Transports

You’re very lucky! You have tons of options on how you can connect your Thing to a wider network. As we’ve mentioned, there are trade-offs and design considerations with each method.

Since the transport you choose has massive implications on your design, you’ll want to nail this down first. It’s also very hard, if not impossible to change transports after you've started. Pick the right transport and the rest of your design will fall into place.  

We’ll start with the oldest/stable transports first and then branch into the more esoteric ones. This isn’t to say you have to stick with the first one we mention - it’s just you’re most likely to be familiar with it!

What is a Transport?

What do we mean by transports?

The most common way we 'split up' a datapath is to use the 7-layer OSI model (not to be confused with the delicious 7-layer burrito).

Below is a generic diagram of the OSI layers:

Now ideally we could pick and choose each layer of our connectivity - so you can swap out any layer to get the perfect balance of power/range/throughput.

But that's not how the real world works. Often times the Transport layer (which, above, is the 4th layer) bleeds through both below and above. While in theory you could run Ethernet over a wireless LoRa link, it would be...strange. We're going to assume that you're using the transport/network/data/physical set together as intended.

Sometimes, these are deeply practical considerations: if you have Ethernet, we assume you've got standard Cat-6 wiring, Ethernet jacks, MAC addressing, and using TCP/IP over it - all those layers are closely tied together and optimized.

If you have a cellular transport, you've got GSM/GPRS/LTE and are using the world-wide cellular network. Doing anything else would require low-level hacking and probably violate the terms of use for the cellular network.

And note that some of the transports don't guarantee end-to-end connectivity, or it may be your responsibility - especially radio.

So, think of transports as a rough sketch of the first 4 layers!

Last updated on 2017-10-25 at 11.39.29 AM Published on 2017-11-01 at 04.23.54 PM