This week, I collected more experimental data for Udit and Rhea to use in tuning the multilateration algorithm. I implemented a time-sync algorithm in the Decawave chips that uses a master chip which periodically broadcast its own timestamp, which the slave anchors adjust their own clocks in response to, before transmitting pulse reception timestamps to the Raspberry Pi.
![](http://course.ece.cmu.edu/~ece500/projects/s20-teamb6/wp-content/uploads/sites/88/2020/04/93302056_220170455915865_1571114262535864320_n-300x250.png)
I also assembled an experimental setup to test the algorithm, setting up the tags and anchors in a 1×1 meter square. I set up two slave anchors that broadcast the time of pulse reception to the Raspberry Pi server, a masterĀ that only broadcasts a timestamp, and a stationary tag that broadcasts its ID. The setup was meant to keep both anchors 1 meter away from both the tag and the master simultaneously.
![](http://course.ece.cmu.edu/~ece500/projects/s20-teamb6/wp-content/uploads/sites/88/2020/04/IMG_0967-300x225.jpg)
Our progress is on schedule for a demo this Wednesday. Udit and Rhea have been making substantial progress towards an effective multilateration algorithm with imperfect data, and I have established a consistent workflow for graphing and collecting timestamp data from multiple anchors and multiple tags simultaneously.
![](http://course.ece.cmu.edu/~ece500/projects/s20-teamb6/wp-content/uploads/sites/88/2020/04/delta-diffs-100cm-NO_outliers-300x167.png)
![](http://course.ece.cmu.edu/~ece500/projects/s20-teamb6/wp-content/uploads/sites/88/2020/04/direct-diffs-100cm-NO_outliers-300x167.png)
This week I will perform any final tests that Udit and Rhea need to fine-tune the multilateration algorithm, and then we will demo on Wednesday.