Frequency Modulated Continuous Wave Radar Overcomes the Normal Inability

Human-aware localization using linear-frequency-modulated continuous-wave radars

J.-M. Muñoz-Ferreras , ... C. Li , in Principles and Applications of RF/Microwave in Healthcare and Biosensing, 2017

5.7.5 Clutter in the Same Range Bin

LFMCW radar sensors can easily deal with unwanted targets situated in range bins in which the desired targets are not located. This is why an improvement in the range resolution is normally pursued, so that this range-isolation capability can be further enhanced. Unfortunately, depending on the application, it is sometimes unavoidable to have clutter in the same desired range bin (e.g., if the vital signs of a person are to be monitored, it is almost impossible not to receive echoes from undesired parts, such as head or limbs). If the radar is coherent, the new dimension Doppler can be exploited to devise specific methods which try to mitigate this in-cell unwanted clutter (see Section 5.6).

Consider the scenario shown in Fig. 5.22, which is analogous to the one depicted in Fig. 5.14, with the inclusion of a stationary clutter scatterer at the same distance of the wanted sinusoidally-vibrating target. The radar prototype is assumed to have a center frequency of f _c=5.8   GHz. The transmitted bandwidth is B=160   MHz and the PRF is 500   Hz, whereas the CPI is 12   seconds. The mean range to the targets is R _c=5.625   m, whereas the sinusoidal motion of the desired target is defined through its range amplitude (peak-to-peak R _pp=50   cm) and its frequency (f _t=0.5   Hz). The amplitude of the desired target is K _d=1, whereas the amplitude of the clutter scatterer is K _c=1.2.

Figure 5.22. Simulated scenario for a point-scatterer target sinusoidally vibrating along LOS and a stationary clutter return.

The slow-time signal for the range bin in which the targets are present can be written as

(5.47) $s_{\mathrm{b}}(\tau )=K_{\mathrm{d}}\cdot \exp \left(j\frac{4\pi f_{\mathrm{c}}}{c}\left(R_{\mathrm{c}}+\frac{R_{\mathrm{pp}}}{2}\sin (2\pi f_{\mathrm{t}}\tau )\right)\right)+K_{\mathrm{c}}\cdot \exp \left(j\frac{4\pi f_{\mathrm{c}}}{c}{R}_{\mathrm{c}}\right).$

The second term in Eq. (5.47) is the zero-Doppler clutter component, whereas the first term corresponds to the wanted signal, which has a Doppler frequency, according to Eq. (5.24), given by

(5.48) $f_{\mathrm{Dop}}(\tau )=-\frac{2\pi f_{\mathrm{c}}{f}_{\mathrm{t}}{R}_{\mathrm{pp}}}{c}\cos (2\pi f_{\mathrm{t}}\tau ).$

For the simulation example in this section, the maximum Doppler frequency can be derived from Eq. (5.48) and turns out to be 30.4   Hz. This value corresponds to the result in Fig. 5.23, which shows the spectrogram for the slow-time signal in the corresponding range bin. The sinusoidal Doppler history associated with the moving target (see Eq. 5.48), together with the zero-Doppler component corresponding to the clutter scatterer, can be seen in Fig. 5.23.

Figure 5.23. Spectrogram (Doppler vs slow-time) for a vibrating target and a stationary clutter scatterer.

The Fourier transform of the slow-time signal Eq. (5.47) is depicted in magnitude in Fig. 5.24. This figure permits the corroboration of the spectrum distribution shown in Fig. 5.12. It becomes clear that a high-pass filter in the slow-time τ permit the mitigation of the clutter zero-Doppler component at the cost of some distortion added to the desired Doppler spectrum.

Figure 5.24. Doppler spectrum of the simulated slow-time signal for the scenario in Fig. 5.22.

