Is there a tool to simulate hundreds of BLE nodes advertising at the same time?
Is there a tool to simulate hundreds of BLE nodes advertising at the same time?
I can see some technical limitations to "hundreds", as the minimum time to send an advertisement and go back to sleep with the beacon example is around 5mS (I got this data by running the SDK 12.2.0 beacon example and looking at the power usage using the Power Profiler Kit ( as I just happen to be looking at this on my own project)
As the beacon has an advertising period of 100mS, the max number of beacons you could simulate by setting the advertising period to zero (0) would be around 20.
In addition, you'd need to change the Mac and other data in between each transmission, which would slow things down further.
So I think its going to be impractical for you to simulate a large number of beacons in this way
In addition to that, BLE uses frequency hopping and you could not truly simulate multiple beacons transmitting simultaneously on multiple channels with just one device.
I guess, technically, it may be possible to build or buy a Software Defined Radio (SDR) transmitter which could simulate this, but I think that would be an extremely specialist piece of equipment.
Hi eBLE and Roger,
I was using several (20) nRF51 based off-the-shelf beacons (either Nordic's beacon kit or other Chinese brands from Aliexpress and similar eshops) with slightly customized BLE FW. With that you could run e.g. 20ms ADV interval simulating 5 beacons with 100ms interval (altering both ADV_IND data and MAC address). This gives you battery of 20x5=100 beacons with 100ms interval. Definitely not 500 but still it gives you "traffic" with ~1 ADV_IND packet per 1ms in average. Good test for collisions and Rx resolution on scanner (FYI there is almost 0% packet loss in this scenario when you use nRF51/52 as scanner, you would need to go to several hundreds of beacons at 100ms interval or lower the interval to 50/20ms to get some significant collision rate).
So n the nutshell with pretty low investment (1 nRF51 beacon can start at 10~15USD) and some FW effort you can simulate pretty interesting load of traffic on BLE ADV channels 37/38/39. Other options would be to develop some BLE FW without SD which fires packets directly through radio peripheral (and so it could go down below 1ms timing resolution between Tx packets so effectively simulate 100~200 of broadcasters with 100ms interval) or to purchase some utterly expensive 2.4GHz network analyzers/synthesizers/test boxes which typically cost thousands or rather dozens of thousand USD.
Cheers Jan
Thanks Jan
My vote would be for a load of nRF51822 boards.
I suspect to buy a professional 2.4Ghz synthesiser you would need to spend tens of thousands of dollars to replicate what you can do with 50 real boards.
I totally agree. There is question if to put more effort into custom FW development (probably starting from DTM and "radio test" examples in the SDK - it is also question of skills indeed;) which could emulate dozens or even hundreds of advertisers on single nRF5x board or if to go with just small modification of generic beacon/advertiser demo app on top of standard BLE stack and then you need to buy, flash and configure probably 10-100 such boards to get to high load of 1 adv. packet per 1ms or more. Either way it will be always much cheaper then other "professional" solutions... and from my real experience I can say it works and it is worth the time and funds saving (I used 10 nRF51 boards in maximum).
You can buy nRF51 modules as cheap as about 2.7$ (ex. on aliexpress or ebay). So, it is cost about 270$ to buy 100 modules + cost of batteries/power supply + programming them. It will be cheaper and more natural than simulated tool.
