You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Martin Florian edited this page Apr 8, 2017
·
2 revisions
Evaluation: Tor traffic
Some additional experiments have been done to check how much traffic Tor generates when idling.
Setup
Android smartphone (Nexus 5X): Doze disabled for PTP (Settings -> Battery -> Battery Optimization -> PTPEvaluationApp -> Not optimized), screen was off during evaluation, started with fully charged battery
PTP was configured to start Tor and create a hidden service but don't send any message.
PTP uses Tor binary from Orbot (0.2.8.9)
Orbot Version 15.2.0-RC-8-multi (uses Tor 0.2.8.9)
No apps have been selected for transparent proxying
date
runtime
setup
bytes total
average bytes
02.02.17
06h16m
PTP
5,43 MB
0,87 MB/h
09.02.17
09h15m
Orbot without HS
6,04 MB
0,65 MB/h
15.02.17
9h16m
Orbot with HS
13,88 MB
1,50 MB/h
18.02.17
6h37m
Orbot with HS
9,40 MB
1,42 MB/h
18.02.17
3h50m
PTP
4,93 MB
1,29 MB/h
19.02.17
3h20m
PTP
2,12 MB
0,64 MB/h
02.03.17
03h11m
PTP
5,35 MB
1,68 MB/h
Conclusion
Tor causes a lot of network traffic to provide a hidden service, even when idling (about 1-1.5 MB/h).
The experiments showed a big variation in the results (between several runs and between PTP and Orbot).
The cause of the variation still needs to be investigated.