The application of the mitigation technique (a fifth-order Chebyshev type-I filter with a 0.5-dB ripple in the passband and a 40-dB attenuation in the 2-Hz-edge-frequency stopband has been employed in this example) permits the recovery of the desired sinusoidal motion of the vibrating scatterer, as shown in Fig. 5.25. Before the mitigation approach of Section 5.6.2, the clutter influence on the desired range history is enormous.

Figure 5.25. Tracked motion pattern for a desired vibrating scatterer before and after applying a clutter-mitigation high-pass filter (R _pp=50   cm).

A more compromised simulation considers that the peak-to-peak amplitude is reduced. Assume now that R _pp=10   mm. For this situation, the maximum Doppler is only 0.61   Hz, which leads to a desired Doppler spectrum almost identical to DC. Fig. 5.26 shows the range history extracted before and after applying a high-pass filter with cut-off frequency of 0.2   Hz. The ground-truth motion of the wanted target is also shown as a reference. It is true that the sinusoidal pattern is observed before the mitigation technique. However, the Doppler filtering amplifies the displacement amplitude, which may help with the detection of the motion. In any case, due to the added distortion, the amplitude of the estimated range evolution after clutter mitigation does not correspond with the ground-truth 10-mm peak-to-peak amplitude.

Figure 5.26. Tracked motion pattern for a desired vibrating scatterer before and after applying a clutter-mitigation high-pass filter (R _pp=10   mm).

As a conclusion, the detection of slow targets when competing with strong stationary clutter is of great difficulty, and any radar in that scenario (e.g., maritime radars trying to detect small and slow boats under heavy clutter) sees its performance degraded.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128029039000059

Radar-based vital signs monitoring

Jingtao Liu , ... Changzhan Gu , in Contactless Vital Signs Monitoring, 2022

9.3.2 Analysis of an FMCW radar IF signal

For FMCW radar, a variety of modulations is possible. The transmitter frequency can vary up and down as follows: sine wave, sawtooth wave, triangle wave, square wave, etc.

Sawtooth modulation is the most used in FMCW radars, so the following analysis is based on the sawtooth wave.

The sawtooth wave of an FMCW radar is a frequency linear-modulation method. The frequency of electromagnetic waves changes linearly with time. The schematic diagram is shown in Fig. 9.12. Fig. 9.12(a) shows the time-domain representation. The transmitted signal in a cycle is usually called a chirp. Fig. 9.12(b) shows the frequency-time diagram of a chirp.

Figure 9.12. Block diagram of a typical FMCW radar system.

Denoting $T_c$ as the chirp repetition period, B is the bandwidth of the chirp, and $S=B/T_c$ is the slope of the frequency modulation. The mathematical expression for the transmitted signal within one frequency ramp interval is:

(9.15) $s_{\mathrm{Tx}}\left(t\right)=\exp \left(\mathrm{j}(2\pi f_{c}t+\pi \mathrm{S}{t}^2+\varphi )\right),$

where $f_c$ is the center frequency of the frequency ramp, ϕ is the initial phase residual, and $t\in [-T_c/2,T_c/2]$ . Suppose that there is a reflection point, and its distance from the radar as a function of time is $R(\tau )$ . Assuming that the movement of the scattering point is relatively slow, $R(\tau )$ can thus be regarded as a constant within a certain period. This is a "stop-and-go" hypothesis, which is extremely common when dealing with slow-moving targets. Therefore, for the scattering point located at $R(\tau )$ , the echo signal received by the FMCW radar is a function of the time delay Δt and a certain amplitude attenuation σ of the transmitted signal. Among them, the time delay is:

(9.16) $\mathrm{\Delta }t=\frac{2R\left(\tau \right)}{c}$

Therefore, the echo signal can be expressed as:

(9.17) $s_{\mathrm{Rx}}\left(t\right)=\sigma s_{\mathrm{Tx}}(\mathrm{t}-\mathrm{\Delta }t)=\sigma \cdot \exp \left(\mathrm{j}(2\pi f_c(t-\mathrm{\Delta }t)+\pi \mathrm{S}{(t-\mathrm{\Delta }t)}^2+\varphi )\right).$

