Thank you for the comprehensive explanation. I'm embarking on my research in the field of beamforming, and your insights have proven invaluable. As a newcomer to this field, would you be able to offer me some suggestions or guidance to help me get started?
Thanks for this great illustration, Emil! The great example of body loss. I have seen the statement that body losses at same 26.5GHz can be as high as 30 dB, In the same time additional hand losses stated 0-5dB and 3dB was assumed for most cases of link budget planning maybe it is up to positioning and the fact that there is quite low possibility of completely blocking one of three antennas in smartphone. BTW do you expect any PAAMs mmW to be integrated in smartphones near future to mitigate UL challenges with coverage?
As far as I’ve understood, the smartphones that support mmW already use phased arrays. For example, iPhones 12-14 in the USA have two modules that are place at different sides to limit the risk that both are covered by hands. These phones are said to support 4-MIMO so I would guess these are two phased array that support dual polarization.
@@WirelessFuture I think there is one commercially used phased array in the market QTM527 from Qualcomm and it is too big and power hungry to be used in smartphones. It is used in bigger FWA CPEs where EIRP can be as high as ~45dBm. I haven`t seen any smartphones supporting 4 layers MIMO right now - 2 layers are commonly used: vertical and horizontal polarization, typical smartphone EIRP for mmW is on the level of 13.5dBm (28GHz) (QTM525) and it is 1T2R maybe using two or three antennas in smartphone but only one is active during PRB. This makes UL the limiting factor for coverage. So the development on the smartphone side would practically improve the mmW coverage a lot.
Hi, very interesting. Can you measure something more. Leaves from trees would be nice and what happens when you increase the distance and put something between. Will the loss be the same ?
The propagation loss (with distance) and penetration loss (by an object) are two independent effects, so they always add up when using the decibel scale. I have returned the hardware to TMYTEK, but I will keep your suggestion in mind if I get my hands on it again!
Very nice demonstration, thank you for this video! Regarding the loss seen with the iphone, could it also be partially due to the hand holding the device? It seems that some of the fingers holding up the phone are almost at the same height as the TX and RX antennas but it may just be the camera angle!
It is mostly the camera angle that is tricky because I only touched the edge of the phone. I don't think my hand had a major impact. I did the same experiment while holding the phone from the top, and the numbers were roughly the same.
The phone back is made of metal. If the aluminum foil is the same width as the phone would the loss be equal? Is any energy 'sneaking' (bending) around the blocking object or are reflections being picked up?
The dynamic range of the measurements were only 17 dB, so I couldn’t measure propagation paths that were weaker than that (compared to the direct unobstructed path). I believe there will be some weaker reflections in the room, but negligible bending around the object. The phone isn’t entirely made of metal, which is why I think some pieces of the signal penetrated it
Thanks for an interesting lab and demonstration. I am thinking about how much the frequency affects the attenuation of the signal in relation to the distance and physical obstacles. This was at 26.5 GHz which is a relatively high frequency. If it were instead a lower frequency, for example 6 GHz, would the attenuation have been significantly less?
In free space, the attenuation with distance is independent of the frequency. The only reason that many pathloss formulas show a dependency is that they assume frequency-dependent receive antenna sizes. I elaborated on this in a blog post: ma-mimo.ellintech.se/2019/10/29/is-the-pathloss-larger-at-mmwave-frequencies/ Physical obstacles will, however, interact differently with radio waves depending on the frequency. I've seen several models (ITU-R P.2346-1, 3GPP TR 38.901) where the penetration losses have the form (f is the frequency): constant*f in dB-scale. The proportionality constant is large for concrete and the human body but small for glass/wood, so it is only some materials where the difference in frequency makes an appreciable difference.
since the frequency is 26 Ghz the optimum distance between antenna elements are 6 mm and you have only four antenna elements so why the transmitter is as big as an iPhone ? another question how big is the bandwidth of the channel you use in this experiment ? i have to admit that you visualize these com;licated things in the best possible way hope you was lecturer in my college
The antenna array is roughly 20 mm wide (≈2 wavelengths at 28 GHz), while the iPhone XR is 75 mm wide. Hence, the iPhone is blocking the path between the transmitter and receiver. If they seem to be equally wide in the video, it is either the perspective or the fact that the array is deployed on a metal sheet that has a width that is more similar to the iPhone. The signal bandwidth wasn't specified in the Developer kit, and the receiver only measure the total signal strength, so I don't know the answer.
