Hands down, the best explanation of Beamforming concept , so far across the internet. Hope the LLMs train on content like these to cater for future generations. Thanks a lot Iain ! Please dont stop making more videos.
Here I am, in my first hardware engineering job for digital audio devices, coming back to the same channel that got me through my first signals and systems class as a sophomore in college. Quality translates, folks! Iain rocks!!!
Solid Gold. Like Einstein said, if you can't explain it simply, you don't understand it enough. What a concise visual representation of the basics of beamforming.
Thanks. It was awesome as I watched it for the second time. In practice, how are these delays adjusted and how do we know (in the receiver) which direction is the direction of interest to adjust those delays? Same question for transmit beamforming.
I had so many "aha!" moments during this video that I ended up liking it multiple times, from going to click the "Like" button, forgetting that I had already liked the video just a few minutes before. Great explanation!
It's like noise cancellation headphones. But instead of subtracting by having a reversed waveform, it adds. And just like good noise cancelling headphones, it has to know the distance between the microphones and earpieces.
I'm so glad it helped. And I'm glad you found it useful to watch a second time. I've always tried to put details into my videos, alongside the overall explanations, that will give people extra benefits if they watch the videos multiple times.
So when the signal is cancelled out on the incoming transmission.. what happens to it...? Just got a bit confused at this point? How is a beam formed if it's cancelled out? Or... Have I misinterpreted what you have said?
Yes, you're right, "cancellation" does take a bit of thinking about. One example in real life is at surf beaches, when a wave is coming in to the shore, and another previous wave is going back down the sand causing a ripple in the reverse direction. When the two waves meet, the peaks of the incoming wave add to the peaks of the returning wave creating a splash. But also, the peaks of the incoming wave are "cancelled" by the troughs of the returning wave, causing the water to be smooth a those points.
Thanks for explaination, now I'm wondering how the transceiver actually tunes to specific wireless client by selecting right delays?. Is it still in PHY or MAC layer ? Also how fast is the switching ? As if we have 4x4 router and two typical 2x2 clients how router assures signal from 3rd device will be still handled without comprising value added of two existing MU-MIMO clients?
Thank you Iain. Beam forming is an Amazing principle. So by adding multiple antennas (n Antennas) and injecting a delay circuits and adders we can form those shaped zones and select which receiver to have the best reception from specific direction. I imagine that if we have a programmable delay circuits with algorithm to control those delays and selecting the proper antennas then we can control those beams. Is that what is used in space division multiple access? We maybe able also to have intelligent algorithm that senses the RSSI levels, provide feedback to apply combinations of selecting antennas, adders and delay circuits and eventually automatically select the sending location. I wonder if such a thing already exists?
Yes, that's right. It's exactly what happens in MU-MIMO. The process you've described happens in the matrix operations in the receiver (for receive beamforming) and the transmitter (for transmit precoding). See "What is Multi-User MIMO Communications (MU MIMO)?" th-cam.com/video/0ncIWlhsu1A/w-d-xo.html
Before watching your video I’ve been reading a lot of blog introducing what is beamforming, but I still cannot understand it. Your illustration is so easy to understand. Thank you Sir!
This Beamforming, given fixed delay, is good only for fixed (position) signal source and fixed receiver - not for moving source or moving receiver, otherwise it will be hard to tune the delay in a way to match the movement of the source, or receiver. Has this technology "Beamforming" been used for moving source, or receiver, at all in practical application?
Thank you very much for the explanation. I have a question. Antennas are fixed on a wireless device. So, the distance between antennas will be the same. How does it work for different frequencies?
Excellent question! The form of beamforming that maximises SNR (that I describe in this video) is inherently narrow band. In practice it doesn't have to be exactly half a wavelength separation, and there are approaches to designing beams that are wide band, but they come with a performance penalty in terms of SNR.
Dear Iain, these are all amazing videos. It is really tough to condense such complex material into short videos but maintain understandability of the it. This is really appreciated. You mentioned radar a couple of times. Do you have any plans to make a basic lecture on that? Maybe just simple doppler/ToF/AoA extraction or maybe something on MUSIC algorithm. I have been searching for a while for a decent explanation of MUSIC but it is hard to find one.
... I just came across this comment again. I've now made a few videos on Radar, and plan to make more. I guess you may have seen them by now, but if not, then the ones I've done so far are: "Why is a Chirp Signal used in Radar?" th-cam.com/video/Jyno-Ba_lKs/w-d-xo.html , "What is a Stepped Frequency Radar Signal?" th-cam.com/video/6JVGb3KpVqs/w-d-xo.html and "How does a Radar Track Manoeuvring Targets?" th-cam.com/video/ibvlKTGQ4zQ/w-d-xo.html
Hi, I am a student with master's degree studying communication in south korea. Thank you very much for the good lecture. You are the hope of the wireless communication. I have two questions. 1. If i want to implement a kind of antenna diversity through the two antennas on the receiving side, It is understood that when the awgn noise added to the signal from the two antennas is different, a gain of 3 dB is obtained. As you explained in the video, when i understand physically, don't the signals entered by the two antennas have the same awgn noise? That's the question I had in mind. Since there is a benefit from beamforming, I think it would be right to have different awgn noise, but I don't understand this clearly. 2. According to what i saw in the MIMO channel-related textbook, if the angle spread (dispersion of PAS) is narrow, that is, if the reception beam is narrow, the correlation with adjacent antennas increases. It states that channel capacity and divercity gains are reduced. Performing beamforming on the transmitting side will cause the beam to be browned, which is expected to result in relatively large correlation between the receiving side adjacent antennas. Then, is it okay to understand that it is a technology that has advantages in terms of energy concentration during tx beamforming, but also has disadvantages because of the high correlation? Thank you.
