the OMRON blood pressure monitor mounted on the left forearm. Figure 6C shows the waveform of one of the heartbeats during the three minute data collection period. Figure 6D shows the waveform of one of the wrist pulses during the three minute data collection period. Figure 6E shows a dynamic time warp ("DTW") waveform analysis showing the distance of each waveform of the heartbeat (the illustration shows the box must be distributed). Figure 6F shows a DTW waveform analysis showing the distances of the various waveforms of the wrist pulse (the illustration shows the box must be distributed). Figure 6G shows the median distance waveform and the maximum distance waveform compared to the respective DTW templates for the heartbeat. Figure 6H shows the median distance waveform and maximum distance waveform compared to the respective DTW templates for the wrist pulse. Figure 7A shows the pulse transit time ("PTT") estimated from the synchronized heartbeat and wrist pulse waveforms. The illustration shows the signal and one cycle of the extracted PTT. Figure 7B shows the probability density distribution of PTT within 3 minutes. Figure 7C shows the blood pressure drawn from the PTT when the subject is sitting down. The star mark shows the blood pressure measured from a commercial blood pressure monitor. Figure 7D shows the blood pressure drawn from the PTT when the subject is experiencing a modest activity and standing. The star mark shows the blood pressure measured from a commercial blood pressure monitor. Figure 8A is a diagram showing one of the conventional RFID systems of an RF front end of an RFID reader that interacts with the tag by backscattering. Figure 8B is a diagram showing one embodiment of the disclosed principles of harmonic RFID back reflection, where Tx represents a transmitter, Rx represents a receiver, LPF represents a low pass filter, and HPF represents a high pass filter. Figure 9A is a photograph of one of the RFID sensor tag wafers of one embodiment of the present invention integrated with an embroidered antenna on a fabric. 9B is a photograph of one of the PCB prototypes of one of the harmonic RFID tags in accordance with an embodiment of the present invention. Figure 10 shows the positioning error of a CDMA (200 tags) one harmonic RFID tag with various system SNRs. Figure 11 shows the sampling rate of harmonic backscattering of CDMA with various SNRs, where the number of tags is within the read range. When the chip code length is proportional to the number of tags, the solid line is from CDMA. The dashed line represents an embodiment with a more efficient semi-orthogonal code. Figure 12A is a diagram showing one of the antenna reflection systems in accordance with an embodiment of the present invention. The splitter is bidirectional and can be replaced by a wideband circulator. The harmonic transceiver and harmonic generator are similar to the harmonic transceivers and harmonic generators used in the passive tag system of Figures 5A and 5B. Figure 12B is a diagram showing one of the measured heartbeat waveforms using the antenna reflection system of Figure 12A. Figure 13 shows calibration results when a near field coherent sense antenna is placed near an asymmetric axis DC motor in accordance with an embodiment of the present invention. The demodulated sinusoidal waveform is a 2D phase projection of a 2D periodic asymmetric rotation. Figure 14 is a cross-section of an electromagnetic wave simulation of a 1.8 GHz co-polarized electric field coupled to the near field of an antenna of a heart motion. Figure 15 is a diagram showing one of the curves of the raw data rotated by the collected motor of the label 1 and the bandpass filtered data. Figure 16 is a diagram showing one of the raw data collected by the tag 2 for one of the heartbeat waveforms having a sampling frequency of 500 Hz. Figure 17 is a diagram showing one of the heartbeat waveforms after the 1.4 Hz to 15 Hz bandpass filtering extracted from Figure 16. Figure 18 is a cycle of the heartbeat waveform extracted from Figure 16, showing the characteristic points. Figure 19 shows one of the signals recorded from a chest label (solid line) and a wrist label (dashed line). The timing difference gives an estimate of blood pressure. Figure 20 is a cycle of the heartbeat waveform and wrist pulse waveform extracted from Figure 17, showing two C-peak points. The delay between peak 1 and peak 2 is about 0.074 s, which translates into a diastolic pressure of about 88 mmHg. Figure 21 is a flow chart showing one of the methods in accordance with another embodiment of the present invention. Figure 22A is a schematic illustration of one of the harmonic RFID backscatter tags. Figure 22B is a photograph of one of the PCB prototypes of the label depicted in Figure 22A. Figure 23 is a photograph showing one of the experimental NCS settings. A first label is in the chest area and a second label is on the left wrist. The sensed waveform is displayed on the screen. The chest strap and wrist strap are only convenient for deployment here. No skin contact or belt tension is required. 24A through 24B are heartbeat and respiratory waveforms extracted from the chest tag NCS signal. (A) The amplitude of the NCS signal is mainly modulated by the heartbeat. The top curve is derived from the normalized amplitude, which passes through the digital BPF (0.8 Hz to 15 Hz) to give the bottom curve. (B) The phase of the NCS signal is mainly modulated by respiratory motion. The top curve normalizes the original phase, which passes through the digital BPF (0.1 Hz to 1.2 Hz) to give the bottom curve. The isolation between the amplitude and phase modulation in the quadrature receiver results in a clear separation between the respiratory waveform and the heartbeat waveform. 