The ballon is still spherical since the wave propagate equally fast in all directions. But the signal energy is unequally distributed over the surface area with beamforming
If even our hands are compltely blocked the mmWave signals... how come the IEEE 802.11ad wifi will work? Is it completely block the wifi signals if someone comes between Wifi router and mobile??
The hand is not blocking the signal entirely. The measurement in the video showed that the loss is larger than 18 dB. In later experiments, I have noticed that the loss is around 25 dB. The main workaround is to position the multiple antennas in the router and mobile to minimize the risk that the antennas are blocked. Iphones with mmWave functionality has antenans at the right side and at the top, so that one shouldn't block both simultaneously.
Hey, I really liked this demonstration. It illustrates the importance of strategically positioning your WI-FI access points inside your house or building a cellphone communication network.
Due to multi path dispersion and the open environment there will be fluctuations in the experiment. If the experiment done in anecohic chamber the experimental results will change
@@WirelessFuture Yes, but with two caveats: 1) ensure constant alignment of the transmitting and receiving antennas (constant value of the angles) when moving the antennas; 2) stay away from the receiver's own noise level.
![Constructive and destructive interference [Part 3, Fundamentals of mmWave communication]](http://i.ytimg.com/vi/hxkojUEKAR0/mqdefault.jpg)
![Constructive and destructive interference [Part 3, Fundamentals of mmWave communication]](/img/tr.png)
![Maximum ratio and zero-forcing beamforming [Part 4, Fundamentals of mmWave communication]](http://i.ytimg.com/vi/w46c0QhByPg/mqdefault.jpg)
A brilliant depiction of mmWave penetration losses...
very interesting, I will wait patiently for your incoming series, when you explain the role of IRS.
excellent demonstration from Prof. Emil Björnson on pathloss consequencies on mmW blocking and pathloss or attenuation!!!
Great informative video. Thank you.
Nice marketing for the MIMO book and a very good mmWave demo ❤
Thanks Emil! This is another excellent video with a very good explanation.
A perfect explanation, straight to the point! :)
Informative and insightful, thank you!
Thank you for your very helpful demonstration!
Thank you for the comprehensive explanation. I'm embarking on my research in the field of beamforming, and your insights have proven invaluable. As a newcomer to this field, would you be able to offer me some suggestions or guidance to help me get started?
Thanks for this great illustration, Emil! The great example of body loss. I have seen the statement that body losses at same 26.5GHz can be as high as 30 dB, In the same time additional hand losses stated 0-5dB and 3dB was assumed for most cases of link budget planning maybe it is up to positioning and the fact that there is quite low possibility of completely blocking one of three antennas in smartphone. BTW do you expect any PAAMs mmW to be integrated in smartphones near future to mitigate UL challenges with coverage?
As far as I’ve understood, the smartphones that support mmW already use phased arrays. For example, iPhones 12-14 in the USA have two modules that are place at different sides to limit the risk that both are covered by hands. These phones are said to support 4-MIMO so I would guess these are two phased array that support dual polarization.
@@WirelessFuture I think there is one commercially used phased array in the market QTM527 from Qualcomm and it is too big and power hungry to be used in smartphones. It is used in bigger FWA CPEs where EIRP can be as high as ~45dBm. I haven`t seen any smartphones supporting 4 layers MIMO right now - 2 layers are commonly used: vertical and horizontal polarization, typical smartphone EIRP for mmW is on the level of 13.5dBm (28GHz) (QTM525) and it is 1T2R maybe using two or three antennas in smartphone but only one is active during PRB. This makes UL the limiting factor for coverage. So the development on the smartphone side would practically improve the mmW coverage a lot.
Thank you for this clear experience.
Hi, very interesting. Can you measure something more. Leaves from trees would be nice and what happens when you increase the distance and put something between. Will the loss be the same ?