It all depends on the characteristics of the channel. And I'm not sure you're understanding about the noise. Noise is introduced primarily by amplifiers in the RF "front-ends" of the transmitter and receiver. If each antenna is connected to its own amplifier, then the noise will be different for each antenna's signal. This video might help: "What are Spatial Diversity and Spatial Multiplexing in MIMO?" th-cam.com/video/MNA0xn7EeyY/w-d-xo.html and "
This is assuming ideal isotropically radiating point source elements (not flat patches). In practice, the spacing that maximises radiation efficiency will depend on the shape and size of the patches, as well as the locations of surrounding back plains etc.
Thank you Iain, For making such a wonderful videos and providing us a valuable knowledge. Can you make a videos on DSP Filters (IIR and FIR Filters)? It will be very helpful.
I know beam forming from acoustic cameras that are using an microphone array to create a picture where you can see the source of noise. This allows to analyze what part of a machine, car, plane and such is emitting what noise. Might be a topic for another video.
I can understand the basic princple of beamforming form your video, but I have a question. Consider 24Ghz freq MIMO radar system, which has the spatial distace between Rx antenna is lambda/2. To acheive +50 degree digital beamforming, the delay time should be (lambda/2)*sin(50degree)/c = 1.6e-11 sec. So, if I want to make 50 degree beamforming, the minimum sampling rate of MIMO system should be 1.6e11 Hz, which is too high to make a real system. I want to know how can I solve this problem.. Thanks.
It is symmetric in the horizontal axis, but normally the array is implemented with a back plane that reflects or absorbs the "backwards pointing" energy (downwards in my picture), so that the array only radiates "outwards" (upwards in my picture). For details on repetitions in the "outwards" beam pattern (upwards in my picture), see: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html
Firstly I would like to start off by saying thank you for sharing your knowledge with your videos which really are helpful to students like me. I watched your beamforming videos (in MIMO), I have a couple of doubts sir. Generally received signal from base station (having single antenna) to communication user(having single antenna) is written as: r = h*(root (P)) * x + n , where h is channel, root P is included to say that transmitted s/g power from base station is P (given E[|x|^2]=1) , x is a transmitted signal, and n is noise. Now, can the above equation be modified for a base station with M antennas as: r = h*W*(root (P)) * x + n , where W is beamforming weights. Is this correct? If so, what does this physically mean? Does that mean total power available at the base station is being divided among all antennas or in general how this transmitted power P is being related to weights in the beamforming vector. I'm confused sir, it would be really helpful if you clarify this.
It depends on the norm of W. Generally we would use a W with norm = 1. Then, yes, it represents the total transmit power being allocated across the antennas.
What is never mentioned in antenna radiation is whether or not the E/M loops being radiated do break down into loops of half wavelength as the loop try to grow in size with distance, In waveguides and cavity resonators it is common to show the stacking of " half wavelength E/M blocks to fit in the available space. A one megacycle wave, has a much larger loop than a 5 GigaHertz signal and though everyone talks about lobes, and directivity, and efficiency, and so on, no one talks about the "stacking of E/M blocks" the size of which depends on the frequency of transmission. So for a given frequency , how many stacked E/M loops exist in it as the wave moves out? I have thought about this for the last 80 years and it seems to me that when one has a directional antenna, or a phased array, all that is happening the system is cutting off and eliminating the peripheral blocks in the stack of E/M blocks in the lobe or in any omnidirectional antenna, I simulated this with a computer algorithm which, when the circumference ( wavefront) of the wave grew larger than a wavelength integers, then the " circumference would accommodate another loop half a wavelength long". The simulation works beautifully and the patterns that emerge make it so obvious, after I saw it, From a central location of the antenna, after the wave settles down, the patterns seems to change from a "polar diagram" to a cartesian diagram where the four symmetrical squares of cartesian coordinates, simultaneously move out their four quarters containing the same pattern of E/M loops as exist in rectangular waveguides contain the higher modes. When I plotted the B and the E field loops far from the antenna they came out to be exactly as occurs in rectangular waveguides with the B loops as normal, and the E loops are exactly the same as the E in the waveguide and the surface currents in a waveguide, It is exactly the same pattern, So it seems, that a centrally placed antenna as a source, will have the near E/M fields going through "a pushy transient pattern" then "a middle field pattern" which I call the settling down zone, and then the far field would resemble the pattern obtained in a rectangular waveguide excited with an electric probe or a magnetic probe as one desires. It is fascination to see the four quarters of cartesian coordinates moving out with additional loops being added as the distance increases, It is remarkable, This stacking of E/M blocks, the size of half a wavelength, in a radiating pattern, is interesting, and in a phased array or a directional antenna, it seems that all one does is to " phase out the peripheral E/M blocks and donate the power to the other half wavelength E/M blocks remaining in the " stacked lobe" There seems to be more going on in radiation than one thinks, and this "stacking effect" needs further discussion . An analogy may be used by looking at Chladni's figures in vibrating sheets or a Jelly block, and after all our radiating medium is not much different from a jelly or a rubber block with its "own impedance" in how it permits our signals to "accelerate " build and decay and reverse those the E/M loops the size of half a wavelength. th-cam.com/video/wvJAgrUBF4w/w-d-xo.html
I'm not sure what you're asking, sorry. Beamforming is a technique for increasing the radiated/received power in a particular direction. In its most basic version, it is tuned for a specific channel/user. In more advanced implementations, it can be tuned for multiple channels/users.
