Almost everyone has a smartphone and they all have Bluetooth (if yours doesn't then you need to get a new phone!), and pretty much all the new ones have BLE. This makes it easy to interface with the IoT, unlike ZigBee which most consumers have never heard of and don't know what it is. Not to mention there isn't an easy way to connect to a random ZigBee device. I think the CSRMesh or a BLE mesh topology is a logical next step to opening up the world of BLE and the IoT for many applications, such as the CSRMesh lighting demo. Why would you want to control just one IoT thing when you could control multiple devices with one click? Nonetheless, from your phone that is always with you. Bluetooth is just easier for people to use and setup because they are familiar with it. If you're a consumer interested in the IoT, then your'e probably not going to want to spend a lot of time setting up every device, and buying extra components. Or God forbid paying someone to setup a network for you. Plus, BLE has excellent range just from a PCB antenna and depending on the application can run off a coin cell for 7-10 years. For these reasons, the project that I am currently working on uses BLE, actually the CSR1010, and we are very interested in the CSRMesh. We looked at all the other options to communicate with our device, including ZigBee. ZigBee is good for some things but I don't see it being a forerunner for the IoT in your home.
I agree. The smart phone revolution is absolute. It's like the Borg. Everyone is walking around with a connection to the 'collective'. The 'collective' is soon to include (in addition to text, Facebook, Internet, ...) the physical world. Android and iOS have BT/BLE and WiFi. BLE is low power enough to run from a coin cell. Mesh (and extended range) is the last key. It's hard to imagine how another standard will have a chance - ANT, Zigbee, ZWave, etc. BT and WiFi will be available for high def analog, video, large files, etc.
Mesh networks are vulnerable to attacks by hackers.
In one of the recent blogs on EEtimes there was a story about how a smart hacker could get entry into a mesh-netwroked lighting system of a house and how he could remotely control all the lights of that house .
It seems to me that end to end security is provided with TCP/iP compatible networks regardless of the mesh. Mesh schemes work ar the ip level or data link level depending on the scheme so end to end security is not impacted here (ie think end to end TLS/SSL). IPSec is also possible depending on the mesh choice.
Wireless encryption offers no end to end security.
Things like 6loWPAN offer ip compression and mesh discovery/routing. With an IP based focus, security mechanisms are preserved.
With end to end security using standard IP approachs, the higher level security protocols still work (https, ssh, sftp, ...)
The issue is murky when thinking about BLE because the security is only provided by wireless encryption and the BLE protocols are not mapped to IP (even though we have PAN in Classic mode). It seems to me that a BLE - IP would solve both the security and the portability issues but it is not really consistent with the BLE "light sensor" end of things.
CoAP would claim to solve this problem, but does not run with BLE. Thread - we don't really know because its a secret.
ZigBee already offers mesh capability. The limit of BTLE, so far, were lower number of MAX connections to a node and limited range. Anyway, allowing mesh capability could definitely heal the range issue, by node-to-node hopping. Anyway, mesh network will now require additional energy consumption for a node, on top of what needed to perform its basic tasks, as it is now supposed to help in carrying out other nodes' information.
In a scenario where we have hundreds of nodes connected to an internet HUB, we will have hundreds of batteries to change sooner or later. Node consumption and mesh protocol effectiveness in energy management will impact battery discharge time and maybe node capability to run w/o battery, possibly harvesting energy from the environment. If mesh-BTLE will proof more efficient in overall energy consumption, it will be the choce for more and more applications.
Apart from the diplomacy expressed in the article, I still cannot dismiss the impression that this move WILL actually wound ZigBee.
Coming to mesh-network security: any reason why ZigBees mesh network would get less issues than mesh-BTLE ?
@Nando Basile you are right, there is the range anxiety issue with BLE/BTSmart. I think this debate of BLE vs. ZigBee is not relevant, each has its place and application when it comes to range and number of nodes as you rightly pointed out. It is indeed BTSmart that is giving a platform to compete with ZigBee in applications with a range overlap, upto ~10m.
ZigBee is better suited for smaller data rates and fewer data types. In its typical applications in industrial automation, sensor networks, the data is typically few 10's of kb and is most often generated from control and or operating instructions. BT on the otherhand has higher varieties of data, audio, video, etc and requires higher data rates of transmission.
One of the advantages where ZigBee stands out is in mesh network configuration with a large number of nodes that require self-healing and resiliency.
My apology for being late here but I do have a question on the "IoT race".
Where is IoT device in here - a smartphone? What else has the IP address - the very definition of an IoT node.
At the very end of the article the 6LoWPAN is mentioned... We are bombarded with these forecasts for 23B to 50B IoT devices by 2020... Soon we will enter 2015 -- where are all these IoT devices / nodes?
BT mesh is solving the range issue, but only partially(at least for TX from the phone, for RX i assume the device will send the info when you pass near it), at least until a home has enough mesh nodes.That's a negative, but also a bit of a complex thing for customers.
On the positive side, it gives incetive for people to buy more nodes, which retailers would like - and might get behind it , which is really valuable.