Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Brief Introduction of my IC Studies

Status
Not open for further replies.

satnetmodem

New Member
##################################
Brief Introduction of my IC Studies
##################################

Why IC Technologies
================

The reasons why I finally come to IC research is that, of all technologies now available in the world, IC technologies are playing the single most important role in human development now, and for the foreseeable days. Without it, we will return back to the mechnical times.

Previous Studies
=============

But I don't have the necessary prerequisites and capacity to conduct IC research, such as an expensive lab or a great team, therefore I have to satisfy myself with some mathematical studies in connection with IC technologies.

Of course, almost all math branches are useful in IC technologies, but what I am interested in mostly is the design stage, and particularly the physical design stage, and over there graph theory has been playing a big role, as such I have spent quite some time on graph theory for IC design.

My first topic was graph theory for channel routing, which is the last phase of the physical design and hence also the last phase of the entire chip design process. I have explored a great number of routing algorithms, particularly for FPGA based logic, but finally I decided not to pursue it further because I consider it a waste of time to focus on traditional, silicon based IC technologies, as it is approaching a scaling limit, that is, there is no way to improve the computing power after the CMOS scale reaches nanometer. In other words, it is very difficult, if not impossible, to produce more powerful processors based on the current silicon chips. Therefore one has to find new materials, new physics and new technologies.

And then I have tried a number of new generations of IC technologies, among them, quantum circuits, graphene based circuits, superconducting chips, resonant tunneling circuits, 3D ICs, spintronics based chips, to name a few. But I found that none of them are able to replace the current IC technologies, at least for the near future.

Nevertheless, I have studied somewhat intensively about the graph theory for quantum computing, and at that time I was particularly interested in the socalled Chimera graph applications to quantum computer, which was devloped by D-Wave, a Canadian quantum computer pioneer. In connection with Chimera graph was my indepth review of graph embedding and related branches of graph theory.

But this study was stopped about end 2017, when I determined to shift my focus from graph theory to really IC technologies.

Current Focuses
==============

After my studies of graph theory for IC technology was cancelled, I've decided to move on with my real IC technologies, and this time it is divided again into two parts: the new generations of IT technologies and the traditional, silicon based computer chips.

It is mentioned before that it seems no breakthrough will be achieved in the near future, however, whatever new generations of IC technologies are going to emerge, they must have something to do with quantum physics, because at the microscopic level, it is always quantum phenomenum which is playing a vital role in the working of the systems. Therefore, I decided to focus on quantum theory and quantum chips as the leading futuristic IC research direction. I plan to write an indepth review of quantum chips in the years to come, beginning with qubits - quantum bits over the next year or two.

On the other hand, I don't want to forget my expertise in silicon IC technologies, therefore since August 2017 I have been working on a design project, first for a hardware monitoring chip, and then recently for a satellite broadband Internet chip. I think satellite Internet will eventually replace the current cable and optical Internet, at least in long run, it will be a scene to come, just as the radio broadcasting has replaced the wired broadcasting many years ago.

For this design project, I have to learn a lot, incl. satellite communication, radio (wireless) communications, and radio frequency modulation and demodulation. My current design project is titled "Chip Design for Satellite Broadband Internet (Satnet). Part I: Modulator and Demodulator (Modem) IP Core". I am currently preparing for a document with the same title, and I hope in a couple of months, the design specifications and design entry will be completed, along with decision about technology of choice (FPGA based, ASIC or others) as well as design tools availability.

For more information about this design project, see 000terminals_indoorunits_modems_chips_toc.txt for details.

Future Plans
==========

The above studies, both quantum chips and satellite broadband Internet related chip design, will remain to be my future orientation of research and design, at least for a decade or so. I will try to write something in these topics, and design some chips or at least some IP cores, as it is a popular fashion now in the industry.


Of course, there are tremendous hurdles and difficulty ahead of my plans, but whatelse has no difficulty if you want to do something?

All the above studies, researches and designs are only my hobby, and I have no intention to commercialize my research and design, and no timetable for these activities and no time-to-market pressure. The only aim is to keep me, my knowledge, and my expertise updated with the development of one of the most important technologies of our age.

I am also well aware that I am far from being specialized in this field. There are a great number of people who are real specialists, both young and experienced. Nevertheless, I will continue my journey on this road and see if I can enjoy my studies someday, somehow.

I am doing all of my above researches on my spare time, because I have to earn a living for myself and as such I can only allocate very limited time to these researches and studies.
 
I think satellite Internet will eventually replace the current cable and optical Internet,
Don't base your future on that!

Satellite links are pretty much useless for serious internet connections, simply due to propagation delays.

"Ping time" is everything when it comes to responsiveness of such as web pages - and especially gaming.
Add the uplink and downlink delays due to speed-of-light propagation to and from a satellite and the delay becomes impractical for many uses.

They could be fine for such as email forwarding and archiving sites etc. where slight delays have no consequence, but not for normal web & internet access.