I doubt that the delay between the second and third antenna equal to Delta. coz , for the first one the sinewave traveling quarter of its period while for the second, half. is it right?
Well the important word is "between". It travels Delta delay _between_ any two neighbouring antennas. Of course, that means that _between_ the first and the third antennas it would be 2 x Delta, and so on.
Well, the speed of light is ~ 3x10^8 m/s. So if our antennas are for example 30 cm apart from each other, then if my calculations are correct the time delay would be around 1 ns (nano second). Can it make such a difference? Are our electronic devices so sensitive and accurate to detect even this? For comparison let me add that for example for LTE the symbol duration is around 71 us (micro second) which made me wondering whether a few nano seconds can make a difference here.
@@iain_explains Good catch:) So you are saying that our mobile phones are so advanced to detect this? That is very interesting. I am no expert in electronics, just curious how it works.
So… I created a machine monitoring daq using LabVIEW, mics and NI cRIO’s… about 4 months into monitoring a dyno i intuitively came up with this concept not knowing I stumbled upon beam forming by looking at hundreds of high speed wave form data… Thank you sir for plainly consolidating my mad-ness!
I never liked a TH-cam video as much as i like this one. Indeed, it's the best explanation I have ever heard. Thank you so much for your work Prof. I wish I could have a professor like you in my university.
@@iain_explains The UK used to have British Satellite Broadcast system in the early 90s which used a phased array on a flat square plate ("squarial") which functioned the same as a dish.
Fantastic explanation. I want to use 2 inverted L antennas to create a phased array. My existing vertical L is a 20m (14Mhz) with a little loading coil to resonate on 40m (7Mhz). How do I reconcile the distance between the antennas to the delay "harness" design required....Is 1/4 wavelength optimal.....could I compromise somehow to get it to work on both bands? I guess I would need a different delay for each of the bands?? Or is it a daft idea! :) 73 de VK2AOE
You'll definitely need to match the delay to the centre frequency of the band .. seperate delays for seperate bands. There are wide-band beamforming techniques, but it's always a trade-off. Beamforming works best when it's narrowband.
@@iain_explains Thanks....I loved your explanation it made the concept clear. As with all Engineering the devil resides in the detail. Thanks for the answer.
Sorry, I'm not sure I understand what you are asking exactly. I didn't represent anything in exponential form in the video. But in general though, a time delay of a sinusoidal signal can equivalently be thought of as a phase shift of the sinusoid - and phase shifts can be written in exponential form as e^{i theta}.
@@iain_explains Thanks so much, I exactly asked this, because I was reading about the steering vectors, they were always represented in the exponential form.
Thank you sir, for all of your videos; those are amazing😍. I have one doubt if we consider BS and UE then beamforming is used at BS only or at the UE side or both?
It can be used by any communication device that has multiple antennas. So yes, it can be used at both the Base Station and the User Equipment, if they have multiple antennas.
Very good point. In these basic channel models the mutual impedance is ignored. When the antenna elements are half a wavelength apart (or more), then this is not too bad an approximation. In reality the presence of nearby scatterers, such as the feed lines of the antenna elements and the mounting platform, also affect the impedance characteristics. In practice though, the channel matrix is estimated from training data, and so all of these factors are accounted for by the beam forming algorithm. They might mean that the performance is less than the ideal channel model predicts, but it doesn't mean it won't work (unless theenna ant elements are much closer than half a wavelength).
@@iain_explains Thank you Sir for the explanation, it has been 16 years my question to my lecturer, but the answer did not satisfy me. In fact, applied antenna for 5G is released from antenna vendor with many matrix combinations, I believe the mutual impedance is not a big things
It's a little bit confusing for me as I thought most of the beamforming is done on the transmitter end while your explanation is mostly from receiver point of view. Would you please elaborate on this?
You might like to watch this video: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html and you'll be able to find the Matlab code I wrote for it on my webpage: iaincollings.com under the "Multiple Antenna Communications" heading on the "Digital Communications" page.
Thanks this made me understand it better! It still left me with a question though. If you look at an angle such that the phase shifting is exactly a whole period, the waves will still add up constructively there. How is it made that these side maxima are so weak in phased array antennas?
Yes, that's a good observation. This effect depends on how far apart the antennas are spaced. Check out this video for more details: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html
If im an Client Router Beamforming as Repeater, will it work beamforming from Ap that dont have mimo ? My Ap Is moving Object here. Please help me answer ?
thank you so much for your video. I have a question? for a mathematic student, which books or courses do you propose to learn these concepts deeply? best.