25A to 25C are (A) synchronized time domain NCS (solid line) and ECG (dotted line) heartbeat signals. The sampling rate of the NCS is 5,000 Sps, and the ECG initially has 512 Sps but increases the sampling to 5,000 Sps. (B) Spectrum of the NCS signal, where the intensity is normalized to a peak of approximately 1 Hz and amplified to clearly show the lower intensity portion. (C) The spectrum of the ECG signal, where the intensity is also normalized to a peak of approximately 1 Hz and amplified to show the lower intensity portion. The high frequency component between 2 Hz and 8 Hz is more prominent than the high frequency component in the NCS. 26A to 26B are (A) time domain NCS signals after high frequency equalization. The triangular markers show sharp peak feature points for heartbeat time interval extraction. (B) Heartbeat interval of NCS and ECG signals. Dotted lines, dotted lines, and solid lines are from the original NCS (NCS1), ECG, and equalized NCS (NCS2). 27A-27B are NCS antenna effects: (A) An EM simulation model of a chest RFID tag in CST Microwave Studio. (B) Analog S ₁₁ parameters for the tag antenna. 28A-28B are analog EM power flows at 2 GHz coupled to the NCS signal in the body of the human body, with (A) having poor impedance matching and (B) having appropriate impedance matching. 29A to 29B are experimental results in various frequencies. (A) The NCS antenna is operated in air (solid curve) and S ₁₁ placed on the chest (dashed curve). (B) Demodulation of the heartbeat signal waveform from the chest tag in Fig. 4, which have different sensing frequencies. The NCS signals from the solid curve (impedance matching condition), the dashed curve, and the dotted curve correspond to the frequencies 1, 2, and 3 in Fig. 5A, respectively. Figure 30 is a photograph of one of the exemplary labels running on an Electronic Product Code (EPC) protocol using a conductive antenna using a conductive ink. Figure 31 shows the experimental results of heartbeat and respiratory waveforms extracted from a chest label using NCS. 32A to 32B are NCS signals when the large body movement starts from 40 s. (A) The amplitude of the NCS signal in the time domain. The illustration is a demodulated heartbeat waveform in the first 40 s of the test subject's sit-in. (B) NCS spectrum (dark gray) during periods of no movement for 10 s to 30 s and NCS spectrum (light gray) for periods of 85 s to 105 s with large body movements. 33A to 33C are extracted heart rates. (A) The third harmonic NCS signal for the heartbeat count (solid line) and the ECG signal (dashed line) for the heartbeat reference of approximately 97 s. (B) Count the heartbeat by NCS (solid line) and ECG (dashed line). The bottom line indicates the error (marked as ERROR). The illustration is the result of a period of 25 s to 55 s when large body movement occurs after 40 s. (C) Instant heart rate curve monitored by NCS (solid line) and ECG (dashed line). Figure 34 shows the NCS settings for breathing, heartbeat, and motion detection. A passive harmonic RFID tag is deployed in the chest region and the harmonic frequencies are backscattered to the reader antenna. Perform real-time demodulation and data analysis on the harmonic reader implemented by SDR. Figure 35A is a graph of one of DC filtered amplitude and phase data. Figure 35B is a graph showing one of the breaths. Figure 35C is a graph showing one of the heartbeats. Figure 35D is a graph showing one of the heartbeats of peak detection. 36A to 36C are graphs showing heartbeat and corresponding motion detection obtained using NCS. Figures 37A-37B are two settings for Near Field Coherence Sensing (NCS) for vital signs of conscious small animals. The signals are collected by a harmonic reader and digital baseband processing is performed using a field programmable gate array (FPGA) and a microcontroller (MCU). (A) A wireless NCS system is implemented by a harmonic RFID system with passive harmonic sensing tags. (B) The wired NCS system replaces the wireless link between the reader and the harmonic tag with an RF cable to reduce interference and facilitate indoor laboratory deployment. Figure 38A is a photograph of one of the experimental settings of an anesthetized rat using one of synchronized NCS and ECG measurements. Figure 38B shows a 5 minute data record from the experimentally set NCS and ECG of Figure 38A. The illustration shows the waveform details for a selected half-second duration. Figure 38C shows one of the heartbeat time intervals extracted from the NCS, which shows a close match with one of the ECG signals. Figure 38D shows a representative respiratory signal that is demodulated from the phase of the NCS signal at about the third to fourth seconds. Figures 39A-39L show vital sign monitoring of conscious small animals using a non-invasive NCS setting. (A) Experimental setup for hamsters. (B) Demodulation of the heartbeat and respiratory signals from the NCS signal. The illustration shows the details of the heartbeat waveform about the eighth second. (C) Heartbeat time interval of approximately 30 seconds. (D) Wired NCS settings for long-tailed parrots. (E) Demodulation of the heartbeat and respiratory signals from the NCS signal. (F) The extracted heartbeat interval of approximately 1.5 minutes. (G) For an NCS setting similar to the four-clawed tortoise of Figure 39D, where the antenna is below the wood chip floor. (H) The normalized raw amplitude of the three minute NCS signal, which is indicative of both breathing and heartbeat due to the shell structure. The illustration shows the waveform details of the overlapping signals. A lightly shaded section indicates a heartbeat, while a darkly shaded section indicates a breath. The strong breathing signal will cover the heartbeat signal during the overlap. (I) The signal processed on the continuous wavelet waveform (CWT) during the same period as (H) exhibits a clear separation of heartbeat and respiration for accurate rate estimation. The inset shows extracted waveform details that clearly indicate each peak in the heartbeat and breath. (J) An NCS setting similar to that of Figure 39D for betta fish. Deploy Tx and Rx antennas in the water near the fish. (K) Demodulation-induced NCS phase signal caused by pectoral fin movement, where the inset shows waveform detail. (L) Demodulation-induced NCS magnitude signal caused by heartbeat. Figure 40 is a drawing depicting one of the systems in accordance with another embodiment of the present invention.

