in order to reduce false negatives:
1. have > 2 RPI opposite of the house/ apartment .. the more the better -- I have 10.
in case you sit on the beacon the signal might not come through your body.. As wifes are always slim, that should not be a problem.
2. do not set the timeout too fast
I have not seen wrong positives.
if you want to do quick downs, try " fast down mode" in the beacons devices edit, set to 20 secs. If no signal for 20 secs the beacon will go to down
as for battery live the tiles seem to be a bit better, but the new ones seem to more sophisticated to save battery power.. that might create a false negative.
the radius are ugly, but you can set the signal strength and repeat rate. for good detection ~1 second repeat is more than enough. don't need 0.2 secs.
but some others can probably chime in with more info..
Karl
and thanks for the nice feed back
new release coming:
- communication through sockets instead of
RPI -> http -> indigo web server -> log file --> sql logger for variable changes --> 'expensive' "variable has changed method" into pibeacon belly
new:
RPI --> tcp/ip socket --> tcp/ip listener --> pibeacon belly .
with message error control (I found a bad ethernet cable on one of my RPIs, 10% of messages required a retransmit). that is not visible in the HTTP channel
- added function cut off low signal if not UP. If up use signal. This should prevent remote beacons coming going UP/down frequency if they are at e.g. -90 dBm.
- some error checking and a lot of tuning
and the bad side: I changed all folder device and variable names where it had a "-" to a "_". learned that indigo does not support those fully (e.g. you can not edit their names )
that will require some maintanece if you have e.g. events setup on variables.
old device names etc will not be touched though.
the stats info will look like:
- Code: Select all
00:10:53 pi TCPIP socket -Stats for RPI-->INDIGO data transfers. Tracking started 2017-10-09 23:11:23. Report TX errors if time between errors is <:100secs
00:10:53 pi TCPIP socket -IP name type first last #MSGs #bytes bytes/MSG bytes/min MSGs/min
00:10:53 -192.168.1.20 pi_IN_0 ok 2017-10-09 23:11:48 2017-10-10 00:10:36 653 254836 390 4282.6 10.9739
00:10:53 -192.168.1.20 pi_IN_Alive ok 2017-10-09 23:17:57 2017-10-10 00:05:42 9 1061 117 17.8305 0.1512
00:10:53 -192.168.1.21 pi_IN_1 ok 2017-10-09 23:11:44 2017-10-10 00:10:41 353 191158 541 3212.5 5.9323
00:10:53 -192.168.1.21 pi_IN_Alive ok 2017-10-09 23:12:55 2017-10-10 00:04:49 6 678 113 11.3940 0.1008
00:10:53 -192.168.1.22 pi_IN_2 ok 2017-10-09 23:11:48 2017-10-10 00:10:40 294 93519 318 1571.6 4.9408
00:10:53 -192.168.1.22 pi_IN_Alive ok 2017-10-09 23:20:20 2017-10-10 00:09:48 5 595 119 9.9992 0.0840
00:10:53 -192.168.1.23 pi_IN_3 ok 2017-10-09 23:11:46 2017-10-10 00:10:50 324 233607 721 3925.9 5.4449
00:10:53 -192.168.1.23 pi_IN_Alive ok 2017-10-09 23:22:30 2017-10-10 00:06:43 5 560 112 9.4110 0.0840
00:10:53 -192.168.1.24 pi_IN_4 ok 2017-10-09 23:11:53 2017-10-10 00:10:34 140 153803 1098 2584.7 2.3528
00:10:53 -192.168.1.24 pi_IN_Alive ok 2017-10-09 23:15:23 2017-10-10 00:07:43 6 690 115 11.5957 0.1008
00:10:53 -192.168.1.25 pi_IN_5 ok 2017-10-09 23:11:44 2017-10-10 00:10:51 198 248927 1257 4183.3 3.3275
00:10:53 -192.168.1.25 pi_IN_Alive ok 2017-10-09 23:17:37 2017-10-10 00:10:39 6 678 113 11.3940 0.1008
00:10:53 -192.168.1.26 pi_IN_6 ok 2017-10-09 23:11:57 2017-10-10 00:10:38 261 163024 624 2739.7 4.3862
00:10:53 -192.168.1.26 pi_IN_Alive ok 2017-10-09 23:12:49 2017-10-10 00:04:39 6 714 119 11.9990 0.1008
00:10:53 -192.168.1.27 pi_IN_7 ok 2017-10-09 23:11:55 2017-10-10 00:10:33 35 5880 168 98.8156 0.5882
00:10:53 -192.168.1.27 pi_IN_Alive ok 2017-10-09 23:11:46 2017-10-10 00:10:19 33 3630 110 61.0035 0.5546
00:10:53 -192.168.1.28 pi_IN_8 ok 2017-10-09 23:12:11 2017-10-10 00:10:47 150 39736 264 667.7783 2.5208
00:10:53 -192.168.1.28 pi_IN_Alive ok 2017-10-09 23:14:58 2017-10-10 00:08:59 6 678 113 11.3940 0.1008
00:10:53 -192.168.1.29 pi_IN_9 ok 2017-10-09 23:11:51 2017-10-10 00:10:36 146 206897 1417 3477.0 2.4536
00:10:53 -192.168.1.29 pi_IN_9 errLength 2017-10-09 19:08:57 2017-10-09 19:51:14 9 13032 1448 297.6812 0.2056
00:10:53 -192.168.1.29 pi_IN_Alive ok 2017-10-09 23:12:49 2017-10-10 00:04:48 6 696 116 11.6965 0.1008
00:10:53 -192.168.1.31 pi_IN_11 ok 2017-10-09 23:11:45 2017-10-10 00:10:46 166 54780 330 920.5983 2.7897
00:10:53 -192.168.1.31 pi_IN_Alive ok 2017-10-09 23:12:57 2017-10-10 00:04:39 6 666 111 11.1924 0.1008
00:10:53 -192.168.1.33 pi_IN_13 ok 2017-10-09 23:11:43 2017-10-10 00:10:52 82 23759 289 399.2789 1.3780
00:10:53 -192.168.1.33 pi_IN_Alive ok 2017-10-09 23:20:37 2017-10-10 00:03:05 8 1124 140 18.8892 0.1344
00:10:53 -192.168.1.34 pi_IN_14 ok 2017-10-09 23:12:46 2017-10-10 00:09:59 42 8190 195 137.6360 0.7058
00:10:53 -192.168.1.34 pi_IN_Alive ok 2017-10-09 23:11:45 2017-10-10 00:01:25 7 826 118 13.8812 0.1176
00:10:53 -192.168.1.36 pi_IN_16 ok 2017-10-09 23:11:50 2017-10-10 00:10:40 159 36729 231 617.2445 2.6721
00:10:53 -192.168.1.36 pi_IN_Alive ok 2017-10-09 23:15:57 2017-10-10 00:09:00 6 666 111 11.1924 0.1008
00:10:53 -192.168.1.38 pi_IN_18 ok 2017-10-09 23:11:59 2017-10-10 00:09:49 38 6802 179 114.3101 0.6386
00:10:53 -192.168.1.38 pi_IN_Alive ok 2017-10-09 23:11:57 2017-10-10 00:10:13 30 3661 122 61.5245 0.5042
00:10:53 -total 3186 1738570 545 29217.3 53.5419
00:10:53 pi TCPIP socket - === Stats for RPI --> INDIGO data transfers == END total minutes measured: 59
besides looking at transmission error see errLength line,
you can also see which RPi is very busy and check why.