Yes, absolutely. Applications include sound source localisation, and 3D surround sound headphones. Also submarines tow arrays of microphones behind them, to do target localisation.
The angle is a simple function of the delay and the antenna spacing (it is briefly explained at the 7 min 30 sec point of the video). sin(theta) = Delta * c / d where theta is measured from the line perpendicular to the line that the array elements are on.
@@iain_explains For pulse radar, this is the case. But for the FMCW radar, it is using the FFT to get the range, as the frequency is proportional to the range in the FMCW waveform.
It can be used to find the direction of a source, by scanning through different phase shifts and looking for which one gives an energy peak/spike. But the distance can only be found by sending a radar pulse, and listening for the echo, or by having multiple arrays spaced widely apart, and triangulating the directions.
I'm glad you liked the video. I'm not offering in-house training or tutoring at the moment, but potentially will in future. For now, the videos will have to do. Please do let me know if there are particular topics that you'd like to hear more about if I haven't covered it so far. Check out my web page which shows a fully categorised list of the videos: iaincollings.com
Best. Explanation. Period.
Period.
Period.Period.
Hands down, the best explanation of Beamforming concept , so far across the internet. Hope the LLMs train on content like these to cater for future generations. Thanks a lot Iain ! Please dont stop making more videos.
Glad you liked it!
Here I am, in my first hardware engineering job for digital audio devices, coming back to the same channel that got me through my first signals and systems class as a sophomore in college. Quality translates, folks! Iain rocks!!!
That's awesome to hear! Thanks for your comment. Good luck in your job. It's always exciting to hear from people starting their careers in industry.
Words are not enough to express how grateful I am to have encountered this channel. Thank you again dear Iain for sharin your knowledge
Thanks for your very nice comment. It's great to hear that you like the videos.
Solid Gold.
Like Einstein said, if you can't explain it simply, you don't understand it enough.
What a concise visual representation of the basics of beamforming.
Thanks for your nice comment. I'm glad you like the video.
Thanks. It was awesome as I watched it for the second time. In practice, how are these delays adjusted and how do we know (in the receiver) which direction is the direction of interest to adjust those delays? Same question for transmit beamforming.
I've got a video coming out on Monday that gives the equation that relates the direction to the delays. Keep an eye out for it.
@iain_explains Thank you so much. Definitely will be so helpful as always
That's exactly "the best explanation I've ever heard". Thank you sir!
Glad it was helpful!
I have a master degree in EE, this is the best explanation of beam forming I have seen without all the fancy equations.
I'm glad you liked the explanation.
I had so many "aha!" moments during this video that I ended up liking it multiple times, from going to click the "Like" button, forgetting that I had already liked the video just a few minutes before. Great explanation!
Thanks. I'm so glad you found the video helpful.
It's like noise cancellation headphones. But instead of subtracting by having a reversed waveform, it adds.
And just like good noise cancelling headphones, it has to know the distance between the microphones and earpieces.
Watched 2 times and understood it like nothing else. It is the best explanation.
I'm so glad it helped. And I'm glad you found it useful to watch a second time. I've always tried to put details into my videos, alongside the overall explanations, that will give people extra benefits if they watch the videos multiple times.
This explain is intuitive and clear. Excellent job.
Glad it was helpful!
One of the best explanations of beam forming ..
Glad you liked it
Excellent. The idea of starting with the recieve case is genius!
Thanks. Glad you liked it!
The most intuisive explaination that I've ever seen!
I'm so glad it was helpful.
Clear explanation sir. thank you
So when the signal is cancelled out on the incoming transmission.. what happens to it...? Just got a bit confused at this point? How is a beam formed if it's cancelled out? Or... Have I misinterpreted what you have said?
Yes, you're right, "cancellation" does take a bit of thinking about. One example in real life is at surf beaches, when a wave is coming in to the shore, and another previous wave is going back down the sand causing a ripple in the reverse direction. When the two waves meet, the peaks of the incoming wave add to the peaks of the returning wave creating a splash. But also, the peaks of the incoming wave are "cancelled" by the troughs of the returning wave, causing the water to be smooth a those points.
Thanks for explaination, now I'm wondering how the transceiver actually tunes to specific wireless client by selecting right delays?. Is it still in PHY or MAC layer ? Also how fast is the switching ? As if we have 4x4 router and two typical 2x2 clients how router assures signal from 3rd device will be still handled without comprising value added of two existing MU-MIMO clients?
I haven't seen a better explanation for this.. hats off sir
Thanks. I'm glad you liked it.
6:40 Shouldn't that be 3 time the delay (and not twice) if we want the signal from non-equatorial source?
Superb explaination with great example. Now I also understood how Beamforming Microphones work in my headphones. Thank you!!
Glad it was helpful!
I watched many videos on this topic, but only this video made me visualise the beam forming. Thanks for making such a complicated topic so easy :)
Glad it helped!
One of the best explanations. Thanks.
Glad you think so!
