Range/power: IEEE802.15.4 (thread) vs. Bluetooth 5

[Reto]

Dear experts,

Comparing:

• IEEE802.15.4, Thread protocol, 250kBit/sec,
• Bluetooth 5, Coded PHY (longest range), 125kBit/sec

Questions:

• Which technology has longer range?
• Which technology has better range/power ratio?
• What current consumption can I expect of nRF52840 when Rx/Tx 802.15.4 thread protocols?

Thanks!

[xeem]

I am also extremely curious what the range and power differences are between thread and BLE 5!

[ovrebekk]

Hi Reto

The RX sensitivity in 125kbps mode is 3dB better than in 250kbps 15.4 mode, so you should expect up to 50% longer range under ideal conditions (if there is a lot of interference then the advantage of improved RX sensitivity is diminished, and TX output power is more important).

The peak current in TX and RX mode is more or less identical, regardless of the radio mode used, and depends mostly on supply voltage, whether or not DCDC is enabled, and on the output power in TX mode. This means that the average current (and power) is determined primarily by the length of the TX packets and RX windows.

For mesh networking I don't expect dramatic differences in average current between 15.4 or BLE based solutions, as both require a high level of radio activity and will not be very low power. Current BLE mesh solutions are based on the 1Mbps bitrate mode, giving 15.4 the edge in range, but in the future there might be BLE mesh solutions based on the 125kbps mode, giving a slight advantage to BLE when it comes to range. Both BLE mesh and Thread allow the implementation of lower power end nodes, that don't have the capability to do meshing, but it is still too early to say which of these implementations will be lowest in terms of average current.

If you only need peer to peer connections then BLE is considerably better, since the BLE protocol is based on highly accurate synchronization between the TX and RX (making the RX windows short and the average current low). BLE also has the advantage that you can scale between 125k, 500k, 1M and 2M bitrate modes, allowing you to reduce the average current when you don't need as long range. The peak currents are similar at higher bitrates, but since the packets are sent much faster over the air the average currents are lower.

TL;DR: For mesh networking I think Thread and BLE is very comparable in terms of range and power requirements, and I would look at the other aspects of Thread vs BLE to decide which protocol to use.

For P2P or star networks BLE has a considerable advantage, giving you lower average currents and multiple bitrate options, and I would recommend BLE over Thread.

Best regards
Torbjørn

[Reto]

What a sophisticated answer! Thanks a lot!

[xeem]

Agreed! Great answer!

Is there an average current measurement for BLE mesh floating around anywhere? Obviously, it's use dependant but a general ballpark would still be really helpful.

[ovrebekk]

In the current BLE mesh implementation the radio is always on, in either TX or RX, so the average current is high ;)
With the nRF52832 running at 3V and using DCDC you are looking at around 6mA average when the radio is on continuously and you have some CPU processing on the side.