it's a great chance to learn this video but really need to take more time to understand the all
@themathguy3149
2 жыл бұрын
Please could you never ever ever stop making videos? You would be an Oasis in the desert if i had found this videos in my signals and systems or microwave courses 🤭
@iain_explains
2 жыл бұрын
I'm glad you like the videos. Better late than never, hey?
@shayaganji1481
2 жыл бұрын
Dear Professor Collings, I really appreciate your work, you have the best methods to explain everything in its best way and I was looking for such a method of teaching for all of my life. thank you very much for your amazing videos and all the deep conceptual explanations
@iain_explains
2 жыл бұрын
Thanks so much for your nice comment. I really appreciate hearing when people find the videos helpful.
@stevens7138
5 ай бұрын
Hi professor, could you please explain why I cannot have 4 streams in a 3x3 MIMO communication even if I have 4 spatial paths? I know from antenna point of view that 3x antenna can only have 2x nulls which means that we can only do 3 beams with zero forcing. But can’t we have some beam to beam interference while still maintain good SNR for 4 streams?
@iain_explains
5 ай бұрын
You would need the beams to be narrow enough to resolve the 4 spatial paths - but to make them that narrow it would require 4 antennas.
@julysky100
5 ай бұрын
Thank you, do you have video to talk about Analog vs. Digital BF?
@iain_explains
5 ай бұрын
Have you seen this one? "What is Hybrid Beamforming?" kzitem.info/news/bejne/14CoqqiQa2Z8aoY
@julysky100
5 ай бұрын
thank you!
@pitmaler4439
2 жыл бұрын
You say the second path values doesn't change the dimension of the matrix H -> when entry of H(1,1) is e.g. 0,34exp(j*0,8) for one path then with the second path H(1,1) changes to a summation of 2 complex numbers e.g. H(1,1) = 0,34exp(j*0,8) + 0,63exp(j*0,2) Is that right? Path1: amplitude: 0,34 phase: 0,8 Path1: amplitude: 0,63 phase: 0,2
@iain_explains
2 жыл бұрын
Yes, exactly.
@manutauer
Жыл бұрын
Thanks for your insightful video. You said min(m,n) determines the degree of freedom. That's the number of max streams. Isn't it possible to send more streams theoretically? Let's say m=4 and n=6. I want to transmit 6 streams. I think with 4 Tx, you can generate a beampattern that has more than 4 beams. We can than chose the 6 strongest beams and place the Rx there. So I could have 6 streams. In this case, I could mix the 6 streams together and transmit over my just 4 Tx. Of course, that wouldn't be very ineffective and I would have (bad) not-narrow beams with perhaps interference problems. But, could that theoretically be done this way? (My antenna pattern can give me more beams than min(m,n) -> some of them are quit small of course - that's what is not really clear to me.) Thanks a lot.
@iain_explains
Жыл бұрын
You must be talking about one of my other videos. Your comment doesn't relate to this video at all. I assume you're talking about one of my MIMO digital communications videos. Assuming that's the case, my answer is that you could send 1000 data streams if you want to (or more!), but it doesn't mean they will be able to be separated at the output of the channel, and successfully decoded.
@geze2004
Жыл бұрын
Great Video, Thanks! H^dagH and its eigenvectors/values show up alot.. but why that is does not jump out at me. Does this have a name and is it covered somewhere?
@iain_explains
Жыл бұрын
It comes up most simply/directly in Least Squares Estimation. See: "What is Least Squares Estimation?" kzitem.info/news/bejne/o5BvvKCjpqegp4I
@maitj3018
2 жыл бұрын
Hi professor, thx. If I have understood it right. In this video you have discussed one particular type of precoding at the Tx (to maximize rate) and MMSE receiver. Can we do it for other combinations of precoding (MRT) and receivers combination (like ZF, MRC). In that case B will be different but is it good for multichannel scenario discussed?
@iain_explains
2 жыл бұрын
Sorry, I'm not really sure what you're asking.