Thank you Iain. Beam forming is an Amazing principle. So by adding multiple antennas (n Antennas) and injecting a delay circuits and adders we can form those shaped zones and select which receiver to have the best reception from specific direction. I imagine that if we have a programmable delay circuits with algorithm to control those delays and selecting the proper antennas then we can control those beams. Is that what is used in space division multiple access? We maybe able also to have intelligent algorithm that senses the RSSI levels, provide feedback to apply combinations of selecting antennas, adders and delay circuits and eventually automatically select the sending location. I wonder if such a thing already exists?
Yes, that's right. It's exactly what happens in MU-MIMO. The process you've described happens in the matrix operations in the receiver (for receive beamforming) and the transmitter (for transmit precoding). See "What is Multi-User MIMO Communications (MU MIMO)?" th-cam.com/video/0ncIWlhsu1A/w-d-xo.html
Interesting facts
Before watching your video I’ve been reading a lot of blog introducing what is beamforming, but I still cannot understand it. Your illustration is so easy to understand. Thank you Sir!
I'm glad you found it useful.
Best beamforming explanation I have seen.
Thanks for your comment. I'm glad you liked it.
You even make it clear to me. Thanks a lot for ur work.
Glad to hear that
can't be explained better than this, so clear.
Thanks for your comment. I'm glad you liked it.
This Beamforming, given fixed delay, is good only for fixed (position) signal source and fixed receiver - not for moving source or moving receiver, otherwise it will be hard to tune the delay in a way to match the movement of the source, or receiver. Has this technology "Beamforming" been used for moving source, or receiver, at all in practical application?
Yes, that's right. When a source/target moves you need to adapt the delays. Then it gets called "Adaptive Beamforming".
Thank you very much for the explanation.
I have a question. Antennas are fixed on a wireless device. So, the distance between antennas will be the same. How does it work for different frequencies?
Excellent question! The form of beamforming that maximises SNR (that I describe in this video) is inherently narrow band. In practice it doesn't have to be exactly half a wavelength separation, and there are approaches to designing beams that are wide band, but they come with a performance penalty in terms of SNR.
It's amazing how good your videos are. With just pen and paper you have made me understand so many concepts! Thank you so much for your contribution.
I'm so glad you like the videos and the format!
Trying to wrap my head around this topic all day, caught your video and boom, mind blown... THANK YOU!
Fantastic! I'm so glad it helped.
wow , these small lectures are insanely good, thank you so much❤❤❤
I'm so glad you're finding them helpful!
phenomenal explanation
Thanks. I'm glad you liked it.
Dear Iain, these are all amazing videos. It is really tough to condense such complex material into short videos but maintain understandability of the it. This is really appreciated.
You mentioned radar a couple of times. Do you have any plans to make a basic lecture on that? Maybe just simple doppler/ToF/AoA extraction or maybe something on MUSIC algorithm. I have been searching for a while for a decent explanation of MUSIC but it is hard to find one.
I'm glad you like the videos. Thanks for the suggestion of a radar and MUSIC topic video. I'll add it to my "to do" list.
... I just came across this comment again. I've now made a few videos on Radar, and plan to make more. I guess you may have seen them by now, but if not, then the ones I've done so far are: "Why is a Chirp Signal used in Radar?" th-cam.com/video/Jyno-Ba_lKs/w-d-xo.html , "What is a Stepped Frequency Radar Signal?" th-cam.com/video/6JVGb3KpVqs/w-d-xo.html and "How does a Radar Track Manoeuvring Targets?" th-cam.com/video/ibvlKTGQ4zQ/w-d-xo.html
Great explanations. Is it possible to visualize near field beamforming (beam focusing) in similar way?
Yes, but in that case the delays are not constant between different antenna elements.
I came here for fun, who else
Hi, I am a student with master's degree studying communication in south korea.
Thank you very much for the good lecture.
You are the hope of the wireless communication.
I have two questions.
1. If i want to implement a kind of antenna diversity through the two antennas on the receiving side,
It is understood that when the awgn noise added to the signal from the two antennas is different, a gain of 3 dB is obtained.
As you explained in the video, when i understand physically, don't the signals entered by the two antennas have the same awgn noise? That's the question I had in mind.
Since there is a benefit from beamforming, I think it would be right to have different awgn noise, but I don't understand this clearly.
2. According to what i saw in the MIMO channel-related textbook, if the angle spread (dispersion of PAS) is narrow, that is, if the reception beam is narrow, the correlation with adjacent antennas increases.
It states that channel capacity and divercity gains are reduced.
Performing beamforming on the transmitting side will cause the beam to be browned, which is expected to result in relatively large correlation between the receiving side adjacent antennas.
Then, is it okay to understand that it is a technology that has advantages in terms of energy concentration during tx beamforming, but also has disadvantages because of the high correlation?
Thank you.
I am the author of this comment.
dispersion of PAS (X)
variance of PAS (O)
It all depends on the characteristics of the channel. And I'm not sure you're understanding about the noise. Noise is introduced primarily by amplifiers in the RF "front-ends" of the transmitter and receiver. If each antenna is connected to its own amplifier, then the noise will be different for each antenna's signal. This video might help: "What are Spatial Diversity and Spatial Multiplexing in MIMO?" th-cam.com/video/MNA0xn7EeyY/w-d-xo.html and "
Curiosity can put u anywhere and some places cant be forgotten .This place too💙
Thank you Pr. for this helpful video. Could you explain Beamspace in MIMO.
