Li-Fi (Visible-light communication) using OFDM

Hi, I am currently in the research phase of my University project.
We are to investigate the different types of modulation techniques and deduce the most suitable one for smart-home remote use. (e.g- powering up the A.C or lights).
Question: Is it possible to implement OFDM modulation on an Arduino ?

Note: I have not found an example or prototype research paper showcasing this ability thus far.
I understand that it requires mathematical calculations such as Discrete Fourier Transform.

I think so. As long as it is purely for research, and not a practical system.

How will you be modulating the light? How receiving it?

Yes, but the devil is in the details - what symbol rate? What subcarrier modulation? Which Arduino(s)? The Uno for instance is unsuitable as it doesn't use a quartz xtal for the clock, OFDM requires very solid timing to correctly separate the subcarriers.

Practical problems like synchronization and clock-recovery are important in a practical system, just generating OFDM is the easy bit.

Yes, and you will need a linear transmission channel. So on/off modulation based systems like existing IR remote TX/RX can't be used.

Hi, thanks for your response.
It is purely for research, so practicality isn't much of a concern.
Biggest hurdle is identifying the hardware capability to modulate/demodulate.
As of now, we are planning to use Single-carrier On-Off Keying, Multi-carrier OFDM, and possibly Colour modulation using Colour-shift keying. On the receiving end there will be a photodiode connected to an Arduino which will be used for demodulation.

Such a thing really exists?

Apologies, I did not phrase it properly. It's Single-carrier (On-off Keying) followed by Multi-carrier (OFDM)

When the multi frequency OOK carriers are combined, do they not require a linear communications path? If so, why use OOK when PSK for example is superior in most regards? What linear transducers can you use? You mentioned photodiode for RX, what is your TX?

Our experiment doesn't require two-way communication, so linear path is required.
Yes that is a good point.
It's just from reviewing research papers that I find this OOK method is the one most commonly used.
I am not familiar with linear transducers yet. Will have to look into more detail into that.
Tx will be a high powered LED- around 3W

Thanks for the response MarkT.
Will need to do more research on what symbol rate we will use and subcarrier modulation.
We were recommended to use Arduino Nano which to my knowledge is the same as UNO but more ports or connections.

Sorry, I think you believe I am referring to some signal path in space. I am talking about amplitude domain linearity. For example OOK is non-linear, doesn't require transmitters, receivers or signal paths to have a linear functional response. But for example, audio does.

So, you need a LED driver that can operate in a linear region, not just on-off. In other words, continuously variable between full off and full on and you need to drive it with a D/A not just a toggle...

The same thing applies to the receiver, since the demod is in software, you need to amplify and present the linear photodiode signal to an A/D.

Thus I suggest that you get yourself an Arduino that has both a D/A and an A/D.

Yeah I had mistaken it for signal path earlier.
LED driver is one of the things I couldn't get my head around earlier.
I understand better now with the explanation you gave.
Still need to do more work on the research part from what I've learned here.
I have read examples of people using MATLAB entirely for the signal processing for modulation and demodulation rather than do it on an Arduino as well.
Thanks for the advice, appreciate your expertise.

Well, I'm sure you would get extra credit for an actual hardware implementation. Do you have anyone with a lot of electronics experience on your team?

For demo purposes, it would be okay if the encoding was slow, due to long delays performing FFT and so on... just send small packets at long intervals. But Nano memory is limited and it has no DAC...

No, unfortunately we do not have a group member with good electronics experience.
Yeah, speed isn't a huge concern as long as the process is successful.
Glad you pointed out the lack of DAC and limited memory on the Nano version.
Will need to put in more research on the hardware specs you mentioned.
I will take leave for now, it was good learning from you.