Bluetooth’s third trick – Mesh networking for IoT


The Bluetooth Special Interest Group (SIG) is working on a mesh networking specification it hopes to release in the second half of 2016 that could be a game changer for the smart home and connected buildings. In this interview with Erret Kroeter, VP of Marketing at the Bluetooth SIG, he offers a preview of what we can expect in terms of range, power consumption, and backwards compatibility from what he calls Bluetooth’s “third trick.”

Can you help put some context around Bluetooth’s mesh networking initiative?

KROETER: We currently have about 30,000 members of the Bluetooth SIG, and we’re growing at a rate of about 300 companies a month. This year we’ll see well over 3 billion Bluetooth products ship, and in 2019 that’ll be close to if not over 5 billion Bluetooth products annually. You’ve got an installed base of about 8 billion Bluetooth devices.

So we’ve got this huge scale and scope. We came out with Bluetooth Low Energy (BLE) technology a couple of years ago and that was a second shot in the arm for Bluetooth. It brought in whole new industries and whole new verticals with different kinds of companies coming into the SIG and saying, “we can use BLE technology in our whatever-you-can-think-of].” That’s when you started seeing Bluetooth in wearables and it wasn’t long after that you saw Bluetooth beacons everywhere.

But one of the things that people are always saying about the technology is that it works great, it’s secure, flexible, easy to use, and consumers love it, but we want to make sure we have good coverage throughout a building or a home. So there was a lot of talk about Bluetooth’s effective range, and a lot of conversations around doing a different kind of networking with Bluetooth. So there was a lot of interest in being able to do a mesh networking topology with Bluetooth in addition to its traditional point-to-point, kind of hub-and-spoke network topology.

Last year we announced the formation of a Mesh Networking Working Group, which is essentially a committee of members that come together and work on a specification that we can standardize. A few companies in the past couple of years have come out with their own versions of Bluetooth mesh networking, and that’s good and shows that there are real products out there from large companies, so it’s definitely workable. But we want a standardized way to create these networks because the power is when anyone or thing can plug into a Bluetooth network that’s existing in the home.

We’ve been working on that Mesh Networking specification throughout last year and into this year. We have well over 100 companies that are coming together to work on this, from big companies to little companies in all different kinds of industries, which again shows the breadth of interest in adding mesh networking to Bluetooth. One of the biggest benefits of that is whole home or whole building coverage so that you can create a mesh network that spans an entire home and ensure your signal inquiry or status update on whatever device is going to get from the bottom part of the basement to the top part of the attic and anywhere in between.

How much will the mesh specification incorporate BLE technology, and what implications does this have on range and power consumption?

KROETER: The mesh networking we roll out in the second half of this year is going to be part of BLE. The objective is to have it backwards compatible all the way back to Bluetooth 4.0, so it’s not going to introduce any incompatibilities. The mesh networking protocol will ride over BLE.

Depending on the transmit power, a BLE signal can actually go up to one kilometer. But the practical range for BLE right now is about 30 meters. Looking at 30-90 meters for a BLE transmission, that’s almost enough to cover a whole home by itself, and these nodes are going to be pretty close together. CSR has claimed that they can theoretically attach 64,000 devices to their mesh network, so even if practically you get a quarter of that, you’re talking about a network that can easily span a home, and probably more like a pretty good sized building.

In terms of power consumption we shouldn’t see a big penalty in terms of power usage, because a lot of it’s just how many times you’re lighting up the radio and transmitting something. BLE can turn on, send a signal, and turn back off in 14 ms, which is an infinitesimal amount of time, so basically most of the time it’s off.

One of the typical scenarios is in lighting control. The lights are already going to be plugged in, so you’re not really worried about power there; You’re using those as a carrier for the signal to bounce along to these nodes. On the other end you’ve probably got a remote switch, and how many times are you going to hit that switch? You’re not going to hit it 100 times a day; you’re going to hit it a few times a day. If an end node asks for a status on whether a window is opened or closed, you’re not going to do that thousands and thousands of times a day either. So we don’t really see in the most likely uses that there’s going to be a big impact on power.

Your throughput likely isn’t going to be an issue either, because you’re talking about little bits of data. You’re not streaming video here. It’s just things like, “Am I on or am I off?” or “What’s the temperature?”

[Figure 1 | Bluetooth mesh networking will allow users to easily provision a range of Bluetooth devices to an existing network in the smart home and connected buildings.]

Where is the Mesh Networking group at this point?

KROETER: What the guys are working on now is making sure they’ve thought through all the different use cases as to how these things are going to light up and talk to each other, remain as quite as possible, and how they’re going to carry traffic over the network, so whether they are going to do a route or just flood the network. Those are a lot of tradeoffs that you’ve got to figure out.

Frankly, the top question we get these days is about how we’re doing mesh and how we’re extending range. So I would expect that you’ll see more companies come into Bluetooth for all of the reasons I laid out before around size, scope, easy to implement, and super low cost because of the number of chips that are being produced, and everybody’s got a controller or provisioning device in the palm of their hand. Then once you get a Bluetooth mesh network up and running, if we’ve done our job right, we have made it simple to bring any kind of product into that network. So, say you’ve got lighting control but you want to add door access sensors, you’ll be able to plug into your existing Bluetooth mesh network.

I think that’s going to be our third trick. We had Bluetooth, we have Bluetooth low energy, and now we’re going to do mesh networking that offers a pretty compelling value proposition. Bluetooth is really the technology that’s going to make the Internet of Things (IoT) accessible to everyone.

Bluetooth Special Interest Group