Your presentations are absolutely brilliant! The progression of the learning curve is excellent, starting with the "mathless", going to the linear array, and ending in the rectangular array.
Thank you for your excellent presentation I read many popular books (Balanis, Van Trees, Spagnolii) but your explaination of the 2D array factor computation is by far the best one.
I just think the "observation window" thing is simply explained by the beamwidth, where if you walk over the same transverse distance at low antenna distance you'll see the beam gain variation, whereas at large distance your same observation window will not see much signal variation, because you'll remain close to the maximum gain "center" of the beam. I also suggest that instead of "painting" a 100lambda sphere with the beam factors, you could draw it as a 3D directivity "blob" pattern, and even in dB. But that's a matter of taste, of course!
Great video! I was just wondering about the simplification at 27:30 where the exp(jz/2) gets simplified away, the z in the numerator and denominator are different according to the formula above so we should have a complex exponential left in the end right?
You are right. To be precise, S_x from this slide should be multiplied by the factor exp( (j/2)*(M-1)*psi_x). My interest in these slides was to highlight how the array amplitude behaves, so this module-1 imaginary factor has no impact in this sense. But yes, an explanation should have been given in this respect. I will take this aspect in consideration in future videos. Thanks for your comment.
Wonderful lecture. How is it possible to generate phase shifts simultaneously like in terms of a circuit? And how is the phase shifted - is it done by a voltage ramp? Thank you so much
Thanks for your comment. The phase may be shifted using a digital or analogue approach. Digital phase shifting is based on the Fast Fourier Transform (FFT). It is the most flexible alternative but requires a powerful Digital Signal Processor (DSP). Analogue Phase shifting makes use of semiconductor devices with properties (e.g. capacitance) that depend on a control voltage as you mentioned in the last part of your question. Typically, 5G systems operating in centimeter wavelengths use digital phase shifting, while millimeter systems use the analogue approach.
@@5glearning772 As you've been kind enough to reply, would you answer a question? For large-N arrays (say N>1,024 elements), it is clear that for any desired steering vector V, there will be M unique 'phase weight sets' K[0,1,...,M] that will produce a steering vector that is equal to V within some error E. b = the number of phase bits ~ 6. Each phase weight set K contains N*2^b values, there are M phase weight sets per steering vector, and we are trying to minimize the number of sets per steering vector to as close to 1 as possible (removing redundant sets). Given that, what is the maximum number M of phase weight sets K such that E is equal across the entire pointing space (hemispheric shell)? In other words, if you were to pre-compute the phase weight sets K (for known N, b ; given V), and you 'binned' weight sets K that resulted in the same vector within error E, how many sets would be required to minimize pointing error E across all possible pointing vectors (a hemisphere) while also minimizing M (number of sets)? e.x. two elements (N=2) with one bit (b=1) of phase shift (0 or 180deg) will have three possible steering vectors. How can the range of *all* steering vectors be compressed into a set of *every possible* steering vector (size = 3) for a given N, V, and b?
I have been thinking on your questions. Probably I am missing some of your points, particularly on weights quantization, but let me present some of my thoughts. Consider, for example, a simple linear array consisting of 8 omini-directional elements with adjacent elements separated by half a wavelength. Assume that you would like to limit the number of steering-weights accepting a penalty in the received power due to misalignment of the steering vector. More precisely, assume that you would accept a 3 dB loss. In that case, it can be demonstrated that you just need 9 weight-sets to cover the full hemisphere. If, on the other hand, you increase the number of elements of the transmit array, the radiation pattern becomes sharper. Therefore, the same steering error leads to a higher loss than in the previous case. This issue can be compensated increasing the size of the weights-set. I hope that this may help.
That is a very great source of information, can anyone direct to a very good book or papers to ponder upon concerned to these topics, will be helpful, Thank you!
Thanks for your comment. Here are some books that you may consult on this subject: Smart Antenna Engineering Ahmed El Zooghby ARTECH Electronically Scanned Arrays - Matlab and Simulation Arik D. Brown CRC Press Antenna Theory - Analysis and Design - 3rd Edition Constantine A. Balanis WILEY (see chapter 16) Array and Phased Array Antenna Basics Hubregt J; Visser WILEY Introduction to Adaptive Arrays - 2nd Edition Robert A. Monzigo et alia SCITECH Phased Array Antennas R. C. Hansen WILEY
Thank you. The simulation code used in this project is not yet in a suitable format for publication. I would need to invest time for this purpose and regretfully I am quite busy right now.
I mainly used the following book: Constatine A. Balanis Modern Antenna Handbook 2008 - Wiley In addition, I consulted the following books: Ahmed El-Zooghby Smart Antenna Engineering 2005 - Artech Arik D. Brown Electronically Scanned Arrays 2012 - CRC Press John Litva et al. Digital Beamforming in Wireless Communication 1977 - Artech Hubregt J. Visser Array and Phased Array Antenna Basics 2005 - Wiley Robert A. Monzigo et al. Introduction to Adaptive Arrays 2011 - SciTech Publishing Zhizhang Chen at al. Introduction to Directio of Arrival Estimation 2010 Artech R. C. Hansen Phased Array Antennas 1998 - Wiley I hope that this may help.
hi thanks for the video. any easy formula to calculate the phase delay into each element for uniform reactangular array antenna to steer the beam to the direction we wanted? thank.
