2
\$\begingroup\$

According to this datasheet, Semtech LoRa can reach 9375 kbps, while CC1125 from TI has a maximum datarate of only 200 kbps, despite both operating on the same frequency (868/915). To reach its highest datarate, LoRa must be used with it's lowest spreading factor, and CC1125 uses narrowband transmission, so the it seems to me that the difference in datarate should not be so great, and especially that it should be higher for CC1125.

What are the key factors I should look into to better understand what datarates are possible for sub-1GHz systems, and thus better evaluate the available systems (bandwidth/spreading/modulation/protocol/etc)?

\$\endgroup\$

3 Answers 3

2
\$\begingroup\$

This is a error in the application note. In the table of page 22, between the SF10 line and the SF9 line, the decimal dot was lost. The table should read:

LoRa SF = 9 -> 1.757 kb/s or 1757 b/s
LoRa SF = 8 -> 3.125 kb/s or 3125 b/s
LoRa SF = 7 -> 5.468 kb/s or 5468 b/s
LoRa SF = 6 -> 9.375 kb/s or 9375 b/s

The LoRa modulation is perfect for low and medium data rates, but cannot be used at high data rate. That why Semtech transceivers support LoRa AND FSK/MSK/OOK, to offer a full range of data rates (and be able to interact with legacy systems).

\$\endgroup\$
2
\$\begingroup\$

It's down to bandwidth. LoRa is wideband, CC1125 is narrowband.

The app note you link to starts, on page 6, with a restatement of the Shannon-Hartley Theorem for data transmission over noisy channels. Nobody answering this question, including myself, is going to do better than restating that capacity theorem. Read that through a few times, and understand it.

As for the actual implementation, LoRa's chirp method of spreading does offer a few neat implementation advantages over other methods, like sweeping doppler, timing and frequency accuracy all into one easy to compensate for error. But that's details, it's not fundamental.

\$\endgroup\$
2
  • \$\begingroup\$ LoRa datarates can vary from 1172 to 37500 bps with the same 500 kHz bandwidth, by varying the spreading factor, so it seems to me that there are other factors also worth considering. \$\endgroup\$
    – Jacob
    Jun 30, 2016 at 10:02
  • \$\begingroup\$ I've asked another LoRa-related question here. Also I've noticed there is now a lora tag, I don't have the points in this SE to add it to the question above, but maybe it applies? \$\endgroup\$
    – uhoh
    Dec 27, 2016 at 11:29
0
\$\begingroup\$

It all comes down to modulation. Different modulations are capable of different data rates. Bandwidth and frequency are properties of a modulation.

Wider bandwidth can make it easier to have higher data rates but two modulations with the same bandwidth very well may have different data rates. Sometimes that bandwidth is used for other things like better receive sensitivity (range) or simpler required hardware.

Frequency is similar in that it can allow for higher data rate, but that doesn't mean it's actually used for such in a particular modulation.

Protocol is a little higher in concept than modulation and not really as relevant (it's analogous to rules for who speaks rather than the language being spoken).

Spreading is a property of the modulation and is one way to use the bandwidth for better range.

At the end of the day, you need to look for high data rate modulations.

\$\endgroup\$

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.