See these articles, as an example:
https://en.wikipedia.org/wiki/Satellite_Internet_access#Challenges_and_limitations
https://arstechnica.com/information...ster-than-advertised-but-latency-still-awful/
 
Last edited:
LEO (Low Earth Orbiting) satellite can solve this problem.

LEO would need either massive power levels (with associated interference between users and across different sats, so not practical) or tracking antennas - and redundant tracking antennas if you were not going to lose the connection each time it had to switch to a different satellite.

As well as only reducing and not eliminating the propagation delays.

Also; downlinks are simple, as the satellite can use multiplex carriers with a large number of logical channels on each.
How do you propose to handle a massive number of independent user uplinks in the limited spectrum available??

Most existing satellite internet system use terrestrial links for the outward data from each user, as far as I am aware.
The actual satellite connection is just a fast downlink.


From a quick search, the functional link time (high enough overhead) of a LEO sat is about 15 mins max..
Remember that they are _not_ geostationary so relative to the ground every orbit has a different path and the same sat may only be in range of one ground station a couple of times each day.

You would need literally hundreds if not thousands to provide continuous, uninterrupted coverage. It's just not in any way practical.
 
You would need literally hundreds if not thousands to provide continuous, uninterrupted coverage. It's just not in any way practical.

Not at all, in fact it's all ready under way (the first satellites were launched earlier this year) - but in addition to existing networks, not to replace them - and intended for areas where there is no existing network, and any Internet at all would be a great improvement.
 
LEO stats have typically 90 - 120 minute orbital periods and very short visibility from any one point. It need a very large amount to have at least one permanently in range of any particular point on earth.

This is the satellite "constellation" planned by just one sat internet company - it appears to have something over 150 individual satellites..
https://4.bp.blogspot.com/-jfk553bt...Yy0ARLgCLcBGAs/s1600/teledesicbig+%281%29.gif
 
LEO stats have typically 90 - 120 minute orbital periods and very short visibility from any one point. It need a very large amount to have at least one permanently in range of any particular point on earth.

This is the satellite "constellation" planned by just one sat internet company - it appears to have something over 150 individual satellites..

No one said it was easy or cheap, but it's happening :D

There are also more than one company launching such systems.

However, as I've already said, it's NOT a replacement for fibre or indeed ADSL.
 
LEO (Low Earth Orbiting) satellite can solve this problem.
No one said it was easy or cheap, but it's happening :D

There are also more than one company launching such systems.

However, as I've already said, it's NOT a replacement for fibre or indeed ADSL.

Please refer to my page at **broken link removed** where you can find some of the pioneers in this field. I guess the day will come when all wired Internet will be wiped out by satellite based wireless Internet. Of course there are numerous challenges to that target, but who knows what will happen in the next few years. Just a decade ago people were still using landline phones, but now only few are still using wired phones. I am optimistic about the satellite future.
 
LEO (Low Earth Orbiting) satellite can solve this problem.

Please refer to my page at **broken link removed** where you can find some of the pioneers in this field. I guess the day will come when all wired Internet will be wiped out by satellite based wireless Internet.

Then you guess wrong, as already explained to you.

Why would you even imagine it's a good idea?.
 
As an IC layout guy for 14 years there is quite a bit of math, but the tool does most of that for you. Oddly the one formula I use the most is Pythagorean theorem to determine the distance between two points. <-- Mainly when creating standard cell libries for a new process design. Although the tool does that for you as well, there are some situations where you must do it manually. With FinFET transistors moving into the 3D aspect of layout, Moore's law is continually pushed to it's limits. What was once recessive characteristics in silicon behavior, are now dominant factors and what was once Dominant are now suppressed.... What's beyond FinFET? I'm not sure, but FinFET's apply vertical techniques to move away from the substrate and avoid leakage. I would imagine some FinFET derivative that also applies horizontal aspects once the substrate has been cleared. Imagine an upside down tree structure. The real problem we will face is that a covalent bond of two Aluminium atoms is only about 3 Angstroms or 0.3nm ... if your technology process is 10nm then that's only about 200 Atoms wide. To put that in a different perspective ... in the time it has taken me to type this(15min), my fingernails have grown 1000nm ... enough "NEW" real estate to fit several transistors :)
 
I guess the day will come when all wired Internet will be wiped out by satellite based wireless Internet.

Only if you have also invented a faster-than-light data communicator.
Propagation delays to/from the sats are the real killer for such concepts. With speed of light restrictions, you simply cannot wish those away.

Satellite internet is a last-resort system for places that cannot get an internet connection by any other means.
That's what it is good for.

The only reason I can see for anyone claiming it will take over is as an investor scam, like the various flying car projects and "free energy" systems that various people have used to sucker non-technical investors in the past, with zero practical products ever appearing.

(Of course, of you had the FTL data link it would not be based on RF so you would not even need the satellites in the first place...)
 
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top