Claims (25)

A method for non-contact measurement of physical and/or internal motion of a body, comprising: providing a first radio frequency ("RF" in a near-field coupling range of one of the first motions to be measured </ RTI> sensing a signal to generate a first measurement signal as the first sensing signal modulated by the first motion; detecting the first measurement signal; and measuring the first measurement signal based on the first measurement signal The first movement. The method of claim 1, wherein the first sensing signal is an ID modulation signal. The method of claim 1, wherein the first sensing signal is an active radio link or a backscattered RFID link. The method of claim 1, wherein the first sensing signal is provided by a wireless tag, and the method includes providing a downlink signal to charge the wireless tag. The method of claim 4, wherein the first sensing signal is one of the harmonics of the downlink signal. The method of claim 1, wherein the first measurement signal is transmitted through the detected far-field radiation after passing through a source of the first motion. The method of claim 1, wherein the first measurement signal is detected to be reflected from a source of the first motion as an antenna reflection. The method of claim 1, wherein measuring the first motion further comprises filtering the first measurement signal to obtain a first motion signal through timing and waveform. The method of claim 2, further comprising: providing a second RF sensing signal to generate a second measurement signal in a near field coupling range of one of the second motions to be measured; detecting the second amount Measuring a signal; and measuring the second motion based on the second measurement signal. The method of claim 9, further comprising measuring the first motion and the second motion measurement based on the synchronization measurement. A system for measuring motion of a body, comprising: a first signal source for generating a first sensing signal; a first antenna electrically communicating with the first signal source and wherein The first antenna is configured to be disposed in a near field coupling range of one of the first motions to be measured to generate a first measurement signal as the first sensing signal modulated by the first motion; And a first receiver for detecting a first measurement signal. The system of claim 11, wherein the first antenna is configured to be disposed within a coupling range of a heart motion, a pulse, a respiratory motion, a bowel motion, or an eye motion. The system of claim 11, wherein the first sensing signal is an ID modulated wave. The system of claim 11, wherein the first sensing signal is an active radio link or a backscattered RFID link. The system of claim 11, wherein the first source and the first antenna are configured as a wireless tag, and a tag reader is configured to transmit a downlink signal to the wireless tag. The system of claim 15 wherein the first receiver is part of the tag reader. The system of claim 15, wherein the wireless tag is configured to be charged by the downlink signal. A system as claimed in claim 15 wherein the frequency of one of the first sensed signals is one of the frequencies of one of the downlink signals. The system of claim 15, wherein the wireless tag modulates the downlink signal using an orthogonal ID such that the first sensing signal is a CDMA signal. The system of claim 11, wherein the first receiver is configured to detect the first measurement signal as a transmission signal. The system of claim 11, wherein the first receiver is configured to detect the first measurement signal as a reflected signal. The system of claim 11, further comprising a filter in communication with the receiver, wherein the filter is configured to demodulate and filter the first measurement signal to obtain a first motion signal. The system of claim 22, wherein the filter is programmed to sample, demodulate, and filter the first measurement signal to derive a motion processor. The system of claim 11, further comprising: a second signal source for generating a second sensing signal; a second antenna in electrical communication with the second signal source and wherein the second antenna Configuring to be disposed in a near field coupling range of one of the second motions to be measured to generate a second measurement signal as the second sensing signal modulated by the second motion; and wherein the receiving The device is further configured to detect the second measurement signal. The system of claim 24, further comprising a processor for measuring a derivative value based on the detected coupled signal and the second coupled signal.