According to Fig. 9.11, the received signal is mixed with the transmitted signal. After that, the resulting mixed signal is low-pass filtered and the intermediate frequency signal is obtained:

(9.18) $s_{\mathrm{IF}}\left(t\right)=f_{LPF}\{s_{\mathrm{Tx}}\left(t\right)\cdot s_{\mathrm{Rx}}\left(t\right)\}=\sigma \cdot \exp \left(\mathrm{j}(\frac{4\pi SR\left(\tau \right)t}{c}+\frac{4\pi f_{c}R\left(\tau \right)}{c}-\frac{4\pi S{R}^2\left(\tau \right)}{c^2})\right).$

It can be seen from Eqs. (9.16)–(9.18) that the obtained intermediate frequency signal is a sinusoidal motion with frequency $f_{IF}$ :

(9.19) $f_{IF}=\frac{2SR\left(\tau \right)}{c}.$

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128222812000184

UWB Chipless RFID Reader: State of the Art

Marco Garbati , ... Romain Siragusa , in Chipless RFID Reader Design for Ultra-Wideband Technology, 2018

2.4.4 Conclusion

An FMCW solution with a reduced sweep time is found to be difficult not only with respect to hardware, because of the UWB VCO's limited speed, but also in terms of tag characteristics. Indeed, a frequency-coded tag implemented with the REP technique is no more than a bunch of parallel passive filters. If the sweep in frequency is too fast, these filters wil not have the time to charge themselves and reach the stable state, and the response will be lower in amplitude, as demonstrated in [KEY 16a]. An SFCW reader can be used to read both time- and frequency-coded tags where the reduced P _R can be compensated by a higher sensitivity compared with an IR-UWB reader. In contrast, the long sweep time will translate into high reading time, especially in the presence of sweep averaging [KOS 12]. An IR-UWB architecture is preferable to design a UWB chipless RFID reader that is compliant with international regulations and capable of reading both time-and frequency-coded tags. It has a higher Pr and a reduced reading time due to the absence of any frequency sweeping in transmission, which makes it ideal for a real-time application. The hardware design of an IR-UWB reader has to be optimized in order to reduce its sampling noise (high input noise bandwidth).

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9781785482922500021

Time-Frequency Methods in Radar, Sonar & Acoustics

In Time Frequency Analysis, 2003

14.4.2 Signal Description

The radar used a linear frequency modulated continuous wave (LFMCW) waveform with a sweep (or waveform) repetition frequency (WRF) of 50Hz. A set of coherent measurements were collected, each of 256 sweeps or 5.12s duration. For hardware limitation reasons there was a short inter-dwell gap of approximately 12 sweeps $\left(12\cdot \frac{1}{50}S\right)$ where no data was recorded between each coherent measurement interval. The radar signal processor pulse compressed each sweep using stretch processing then formed 20 digital beams (10 for each arm of the "L" array). Doppler analysis of the 256 sweeps in each successive coherent processing interval was performed for each range cell in each beam direction.

The selection of WRF (50Hz) meant that Doppler measurements of what was a very high velocity target were ambiguous for most of the flight. The long (with respect to target kinematics) coherent integration time (CIT) increased radar sensitivity although the target acceleration decreased the coherent processing gain achieved and limited the accuracy of velocity measurements. The radar used existing software for range, beamforming and Doppler processing which assumed constant velocity targets and rather than modify the software to support accelerating targets it was considered that some form of post event acceleration analysis could be used to mitigate the accelerating target mismatch loss.

Received radar data was displayed in real-time during the experiment and also recorded on tape for subsequent analysis. The data for off-line analysis was range processed (sweep compressed) and beamformed but not Doppler processed. It was organized into a sequence of dwells of data where each dwell contained two sets of 10 formed beams (one set for each arm of the "L" array) with a group of 20 range cells of interest and the complex signal corresponding to each of the 256 sweeps in a CIT for each range and beam. Sequences of dwells were collected into a single file.