Best regards
Thanks for the suggestion. I'll put it on my "to do" list.
Best explanation in existence.
Thanks. I'm really glad you liked it.
The distance between the center of two patches should be d or the distance between two adjacent sides of each patch?
This is assuming ideal isotropically radiating point source elements (not flat patches). In practice, the spacing that maximises radiation efficiency will depend on the shape and size of the patches, as well as the locations of surrounding back plains etc.
I like the simplicity!
Thx from korea! BEST explanation!
Glad you think so!
Thank you Iain, For making such a wonderful videos and providing us a valuable knowledge. Can you make a videos on DSP Filters (IIR and FIR Filters)? It will be very helpful.
Thanks for the suggestion. They're on my "to do" list.
Please upload more video on analog and digital communication, radios,SatCom,latest technologies
Thanks for the topic suggestions. I've got them on my "to do" list (but it's getting to be a long list ...)
I know beam forming from acoustic cameras that are using an microphone array to create a picture where you can see the source of noise. This allows to analyze what part of a machine, car, plane and such is emitting what noise. Might be a topic for another video.
Thanks for the suggestion. I'll add it to my "to do" list.
Fantastic lecture. Thank you.
I'm glad you liked it.
I can understand the basic princple of beamforming form your video, but I have a question.
Consider 24Ghz freq MIMO radar system, which has the spatial distace between Rx antenna is lambda/2.
To acheive +50 degree digital beamforming, the delay time should be (lambda/2)*sin(50degree)/c = 1.6e-11 sec.
So, if I want to make 50 degree beamforming, the minimum sampling rate of MIMO system should be 1.6e11 Hz, which is too high to make a real system.
I want to know how can I solve this problem.. Thanks.
frankly it is the best explanation
Thanks. I'm glad you liked it.
awesome explanation, I thought at the begneinggg how hw said best explanation , but he worth the name love the material Mr.Iain
Glad you liked it!
Dear Sir your videos are brilliant. Your explanations very clear .Thank you for sharing your knowledge
Thanks for your nice comment. I'm glad the videos are helping.
Hi ! Why the beampatern is not symetric to the vertical axis ? Thanks !
It is symmetric in the horizontal axis, but normally the array is implemented with a back plane that reflects or absorbs the "backwards pointing" energy (downwards in my picture), so that the array only radiates "outwards" (upwards in my picture). For details on repetitions in the "outwards" beam pattern (upwards in my picture), see: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html
@@iain_explains ok thanks for your reply :)
I just saw your other video! Will watch that now!
Great. I hope it helps.
Beam forming technique is only used into half duplex system?
Glad that I’ve found this channel and yeah best explanation I’ve ever heard
Welcome aboard! I'm glad you liked the explanation.
I am trying to understand the MUSIC algorithm. Any link to easily understand it?
Firstly I would like to start off by saying thank you for sharing your knowledge with your videos which really are helpful to students like me.
I watched your beamforming videos (in MIMO), I have a couple of doubts sir.
Generally received signal from base station (having single antenna) to communication user(having single antenna) is written as: r = h*(root (P)) * x + n , where h is channel, root P is included to say that transmitted s/g power from base station is P (given E[|x|^2]=1) , x is a transmitted signal, and n is noise.
Now, can the above equation be modified for a base station with M antennas as: r = h*W*(root (P)) * x + n , where W is beamforming weights. Is this correct? If so, what does this physically mean? Does that mean total power available at the base station is being divided among all antennas or in general how this transmitted power P is being related to weights in the beamforming vector.
I'm confused sir, it would be really helpful if you clarify this.
It depends on the norm of W. Generally we would use a W with norm = 1. Then, yes, it represents the total transmit power being allocated across the antennas.
@@iain_explains Thanks alot for ur reply sir.
What is never mentioned in antenna radiation is whether or not the E/M loops being radiated do break down into loops of half wavelength as the loop try to grow in size with distance, In waveguides and cavity resonators it is common to show the stacking of " half wavelength E/M blocks to fit in the available space. A one megacycle wave, has a much larger loop than a 5 GigaHertz signal and though everyone talks about lobes, and directivity, and efficiency, and so on, no one talks about the "stacking of E/M blocks" the size of which depends on the frequency of transmission. So for a given frequency , how many stacked E/M loops exist in it as the wave moves out? I have thought about this for the last 80 years and it seems to me that when one has a directional antenna, or a phased array, all that is happening the system is cutting off and eliminating the peripheral blocks in the stack of E/M blocks in the lobe or in any omnidirectional antenna,
I simulated this with a computer algorithm which, when the circumference ( wavefront) of the wave grew larger than a wavelength integers, then the " circumference would accommodate another loop half a wavelength long".