@wumeng8619
2 жыл бұрын
Hi Professor Collings, it seems you have re-uploaded this video and my previous comments were gone. Just thinking about how to map MIMO layers to physical channels that different layers experience spatially separated paths, and in case of highly imbalanced channel gains of these paths, would water-filling imply no usage of high rank transmission? In 5G with large number of TRX at gNB, due to high BF gain, UE could still report high RI even though the MIMO channel is ill-conditioned with such path imbalance. How would power be better allocated in this practical consideration? In addition, if different paths are assumed to arrive Rx almost simultaneously, i.e., within the delay resolution such that they cannot be separated as delayed taps, my understanding is doing eigenvalue decomposition would result in only 1 dominant eigenbeam with multiple peaks pointing in these paths (assuming co-located Rx antennas e.g., at UE without BF capability). How to map layers to different paths in this case? Last but not least, I have to say you make great videos to explain basic things thoroughly with just a pen and piece of paper, thanks, loving your channel!
@iain_explains
2 жыл бұрын
Hi yes, sorry about re-uploading. I realised I had said something in the original version which was correct for MC-BF, but I'd said it in a way that implied the result was more general than it was. So I decided to edit that bit out, to avoid possible confusion. In terms of your questions, the eigenvalues determine whether the water filling solution puts power into all of the eigenvectors or only a subset. It's impossible to say anything more specific than that, without actually calculating the solution for a specific channel matrix. In terms of your comment about the delayed taps - that's not the case. I probably should have been more explicit about pointing out that everything in the video is for narrowband channels (in other words, no ISI). The "shape" of the "beams" is not as simple as either narrow beams in the direction of each of the collections of paths, or as you suggest they might be, having multiple peaks in the direction of multiple paths - although in some situations they might be. It is simply a case of what the eigenvectors of H^dagH turn out to be. In a general sense, they will be directed towards the main collections of paths, but the exact shapes are hard to say anything about in general. Thanks for your comments. I'm glad you like the videos.
@nazihomayounnasab6419
Жыл бұрын
Dear Professor Collings, at first, I would like to thank you for the great informative videos and the simple deep conceptual method of teaching. I have a request, is it possible that you have an explanation about 5G channel modelling? Thank you again and Best Regards
@iain_explains
Жыл бұрын
I'm glad you like the videos. Thanks for the suggestion. I'll add it to my "to do" list.
@Saens406
2 жыл бұрын
so you have the transformation B on your signal to precode the information, which will create beam patterns. But you also have to apply the pinv(A) at the receiver. I don't get why. Is it some sort of match filter to max the SNR or is it needed for the beam forming to work? Also what is u_H? Also how the antennas know which path to direct the beam forming from as the channel impulse response is the sum of the 2 paths.
@iain_explains
2 жыл бұрын
The matrix that gets applied at the receiver combines the signals from all of the receive antennas. It is a good idea to combine them in a way that achieves your aims (eg. maximising SNR). The matrix U_H is explained at the following time point in the video: kzitem.info/news/bejne/xmypyYZmfGaDdJg The antennas "know" which path(s) to direct the beamforming by calculating the U_H matrix.
@ali_abdi
6 ай бұрын
Super! Thank you
@iain_explains
6 ай бұрын
Glad you liked it.
@MrHjld
2 жыл бұрын
Hi Professor Iain, are you accepting any PhD students?
@iain_explains
2 жыл бұрын
Yes I sure am. Send me an email, I'd like to hear from you.
@ahmadzaklouta
2 жыл бұрын
Thanks, Can you make a more detailed video about B matrix?
@iain_explains
2 жыл бұрын
Thanks for the suggestion. I've put it on my "to do" list.
@asinegaasinega
Жыл бұрын
so this is the part that's confusing to me here. Is this what's called a spatial stream? (is there a book you would recommend to study this? I dont mind the math but most of the time, I have found writers at this level are trying to impress someone. I can't stand that. Maybe if you would recommend a book that you have come across?) Thank you for doing this. This and equalization are the two aspects of comms I really need to hunker and get a deep deep understanding.
@iain_explains
Жыл бұрын
Yes, that's right, there are r spatial streams.
@asinegaasinega
Жыл бұрын
@@iain_explains thank you again for putting these up. I am very grateful
@DsBoyan
2 жыл бұрын
Great video! Love your channel! Thank you for making such information publicly available.
Пікірлер: 41