Hi, I was wondering how much power we could save if we buy the TTK1100 for TDoA stack. In APH011 described picture below, showing current spikes when TWR exchanges happen.
My question is, if we use TDoA, then the tag only need to send blinks, therefore instead of 6 current spikes, it will be reduced into 1 single tx spike.
Is this line of thought correct? Can anyone give me numbers so that I can ensure other engineer in my team?
Correct, for TDoA you only need that initial Tx spike, the tag can then go back to sleep. The receivers will probably need to be powered up to some level all the time in order to maintain good clock synchronisation but since they will normally be wired into the building that’s not usually an issue.
One potential issue if only powered up for a short time is how you handle splitting time between tags without being awake to be allocated a time. Since each transmit is so short for a small number of tags you can transmit at a random time and the risk of a collision is tiny. As the number increases you may need to start to do something about it.
Thanks for the additional insight to TDoA collisions! I’m pretty much interested in TDoA for it uses less power, but is there any drawbacks that I might miss?
So far, I have heard how the TDoA could have collisions when the tags are numerous, or how TDoA provides less accuracy and robustness compared to TDMA TWR.
The collisions issues exists for TWR as well, in fact it’s worse since TWR takes longer per tag. But in TWR the tag is receiving as well and so there is more opportunity to pass timeslot information to it.
With TDoA and anything under ~1000 updates per second (total over all tags) it’s probably easier to ignore the risk of collisions. Ensure that the time between transmits is slightly random (so two tags that do collide probably won’t collide on their next transmit) and the losses due to collisions will be low enough that for most applications they don’t matter.
TDoA you need one more anchor to see the message and your accuracy is dependent on getting good clock sync between anchors. The easiest way to do this is a reference clock hard wired to each anchor and a tag at a known location. Other methods exist but are more complex and/or less accurate.
TDoA you also need all the anchors to be connected to some central position calculation server so those connections and server act as a single point of failure for the whole system. TWR on the other hand is a lot more robust, unless you have a central system allocating time slices there is no single point of failure that will bring the whole system down.
TDoA is better for moving items since all measurements are simultaneous, there is no risk of motion between measurements. TWR on the other hand gives lower latency if the tag needs to know it’s location since that is where the data is collected, no additional server and network delays.
Having said that the system I created was for tracking things doing >30m/s and >1g accelerations and we went with TDMA DS TWR over TDoA. We do however measure ranges very quickly and correct for the different measurement times. Plus we only need to cope with a small known number of tags which makes life easier. The simplicity of setup (anchors can be pre-configured off site and require only power) ended up being the deciding factor, we can be up and running on site in around 30 minutes.
If I wanted to track/locate signal strengths what would be the best program to use for the job. Using a directional antenna..?
You can come up with an approximation of the received signal strength based on a number of device registers but it’s not very accurate or consistent.
What exactly are you trying to do? Normally when someone wants to monitor signal strength it’s for some reason other than curiosity as to what the signal strength is. What is your end goal?