A typical range-Doppler map seen by the operator is shown in Figure 14.4.1. It corresponds to one beam in the direction of the target and 20 range cells stacked vertically. For each range cell the Doppler spectrum has been determined from the appropriate range samples of each of the 256 pulse compressed sweeps, with one spectrum per range cell. Although not shown, the operator actually sees 20 such figures per dwell where each one corresponds to a different beam direction.

Fig. 14.4.1. Range-Doppler map showing the accelerating target smeared in Doppler (from −22Hz to 0Hz) in range cells 6 and 7. The direct wave from the transmitter and ground clutter is visible surrounding 0Hz Doppler and centered in range cell 2. The coasting spent first stage of the two stage TBM can be seen at range cell 6 with 10Hz Doppler.

This particular example covers the time interval from launch time plus 16s to launch time plus 21s (T+16s: T+21s). The range-Doppler map shows the accelerating target smeared in Doppler (from −22Hz to 0Hz) in range cells 6 and 7. The target velocity is such that this is an ambiguous Doppler measurement. The coasting spent first stage of the two stage TBM can be seen at range cell 6 with 10Hz Doppler. The direct wave from the transmitter as well as ground clutter is visible surrounding 0Hz Doppler and centered in range cell 2. An injected calibration signal can be seen in range cells 2 and 3 at -25Hz and +25Hz. Receding targets appear to incorrectly have positive Doppler however this is a frequency inversion artifact of the hardware design of the radar.

It is clearly difficult to determine the true time-varying velocity of the target since it changes significantly during the radar CIT. We are interested in determining the instantaneous Doppler law and hence the time-varying velocity of the target throughout this and all other dwells which contain the target during the powered segment of flight.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780080443355500358

Systems and Applications

NICHOLAS FOURIKIS , in Advanced Array Systems, Applications and RF Technologies, 2000

1.6.5 LPI Radars

We have already seen that PD radars utilizing pulse compression techniques, monocycle radars, and spread-spectrum systems can have the LPI characteristic in the previous section. Ideally, designers would like to transmit as much power as is necessary to attain a certain radar range requirement without being detected by spectrum surveillance or electronic support measures (ESM) systems. The availability of anti-radiation missiles (ARMs) emphasized the need for LPI radars. The Gulf War certainly proved the usefulness of both the ESM systems and the ARMs on the one hand and the vulnerability of conventional radars on the other.

In this section we shall consider the following three types of LPIRs:

•

Frequency modulated continuous wave (FMCW) radars

•

FM interrupted continuous wave (FMICW) radars

•

Staring radars

In sections 1.6.6 and 2.5.7 we shall consider bistatic/multistatic radars, some of which have exceptional LPI capabilities.

1.6.5.1 FMCW and FMICW radars

The most straightforward LPI radar is the continuous wave (CW) radar, which has a 100% duty cycle. The radar can, in theory, perform the prime radar objectives of surveillance and detection and has LPI capability because its instantaneous power is low.

The most easily realizable CW radar is the frequency modulated (FMCW) radar where the transmitter signal is frequency modulated by a linear waveform [81,82]. The received signal has the same modulation but is delayed relative to the transmitted signal. If Δf is the peak-to-peak frequency deviation, SL is the slope of the ramp, and δf ₁, is the frequency difference between the transmit and receive ramps, the target range can be measured with a range resolution of δR given by

$\begin{array}{lll}R=\frac{\delta f_{1}c}{2(\text{SL})}\hfill & \text{and}\hfill & \delta R=\frac{c}{2\Delta f}\hfill \end{array}$

An excellent account of the developments related to FMCW radars is given in reference [83].