Thanks for your comment. The same concepts described for the radio field may be applied to Optics as well. From a practical perspective, it must be remembered that the wavelengths used in 5G radio are in the interval between 1 and 10 cm, hence several orders of magnitude longer than even infrared light. Furthermore, beam-forming at mm wavelengths is currently based on analogue techniques due to today's limitations in semiconductor technology. Therefore, current transceivers in these wavelengths cannot take advantage of the flexibility and richness that digital beam-forming using Fourier transform entails. Of course, this should not discourage research. On the contrary, it is research that paves the way for commercial applications one or two decades later. A final point: in this video we are considering that transmission and reception powers are enough to ensure that a large number of photons are acting. If this is not the case (as happens frequently in Optics), quantum-mechanical effects must be taken into account.
@@5glearning772 Thank you for the detailed response. I will certainly take head of your points in my work. I know time is a precious commodity for people like yourself...but do make more lectures to pass on the knowledge. You have a rare ability of presenting complex things in a clear and simple manner.
is it possible to have a little explanation about the difference between the grating lobe in array design and its difference with the beam related to the space harmonic in leaky wave antenna please?
This interesting subject is discussed in the book "Adaptive Antenna Arrays for Radar and Communications" written by Allan J. Fenn (ISBN: 978-1-59693-273-9).
Your presentations are absolutely brilliant! The progression of the learning curve is excellent, starting with the "mathless", going to the linear array, and ending in the rectangular array.
Excellent concept and graphical representation of 3D Beam-forming. Very good job. Thank you.
Thank you very much!
Thank you for your excellent presentation I read many popular books (Balanis, Van Trees, Spagnolii) but your explaination of the 2D array factor computation is by far the best one.
Many thanks!
Great video! Waiting on that video about precoding!
Thank you!
Excellent concept and graphical representation of 3D Beam-forming
Thank you very much!
I just think the "observation window" thing is simply explained by the beamwidth, where if you walk over the same transverse distance at low antenna distance you'll see the beam gain variation, whereas at large distance your same observation window will not see much signal variation, because you'll remain close to the maximum gain "center" of the beam.
I also suggest that instead of "painting" a 100lambda sphere with the beam factors, you could draw it as a 3D directivity "blob" pattern, and even in dB. But that's a matter of taste, of course!
Amazing video. Well done. 👏🏾👏🏾👏🏾👏🏾👏🏾
Thank you!
Very Good Presentation. ThanKs!
Excellent animation and explanation
Thank you!
Wonderful explanation
Thank you!
You are the GOAT! Thank you man
Thank you!
Great video! I was just wondering about the simplification at 27:30 where the exp(jz/2) gets simplified away, the z in the numerator and denominator are different according to the formula above so we should have a complex exponential left in the end right?
You are right. To be precise, S_x from this slide should be multiplied by the factor exp( (j/2)*(M-1)*psi_x). My interest in these slides was to highlight how the array amplitude behaves, so this module-1 imaginary factor has no impact in this sense. But yes, an explanation should have been given in this respect. I will take this aspect in consideration in future videos. Thanks for your comment.
Congratulations!!
Thank you!
Greatest! let me understand why the path difference need to multiply the spherical axes factor.
Thanks very much!
Wow that was good!
Thank you!
Wonderful lecture. How is it possible to generate phase shifts simultaneously like in terms of a circuit? And how is the phase shifted - is it done by a voltage ramp? Thank you so much
Thanks for your comment.
The phase may be shifted using a digital or analogue approach.
Digital phase shifting is based on the Fast Fourier Transform (FFT). It is the most flexible alternative but requires a powerful Digital Signal Processor (DSP).
Analogue Phase shifting makes use of semiconductor devices with properties (e.g. capacitance) that depend on a control voltage as you mentioned in the last part of your question. Typically, 5G systems operating in centimeter wavelengths use digital phase shifting, while millimeter systems use the analogue approach.
@@5glearning772 Thanks a lot mate..
Great man
Thank you so much!
I'm new on this area. And you are very good at lecturing, thank you very much. Are there any pdf file that i can reach for more to learn?
Thanks for you comments. Only the video version is currently available, due to time limitations.
Thank you
You are welcome
dude. that was really good!
Thank you.
@@5glearning772 As you've been kind enough to reply, would you answer a question?
For large-N arrays (say N>1,024 elements), it is clear that for any desired steering vector V, there will be M unique 'phase weight sets' K[0,1,...,M] that will produce a steering vector that is equal to V within some error E. b = the number of phase bits ~ 6. Each phase weight set K contains N*2^b values, there are M phase weight sets per steering vector, and we are trying to minimize the number of sets per steering vector to as close to 1 as possible (removing redundant sets).
