1
\$\begingroup\$

Disclaimer: I am decent with math/theory, but I have zero hardware knowledge (apart from some basic microcontrollers).

I'm looking to build a (possibly crude, but functional) hardware prototype of a receiver with an antenna array. The application involves frequencies in the 2.4 GHz ISM band. I am experimenting with array processing algorithms and would like to be able to play around with the array geometry a bit. I would also like to do most of the processing in software. Given my background though, I have very little clue how to start. In particular, I've been trying to read data sheets for various types of components, but I can't figure out what the components expect as input and what they output.

Here's what I've figured out so far:

I'll need a couple of antennas, of course, close enough so that I don't get spatial aliasing. The antennas will need to be connected to an RF front end, which does the downconversion. At some point, I'll need an ADC for sampling, and then I'll need a 'bridge' into software.

Assuming what I just described at least makes a little sense, here are my questions:

  1. How do I bridge the RF front end and the ADC?
  2. How do I bridge the ADC and the software part?
  3. Can I make things a bit simpler?, e.g. can I try oversampling the received antenna signals and then just doing everything in software from there?

Additionally, is there a decent resource for learning about practical RF circuitry and prototyping? I think that is perhaps where I am weakest and need the most help.

Thanks!

\$\endgroup\$
  • \$\begingroup\$ As far as arrays, what exactly are you trying to do? There are quite a few difficulties in designing a working array at 2.4 GHz. \$\endgroup\$ – Yuriy Dec 18 '13 at 5:13
  • \$\begingroup\$ Yuriy, I want to test out some angle of arrival algorithms (beyond MATLAB simulation). I can't be more specific about the application, unfortunately. \$\endgroup\$ – user34309 Dec 18 '13 at 10:53
2
\$\begingroup\$

1- an RF Receiver is what you want to sit between the Antenna and ADC. Something like the Lime LM6002 or the AD9361 transceiver IC's might make this a lot easier. There are Eval boards that mate up with FPGA eval boards for an easy (Expensive) solution. Do you need to have these two channels to be coherent?

2- An ADC might have 2 channels (I and Q) of digitized data clocked at 75 or 100 MHz, depending on the bandwidth of the signal you're trying to receive. Perhaps at 12 to 14 bits of bit width, so 1.8GBit/Sec of data to process. This generally goes into an FPGA to do downconversion and digital filtering, with the lower data rate info from the FPGA moved into a DSP for signal processing.

3-You can't really oversample a 2.4 GHz RF signal with a current technology ADC.

There are several things that might make this easier. Doing this on the FM broadcast band would make this much simpler, as the bandwidths and frequencies are much lower, and easier to deal with. Using narrow band channels for your Experimentation would also help, as that would reduce the RF BW, and reduce the bitrate out of the FPGA.

(Edited to add TV Tuner Dongles) Another dirt cheap possibility is to use a TV Tuner dongle which would work for FM broadcast band. Using that and GNU Radio would take care of a lot of the hardware issues, but wouldn't support the 2.4 GHz ISM band. Some more details here http://sdr.osmocom.org/trac/wiki/rtl-sdr

\$\endgroup\$
  • \$\begingroup\$ I don't understand what you mean by eval boards that "mate up" - could you please elaborate on that? I also don't quite get what two channels (question 1) we're talking about. \$\endgroup\$ – user34309 Dec 18 '13 at 11:11
  • \$\begingroup\$ For example, the Zedboard, which is a development board for the Xilinx Zynq FPGA, has connectors which will mate up with the Analog Devices AD9361 transceiver link which would give you an off the shelf eval board instead of spinning your own boards, which would be fun but probably a real yak shaving detour from what you want to do Receiver interfaces to FPGA's tend to be done in IQ format, with the RF channel downconverted to 2 quadrature channels to make the signal processing easier. \$\endgroup\$ – rfdave Dec 18 '13 at 12:02
  • \$\begingroup\$ I looked a bit more into software-defined radio (hackRF, USRP, bladeRF), and they do seem promising. They're a bit more expensive than I expected, but that appears to be the norm with transceivers at the 2.4 GHz band. Thank you for the help! \$\endgroup\$ – user34309 Dec 19 '13 at 1:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.