Acceptance of FMCW radar has been delayed because of the technological problems associated with its realization. More explicitly, the difficulties related to the realization of monostatic FMCW radars have centered around the isolation required between the transmit and receive EM waves. High isolation is required between the transmit and receive signals when the typical corresponding power levels are of the order of watts and picowatts, respectively. A satisfactory solution to this problem has been proposed and implemented [84].

Frequency modulated interrupted continuous wave (FMICW) radars resemble monostatic radars where the antenna is switched between the receiver and the transmitter. The system overcomes the difficulty of isolating the receiver from the transmitter and, provided certain conditions on the switching speed and the FMCW parameters apply, the essential characteristics of the FMCW are preserved [85].

1.6.5.2 Staring Radars

So far we have considered scanning radar systems in which one inertialess beam scans a given surveillance volume; all the power of the system is focused onto one pixel, which in turn is scanned through the surveillance volume. Scanning systems are therefore easily detected by radar warning receivers (RWRs) that work in conjunction with jammers capable of disabling the radar.

An LPI system is realized if the transmitted power is distributed evenly throughout the surveillance volume and contiguous staring beams receive the scattered radiation. As the power toward any direction is low, detection by a RWR is more challenging.

If P _T, M, and T ₂ are the power transmitted, the number of beams covering the whole surveillance volume, and the time it takes to complete the surveillance volume, respectively, the energy per pixel for the scanning system is equal to P _T(T ₂/M), and for the staring system is (P _T/M)T ₂. The two systems therefore provide the same energy per pixel. The overall effect upon detection, however, is primarily dependent upon the performance of the integration process employed and the target and noise statistics. While one can use coherent integration to ensure maximum efficiency, the target and noise statistics are application dependent.

The following comparisons between the scanning and staring systems can be made:

•

The information update rate for the staring system is higher than that corresponding to the scanning system.

•

The staring system uses parallel processing of the incoming information, while serial processing is used in scanning systems.

A small experimental radar that flood-illuminates the surveillance volume and utilizes staring receive beams has been reported [86].

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780122629426500032

RF and camera-based vital signs monitoring applications

Li Zhang , ... Hong Hong , in Contactless Vital Signs Monitoring, 2022

14.1.1 RF sensor

Measurement of human physiological signs based radar has been explored with three types of electromagnetic radar systems; continuous-wave, frequency-modulated (FM), and ultra-wide band (UWB). Many studies have focused on using Doppler radar at various frequencies, output powers, and distances.

Several pros are associated with electromagnetic radar-based methods, which can be listed as follows:

Reliable results with stationary objects have been achieved. Even if there is some non-metallic material, such as wood, clothing, glass, or water, between the radar and the subjects being measured, the heart and lung signals can be extracted. There is a possibility to extract the cardiorespiratory signal from medium and long distances (up to 30 m) [3,4].

While RF-based sensors have several pros over current vital signs monitoring, there are limitations, including:

When the movement of the surface of interest is very small (small Doppler movements) or superimposed on other movements (such as the subject's head and arms), the radar-based method is prone to motion artifacts and noise. Radar-based methods restrict subjects' movement and need an exposed Region of Interest (ROI) to analyse, which is inappropriate for long-term monitoring [12]. At distances of greater than one m, radar-based methods are prone to degradation due to the motion artifacts and noise caused by the increased free-space loss. For example, SNR decreases to about 50% as the distance of the subject from the electromagnetic radar sensor increases from 0.25 m to 2.5 m [21]. Detecting the vital signs of multiple people at the same time is prone to interference with radar signals. It is more difficult to obtain vital sign data from people with high body fat. While studies of slim people have produced usable data, detection of data from overweight people is more challenging [14].