Given that, what is the maximum number M of phase weight sets K such that E is equal across the entire pointing space (hemispheric shell)?
In other words, if you were to pre-compute the phase weight sets K (for known N, b ; given V), and you 'binned' weight sets K that resulted in the same vector within error E, how many sets would be required to minimize pointing error E across all possible pointing vectors (a hemisphere) while also minimizing M (number of sets)?
e.x. two elements (N=2) with one bit (b=1) of phase shift (0 or 180deg) will have three possible steering vectors. How can the range of *all* steering vectors be compressed into a set of *every possible* steering vector (size = 3) for a given N, V, and b?
Sure
@@5glearning772 haha thanks!!!
I have been thinking on your questions. Probably I am missing some of your points, particularly on weights quantization, but let me present some of my thoughts.
Consider, for example, a simple linear array consisting of 8 omini-directional elements with adjacent elements separated by half a wavelength. Assume that you would like to limit the number of steering-weights accepting a penalty in the received power due to misalignment of the steering vector. More precisely, assume that you would accept a 3 dB loss. In that case, it can be demonstrated that you just need 9 weight-sets to cover the full hemisphere.
If, on the other hand, you increase the number of elements of the transmit array, the radiation pattern becomes sharper. Therefore, the same steering error leads to a higher loss than in the previous case. This issue can be compensated increasing the size of the weights-set.
I hope that this may help.
That is a very great source of information, can anyone direct to a very good book or papers to ponder upon concerned to these topics, will be helpful, Thank you!
Thanks for your comment.
Here are some books that you may consult on this subject:
Smart Antenna Engineering
Ahmed El Zooghby
ARTECH
Electronically Scanned Arrays - Matlab and Simulation
Arik D. Brown
CRC Press
Antenna Theory - Analysis and Design - 3rd Edition
Constantine A. Balanis
WILEY
(see chapter 16)
Array and Phased Array Antenna Basics
Hubregt J; Visser
WILEY
Introduction to Adaptive Arrays - 2nd Edition
Robert A. Monzigo et alia
SCITECH
Phased Array Antennas
R. C. Hansen
WILEY
@@5glearning772 Thanks a ton!
nice presentation and explanation. Is it possible to put matlab code that you used in the clipe, please.
Thank you.
The simulation code used in this project is not yet in a suitable format for publication. I would need to invest time for this purpose and regretfully I am quite busy right now.
Can you share the references you used for these videos? Thank you!
I mainly used the following book:
Constatine A. Balanis
Modern Antenna Handbook
2008 - Wiley
In addition, I consulted the following books:
Ahmed El-Zooghby
Smart Antenna Engineering
2005 - Artech
Arik D. Brown
Electronically Scanned Arrays
2012 - CRC Press
John Litva et al.
Digital Beamforming in Wireless Communication
1977 - Artech
Hubregt J. Visser
Array and Phased Array Antenna Basics
2005 - Wiley
Robert A. Monzigo et al.
Introduction to Adaptive Arrays
2011 - SciTech Publishing
Zhizhang Chen at al.
Introduction to Directio of Arrival Estimation
2010 Artech
R. C. Hansen
Phased Array Antennas
1998 - Wiley
I hope that this may help.
hi thanks for the video.
any easy formula to calculate the phase delay into each element for uniform reactangular array antenna to steer the beam to the direction we wanted?
thank.
Great lecture of a hot research topic. Is it reasonable to apply similar concepts to the so-called metasurfaces in the optical regime?
Thanks for your comment.
The same concepts described for the radio field may be applied to Optics as well.
From a practical perspective, it must be remembered that the wavelengths used in 5G radio are in the interval between 1 and 10 cm, hence several orders of magnitude longer than even infrared light.
Furthermore, beam-forming at mm wavelengths is currently based on analogue techniques due to today's limitations in semiconductor technology. Therefore, current transceivers in these wavelengths cannot take advantage of the flexibility and richness that digital beam-forming using Fourier transform entails. Of course, this should not discourage research. On the contrary, it is research that paves the way for commercial applications one or two decades later.
A final point: in this video we are considering that transmission and reception powers are enough to ensure that a large number of photons are acting. If this is not the case (as happens frequently in Optics), quantum-mechanical effects must be taken into account.
@@5glearning772 Thank you for the detailed response. I will certainly take head of your points in my work.
I know time is a precious commodity for people like yourself...but do make more lectures to pass on the knowledge. You have a rare ability of presenting complex things in a clear and simple manner.
is it possible to have a little explanation about the difference between the grating lobe in array design and its difference with the beam related to the space harmonic in leaky wave antenna please?
Sorry for being unable to provide an answer to your question. I have no experience on leaky-wave antennas.
thank you
Welcome
Still having curiosity on impedance of mutual coupling among antenna array. How to calculate that in real case?
This interesting subject is discussed in the book "Adaptive Antenna Arrays for Radar and Communications" written by Allan J. Fenn (ISBN: 978-1-59693-273-9).