The simulation works beautifully and the patterns that emerge make it so obvious, after I saw it, From a central location of the antenna, after the wave settles down, the patterns seems to change from a "polar diagram" to a cartesian diagram where the four symmetrical squares of cartesian coordinates, simultaneously move out their four quarters containing the same pattern of E/M loops as exist in rectangular waveguides contain the higher modes. When I plotted the B and the E field loops far from the antenna they came out to be exactly as occurs in rectangular waveguides with the B loops as normal, and the E loops are exactly the same as the E in the waveguide and the surface currents in a waveguide, It is exactly the same pattern,
So it seems, that a centrally placed antenna as a source, will have the near E/M fields going through "a pushy transient pattern" then "a middle field pattern" which I call the settling down zone, and then the far field would resemble the pattern obtained in a rectangular waveguide excited with an electric probe or a magnetic probe as one desires. It is fascination to see the four quarters of cartesian coordinates moving out with additional loops being added as the distance increases, It is remarkable,
This stacking of E/M blocks, the size of half a wavelength, in a radiating pattern, is interesting, and in a phased array or a directional antenna, it seems that all one does is to " phase out the peripheral E/M blocks and donate the power to the other half wavelength E/M blocks remaining in the " stacked lobe" There seems to be more going on in radiation than one thinks, and this "stacking effect" needs further discussion . An analogy may be used by looking at Chladni's figures in vibrating sheets or a Jelly block, and after all our radiating medium is not much different from a jelly or a rubber block with its "own impedance" in how it permits our signals to "accelerate " build and decay and reverse those the E/M loops the size of half a wavelength. th-cam.com/video/wvJAgrUBF4w/w-d-xo.html
Did you say you’ve been thinking about this for 80 years?! That’s impressive. What antenna systems were you working on 80 years ago? And since then?
Thank you for the introduction
You're welcome
Thanks Mr. Iain for explaining in such an easy way the Beamforming.
Glad it was helpful!
amazing explanation very clear! but if im using only one channel/device does it better to turn off beamforming for smoother connection? thank you!
I'm not sure what you're asking, sorry. Beamforming is a technique for increasing the radiated/received power in a particular direction. In its most basic version, it is tuned for a specific channel/user. In more advanced implementations, it can be tuned for multiple channels/users.
I doubt that the delay between the second and third antenna equal to Delta. coz , for the first one the sinewave traveling quarter of its period while for the second, half. is it right?
Well the important word is "between". It travels Delta delay _between_ any two neighbouring antennas. Of course, that means that _between_ the first and the third antennas it would be 2 x Delta, and so on.
Well, the speed of light is ~ 3x10^8 m/s. So if our antennas are for example 30 cm apart from each other, then if my calculations are correct the time delay would be around 1 ns (nano second).
Can it make such a difference? Are our electronic devices so sensitive and accurate to detect even this?
For comparison let me add that for example for LTE the symbol duration is around 71 us (micro second) which made me wondering whether a few nano seconds can make a difference here.
Well, it's no different to how your brain uses your eyes to work out how far away something is. And your eyes are even less than 30 cm apart!
@@iain_explains Good catch:) So you are saying that our mobile phones are so advanced to detect this?
That is very interesting. I am no expert in electronics, just curious how it works.
This is really good one 👍
I'm glad you like it
Hello,
What about some source(not interested) broadcasting signals in the angle other than straight?
Sorry, I don't know what you're trying to ask.
How far is far enough for considering the wave front to be flat?
So… I created a machine monitoring daq using LabVIEW, mics and NI cRIO’s… about 4 months into monitoring a dyno i intuitively came up with this concept not knowing I stumbled upon beam forming by looking at hundreds of high speed wave form data…
Thank you sir for plainly consolidating my mad-ness!
I'm so glad my video helped you.
I never liked a TH-cam video as much as i like this one. Indeed, it's the best explanation I have ever heard. Thank you so much for your work Prof. I wish I could have a professor like you in my university.
Thanks so much for your very nice comment. I'm so glad you liked the video.
Very useful. Very well explained.
Glad it was helpful!
Hi there Iain, is this what is called a phased array?
Yes, that's right.
hello, I'm from Brazil and I wanted to know if beamforming also improves latency/ping in online games, thanks
Great explanation, Iain
Glad you liked it
Great explanation. I expect that by changing the delay one could cause the beam to sweep.
Yes, that's what's called Adaptive Beamforming.
Could you use this as an alternative to a dish antenna?
Yes, that's right.
@@iain_explains The UK used to have British Satellite Broadcast system in the early 90s which used a phased array on a flat square plate ("squarial") which functioned the same as a dish.
Thanks for perfect explanation sir
Glad you liked it.
It's just awesome! Didn't know there is a similar beamforming effect in the receiving end as there is in the transmitting side.
Glad I could help!
Fantastic explanation. I want to use 2 inverted L antennas to create a phased array. My existing vertical L is a 20m (14Mhz) with a little loading coil to resonate on 40m (7Mhz). How do I reconcile the distance between the antennas to the delay "harness" design required....Is 1/4 wavelength optimal.....could I compromise somehow to get it to work on both bands? I guess I would need a different delay for each of the bands?? Or is it a daft idea! :) 73 de VK2AOE
You'll definitely need to match the delay to the centre frequency of the band .. seperate delays for seperate bands. There are wide-band beamforming techniques, but it's always a trade-off. Beamforming works best when it's narrowband.
@@iain_explains Thanks....I loved your explanation it made the concept clear. As with all Engineering the devil resides in the detail. Thanks for the answer.
I'm glad I've been able to help.
Can you please explain how did you represent it in an exponential form ?