Based on the limitations shown, radar would likely need to be supplemented with other technologies for a complete remote non-contact vital signs monitoring system.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128222812000238

Radar and Inverse Scattering

Hsueh-Jyh Li , Yean-Woei Kiang , in The Electrical Engineering Handbook, 2005

10.10.2 Frequency-Modulated CW Radar

The CW radar can measure the Doppler frequency of the target, but it cannot measure the target range. The frequency-modulated CW radar (FM–CW) can measure both the range and Doppler frequency of the target. In the FM–CW radar, the transmitted frequency changed as a function of time in a known manner.

Figure 10.4(A) shows the principle of triangular FM–CW ranging on a single target with no Doppler shift. The range information is contained in the frequency difference between the signal echo and the radar's present transmitting frequency. If there is no Doppler frequency, the difference frequency is a measure of the target range, which is given by:

FIGURE 10.4. Principle of Triangular FM–CW Ranging on a Single Target with Different Doppler Shifts

(10.35) $R=\frac{cT{f}_r}{2B},$

where B is the bandwidth of the transmitted signal, T is the period of the modulation wave, f_r is the frequency difference between the signal echo and the present transmitting signal, and c is the light speed.

If there is a Doppler shift, there is a received frequency-time relationship, as shown in Figure 10.4(B). There are two difference frequencies: the upper beat frequency, f_b (up), and the down beat frequency, f_b (down). The range frequency f_r and the Doppler frequency f_d can be extracted by:

(10.36) $\begin{array}{r}{f}_r=\frac{1}{2}[f_b(\text{up})+f_b(\text{down})].\\ f_d=\frac{1}{2}[f_b(\text{down})+f_b(\text{up})].\end{array}$

FM–CW radars can be used in airborne applications. For example, an FM–CW altimeter can be placed in the aircraft to measure height above the surface of the earth. A Doppler navigation radar can measure the vector velocity relative to the frame of reference of the antenna assembly. A Doppler navigation radar having forward and rearward beams is called a Janus system. With the Janus system, the angular displacement of the aircraft heading and the speed along the ground track can be measured.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780121709600500475

Pulsed Inductive Thermal Wave Radar (PITWR)

Yunze He , ... Ruizhen Yang , in Transient Electromagnetic-Thermal Nondestructive Testing, 2017

11.1.1 Background

In the field of photothermal radiometry, Mandelis et al. proposed thermal-wave radar (TWR), which combined linear frequency–modulated continuous-wave excitation and matched filtering processing. This was based on the Hilbert transform used to detect human dental demineralization lesions and osteoporotic bone loss [218], suggesting a significant improvement in the dynamic depth-resolution range of subsurface defects [217,235] 217 235 . This approach has since been introduced in thermography nondestructive testing. In 2014, J. Liu and coworkers used thermal-wave radar imaging (TWRI) for CFRP inspection [236]. Mulaveesala and coworkers highlight the Hilbert transform-based time-domain phase analysis scheme introduced for testing and evaluation of subsurface defects in a mild steel sample [237]. Mandelis and coworkers also proposed truncated-correlation photothermal coherence tomography (TC-PCT) based on CCMF and frequency doubler with exclusive-or (EX-OR) logic algorithm, which enabled three-dimensional (3D) visualization of subsurface features [238–240] 238 239 240 . Unlike other thermographies, TC-PCT is a coherence tomography, a 3D tomographic imaging method that intends to rival optical coherence tomography. As such, it stands apart from other thermographic methods that are not 3D building tomography. There are many advantages to this unique high-axial resolution technique, which breaks through the thermal diffusion limitations of other thermographic principles with an evolving temporal filter. However, the variable eddy current excitation signals and advanced radar processing methods have not been used in eddy current thermography to solve the discrepancy between dynamic range and depth resolution or to improve defect detectability.