The propagation loss (with distance) and penetration loss (by an object) are two independent effects, so they always add up when using the decibel scale. I have returned the hardware to TMYTEK, but I will keep your suggestion in mind if I get my hands on it again!
Very nice demonstration, thank you for this video! Regarding the loss seen with the iphone, could it also be partially due to the hand holding the device? It seems that some of the fingers holding up the phone are almost at the same height as the TX and RX antennas but it may just be the camera angle!
It is mostly the camera angle that is tricky because I only touched the edge of the phone. I don't think my hand had a major impact. I did the same experiment while holding the phone from the top, and the numbers were roughly the same.
The phone back is made of metal. If the aluminum foil is the same width as the phone would the loss be equal? Is any energy 'sneaking' (bending) around the blocking object or are reflections being picked up?
The dynamic range of the measurements were only 17 dB, so I couldn’t measure propagation paths that were weaker than that (compared to the direct unobstructed path). I believe there will be some weaker reflections in the room, but negligible bending around the object. The phone isn’t entirely made of metal, which is why I think some pieces of the signal penetrated it
Thanks for an interesting lab and demonstration. I am thinking about how much the frequency affects the attenuation of the signal in relation to the distance and physical obstacles. This was at 26.5 GHz which is a relatively high frequency. If it were instead a lower frequency, for example 6 GHz, would the attenuation have been significantly less?
In free space, the attenuation with distance is independent of the frequency. The only reason that many pathloss formulas show a dependency is that they assume frequency-dependent receive antenna sizes. I elaborated on this in a blog post: ma-mimo.ellintech.se/2019/10/29/is-the-pathloss-larger-at-mmwave-frequencies/
Physical obstacles will, however, interact differently with radio waves depending on the frequency. I've seen several models (ITU-R P.2346-1, 3GPP TR 38.901) where the penetration losses have the form (f is the frequency): constant*f in dB-scale. The proportionality constant is large for concrete and the human body but small for glass/wood, so it is only some materials where the difference in frequency makes an appreciable difference.
since the frequency is 26 Ghz the optimum distance between antenna elements are 6 mm and you have only four antenna elements
so why the transmitter is as big as an iPhone ?
another question how big is the bandwidth of the channel you use in this experiment ?
i have to admit that you visualize these com;licated things in the best possible way hope you was lecturer in my college
The antenna array is roughly 20 mm wide (≈2 wavelengths at 28 GHz), while the iPhone XR is 75 mm wide. Hence, the iPhone is blocking the path between the transmitter and receiver. If they seem to be equally wide in the video, it is either the perspective or the fact that the array is deployed on a metal sheet that has a width that is more similar to the iPhone.
The signal bandwidth wasn't specified in the Developer kit, and the receiver only measure the total signal strength, so I don't know the answer.
1:09 With beamforming, is the balloon an oblong shape, rather than spherical?
The ballon is still spherical since the wave propagate equally fast in all directions. But the signal energy is unequally distributed over the surface area with beamforming
If even our hands are compltely blocked the mmWave signals... how come the IEEE 802.11ad wifi will work? Is it completely block the wifi signals if someone comes between Wifi router and mobile??
The hand is not blocking the signal entirely. The measurement in the video showed that the loss is larger than 18 dB. In later experiments, I have noticed that the loss is around 25 dB. The main workaround is to position the multiple antennas in the router and mobile to minimize the risk that the antennas are blocked. Iphones with mmWave functionality has antenans at the right side and at the top, so that one shouldn't block both simultaneously.
Hey, I really liked this demonstration. It illustrates the importance of strategically positioning your WI-FI access points inside your house or building a cellphone communication network.
Thank you sir
Thank You sir...
Due to multi path dispersion and the open environment there will be fluctuations in the experiment. If the experiment done in anecohic chamber the experimental results will change
Yes, in an anechoic chamber with careful measurements, one should get a perfect fit to the Friis propagation formula.
@@WirelessFuture Yes, but with two caveats:
1) ensure constant alignment of the transmitting and receiving antennas (constant value of the angles) when moving the antennas;
2) stay away from the receiver's own noise level.