Sorry, I'm not sure I understand what you are asking exactly. I didn't represent anything in exponential form in the video. But in general though, a time delay of a sinusoidal signal can equivalently be thought of as a phase shift of the sinusoid - and phase shifts can be written in exponential form as e^{i theta}.
@@iain_explains Thanks so much, I exactly asked this, because I was reading about the steering vectors, they were always represented in the exponential form.
The best explanation I've seen 👏
Thanks. I'm glad you liked it.
Thank you sir, for all of your videos; those are amazing😍. I have one doubt if we consider BS and UE then beamforming is used at BS only or at the UE side or both?
It can be used by any communication device that has multiple antennas. So yes, it can be used at both the Base Station and the User Equipment, if they have multiple antennas.
@@iain_explains Thank you 😊
How about calculation mutual impedance among antennas? Is it ignored?
Very good point. In these basic channel models the mutual impedance is ignored. When the antenna elements are half a wavelength apart (or more), then this is not too bad an approximation. In reality the presence of nearby scatterers, such as the feed lines of the antenna elements and the mounting platform, also affect the impedance characteristics. In practice though, the channel matrix is estimated from training data, and so all of these factors are accounted for by the beam forming algorithm. They might mean that the performance is less than the ideal channel model predicts, but it doesn't mean it won't work (unless theenna ant elements are much closer than half a wavelength).
@@iain_explains Thank you Sir for the explanation, it has been 16 years my question to my lecturer, but the answer did not satisfy me. In fact, applied antenna for 5G is released from antenna vendor with many matrix combinations, I believe the mutual impedance is not a big things
Good explanation… 😊 thank you
Glad it was helpful!
Best explanation ! Thank you !
Glad it was helpful!
Thanks , well explained.
Glad it was helpful!
It's a little bit confusing for me as I thought most of the beamforming is done on the transmitter end while your explanation is mostly from receiver point of view. Would you please elaborate on this?
Beamforming can (and is) done at either (or both) the transmitter and/or the receiver.
Very clear explanation sir. What if I want to make a plot like that? Do you have an idea to make a beam pattern plot using Matlab or Python? thanks
You might like to watch this video: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html and you'll be able to find the Matlab code I wrote for it on my webpage: iaincollings.com under the "Multiple Antenna Communications" heading on the "Digital Communications" page.
Thanks this made me understand it better! It still left me with a question though. If you look at an angle such that the phase shifting is exactly a whole period, the waves will still add up constructively there. How is it made that these side maxima are so weak in phased array antennas?
Yes, that's a good observation. This effect depends on how far apart the antennas are spaced. Check out this video for more details: "How does Antenna Spacing affect Beamforming?" th-cam.com/video/amepyf9-E1w/w-d-xo.html
Excellent info. I wanted to understand beamforming microphone this explains a lot!
Glad it was helpful!
does beamforming work with a single antenna receiving the signal on different anguls?
Yes. This video explains it: "What is Multi Channel Beamforming?" th-cam.com/video/e6scR0C4RAc/w-d-xo.html
If im an Client Router Beamforming as Repeater, will it work beamforming from Ap that dont have mimo ?
My Ap Is moving Object here.
Please help me answer ?
thank you so much for your video. I have a question? for a mathematic student, which books or courses do you propose to learn these concepts deeply? best.
Here's a good book: D. Tse and P. Viswanath, “Fundamentals of Wireless Communication”
@@iain_explains Thank you so much
great video, thanks
excellant!such a clear lecture to talk about beamforming. Thx very much
Glad it was helpful!
Wow. That is the best explanation I’ve ever heard. Thank you so much.
That's great to hear. Glad it was helpful!
Can you use this for multiple microphones?
Yes, absolutely. Applications include sound source localisation, and 3D surround sound headphones. Also submarines tow arrays of microphones behind them, to do target localisation.
Thank you for the video. How is this used to calculate angles for data coming from say a short range radar?
The angle is a simple function of the delay and the antenna spacing (it is briefly explained at the 7 min 30 sec point of the video).
sin(theta) = Delta * c / d
where theta is measured from the line perpendicular to the line that the array elements are on.
@@iain_explains I see, Thank you. So fft gives the frequencies and that then is used with that formula to calculate the range?
In radars, the range is generally calculated by sending a pulse and then measuring the time it takes for it to bounce back.
@@iain_explains For pulse radar, this is the case. But for the FMCW radar, it is using the FFT to get the range, as the frequency is proportional to the range in the FMCW waveform.
Such a simple yet clear explanation, which even hour long fancy videos are unable to provide
Glad you liked it
Can this be used to determine the location and distance to the sources?
It can be used to find the direction of a source, by scanning through different phase shifts and looking for which one gives an energy peak/spike. But the distance can only be found by sending a radar pulse, and listening for the echo, or by having multiple arrays spaced widely apart, and triangulating the directions.
Thanks a ton Ian. Not what I thought beamforming was. Excellent explanation of the theory of beamforming. Do you offer any in house training?
I'm glad you liked the video. I'm not offering in-house training or tutoring at the moment, but potentially will in future. For now, the videos will have to do. Please do let me know if there are particular topics that you'd like to hear more about if I haven't covered it so far. Check out my web page which shows a fully categorised list of the videos: iaincollings.com