In in situ applications, materials under test (MUT) may have oil, coating, or an oxidation layer on their surface. This significantly changes the thermal emissivity of the sample and introduces illusory temperature inhomogeneity, resulting in false alarms. Removing the influence of surface emissivity variation is vital, and several solutions were suggested in previous studies. Firstly, inductive excitation is an emissivity-independent way of subsurface heating depending on the induction frequency and the electrical properties of the MUT, which removes the effect of emissivity variation in the heating process. Aside from spraying water or paint, L. Bai et al. proposed a two heat balance states-based normalization method to remove the influence of surface emissivity variation [212]. Y. He and coworkers proposed eddy current pulsed phase thermography (ECPPT) to reduce the variation of surface emissivity [141]. However, the advanced CC phase was not investigated in eddy current thermography for suppression of emissivity variation, although the CC phase in TWR is a powerful emissivity-normalized parameter [217,218] 217 218 .

In this chapter, a powerful PI-TWR is proposed by introducing CC matched filtering in ECPT for: (1) subsurface defect evaluation and suppression of emissivity variation in steel, and (2) CFRP evaluation.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128127872000113

RF/wireless indoor activity classification

A. Rahman , ... V. Lubecke , in Principles and Applications of RF/Microwave in Healthcare and Biosensing, 2017

7.3.1 Sensing Techniques

Researchers have reported different types of short-distance radar for different applications. Ultra-wide band radar, continuous-wave (CW) radar, frequency modulated CW radar, and pulsed radar, to cite a few. Radars also vary based on the operating frequency (X band, K band, etc.). Directivity of sensing also a diversity; this diversity is mainly introduced by antenna design. However, the sensing principle is more or less unique regardless of the type of each radar: they all sense the back-scattered signal reflected from the target.

The single-channel radars are limited by their inability to produce displacement sensing in all positions. However, this limitation is overcome using a quadrature radar system, which in principle provides a stereo vision. The outputs of the quadrature radars are called in-phase (I) and quadrature phase (Q). The combination of I and Q can provide useful insight in characterizing a target's motion. Fig. 7.3 shows the experimental results of two different kinds of motion patterns of a linear moving platform. It is evident from the plots that distinguishing the patterns of I and Q waves holds the information regarding the motion patterns. Additionally, IQ plots are apparently different, indicating that these could be useful parameters in RF activity classification.

Figure 7.3. Time domain data and corresponding IQ plot, for Doppler radar measurement of sinusoidal motion in (A), and (B) shows the same for a series random linear movement.

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128029039000072

IRACON channel measurements and models

Sana Salous , ... Katsuyuki Haneda , in Inclusive Radio Communications for 5G and Beyond, 2021

3.2.2 Outdoor-to-indoor propagation

Building entry loss (BEL) refers to the attenuation of the signal as it propagates from outside to inside a building. This is particularly important in the mmWave band which exhibits significantly higher attenuation than below 6 GHz bands.

To study BEL as a function of frequency O2I measurements were conducted at 3, 10, 17, and 60 GHz [DCCS17b]. Both BEL and DS were estimated for various Rx positions in different rooms. The results reveal a strong variation in signal attenuation from 5 to 40 dB depending on the composition of the window material. Relatively low and non-frequency-dependent attenuation (around 5 dB) is observed for non-coated glass windows while a high attenuation (30 dB), increasing with frequency, is recorded for coated glass windows (see Fig. 3.12.) The estimated DS values are relatively low (below 30 ns for all the measurements) and uniformly distributed across the different frequency bands.

Figure 3.12. Building Entry Losses across the four bands [DCCS17b].

In [SMVB20], data sets at 26 GHz from three different measurement campaigns in the same site but with different experimental setups were used to characterize CL, BEL in isolated building and building entry loss over a cluttered path. The measurements were performed to estimate BEL by placing the transmitter: in front of the building on a street canyon, in front of the building on an open field, or at a high level in a cluttered environment with NLOS. The data were processed to estimate median values of the BEL. From the three sets of data the difference between the estimation of BEL from a transmitter placed outside in LOS to the facade of the building or from a transmitter above a cluttered environment with NLOS gave values which differ by less than 1.8 dB which indicates that either method can be used for estimating the median value of BEL. The median of clutter loss estimated from two different data sets was found to be 18 dB, which confirms that both experimental setups and the methodology are valid for this kind of measurements. With these results, it can be concluded that at 26 GHz, the median values of BEL and CL can be treated as multiplicative, i.e., the overall excess path loss is the sum in dB of the individual losses. A summary of the obtained results is shown in Table 3.6 [MVS19].

Measurements of BEL for a "traditional" building (non-coated window glass) for radio frequencies in the range 2-60 GHz and for different elevation angles by locating the transmitter and receiver at different floors at two orthogonal sections of a building are reported in [DMS18]. It is proposed to use a linear model to account for the angle of elevation dependence of BEL. $L_{ef}=k|\theta |(\text{dB})$ . The results give a value of k greater than 24 dB/90^∘ which is in contrast to the k value in the ITU and mmMagic models of 20 dB/90^∘.

Simulations and measurements using a VNA and directional horn antennas were used to study infra-red reflector glass over a frequency range 11-67 GHz [BCX17]. Different situations are analyzed specifically to assess the losses due to diffraction, penetration and the possible exploitation of gaps in the window pane by placing the antennas on either side of the window. The results showed that the penetration loss through the glass is in excess of 30 dB which is further exacerbated by the problem of angle of incidence. The opportunity to exploit gaps in the building infrastructure was shown to reduce the penetration loss in excess of 10 dB. This benefit is greater as frequency increases, though there is more vulnerability to scattering where it is more beneficial to rely on reflection from the far end of a window as opposed to the diffraction off the near end. (See Fig. 3.13.)

Figure 3.13. Analysis of gap effects in the range a) 22-30 GHz and b) 50-67 GHz.

To measure BEL in a typical thermally efficient building, and a traditional building measurements were performed between 0.4 to 73 GHz using two custom channel sounders based on the frequency swept continuous wave (FMCW) architecture [TSRC18]. BEL was then estimated by taking the difference between the received power inside the house and the mean power of the locations measured outside the house. Directional antennas were used outside the building as transmitters and omnidirectional antennas as receivers inside the two types of houses. The CDF of the BEL did not display significant frequency dependence between 25 and 73 GHz. Comparative measurements indicated that opening external windows reduces the penetration loss by 10 dB in the millimeter wave bands between 25-73 GHz. A difference of 2-16 dB between the two types of build is measured across the entire frequency range. (See Fig. 3.14 and Table 3.5.)

Figure 3.14. CDF for the traditional house for the higher frequency bands [TSRC18].

Table 3.5. Measured BELs values.

Penetration Loss [dB] Average	Omni. Dir.			Omni. Dir.
	3 GHz	10 GHz	17 GHz	17 GHz
Corridor	20	20	21	21
Flat	27	30	33	35
Rest room	11	9	8	9
Large Office 1	14	17	18	18

Delay Spread [ns] Average	3 GHz	10 GHz	17 GHz	17 GHz
Corridor	37	34	37	34
Flat	41	38	39	35
Rest room	33	30	36	38
Large Office 1	29	31	31	31
Large Office 2	49	42	42	40

Table 3.6. Measurement campaigns, experimental setups and results.

Campaign			Measured values [dB]
	Experimental setup	Losses characterized	No GTx correction	GTx correction
Measurements 2019	Durham University	BEL over cluttered path, NLOS	44.03	42.6
	Channel Sounder	BEL isolated building, LOS	45.9	42.9
		Clutter loss	18.07	18.07
Measurements 2018	CW + downconverter	BEL isolated building, LOS	44.27
Measurements 2017	CW at 26 GHz	Clutter loss		18.21

•

Strong variations between 10 and 40 dB depending on the room

Read full chapter

URL:

https://www.sciencedirect.com/science/article/pii/B9780128205815000092

fosterenwhat1944.blogspot.com

Source: https://www.sciencedirect.com/topics/engineering/frequency-modulated-continuous-wave

Share this post