Sudden drop in WiFi numbers
TL,DR: Title says it. Below is the long version of saying the bug is either at the central database end, or the app is making the data unreadable by the database after a certain number of scans.
Hello. Been wardriving for a couple years. I was looking forward to getting into the Top 100 club soon, but ran into an issue. Sometime around spring/early summer I noticed a sharp drop in the amount of WiFi signals I was receiving credit for.
As far as I can tell everything is working on my end. The number of Bluetooth signals I have been racking up indicated that I am continuing to hit new areas. The proportion of WiFi vs Bluetooth signals I'm picking up is roughly the same as always. Checked some random csv files and the WiFi signals all have coordinates.
So either there's a new guy running around Chiefs Kingdom with only WiFi enabled and I just happen to be driving through the same random unmapped hoods they are right after them (unlikely), or there's a bug somewhere in the process that is misreading the data
I ran some tests, occasionally using up to 3 phones at a time (you should be able to tell when because the start/upload times will be right after each other). This issue persists regardless of device or app version. Persisted during a recent trip where I drove from KC to south of Houston (and I drove through some ghetto areas in DFW and Houston metros I know nobody wardrives).
I made sure route logging was enabled, though it seems to only show 10%-20% of where I actually drove with Wigle running.
That's about it
Device list:
Samsung Galaxy S-5, rooted, running (I think) LOS 15/Nougat 7.1, Configured for Nethunter. No longer using due to damaged sd card slot.
Samsung Galaxy A-12, NOT rooted, running Android 12 stock
Samsung Galaxy Note 20 Ultra, NOT rooted, running Android 13 stock
Moto G7 Play, rooted, running Android 10, Configured for Nethunter
Moto G Stylus (2021), rooted, running Android 11, configured for Nethunter
Cricket Dream, rooted, running Android 11, configured for Nethunter
I was running different versions of the app without realizing it. Wasn't aware the version in the Nethunter store was different that the F-Droid version which was different than the Play store version. The Moto G Stylus and Note 20 I know are using the latest Play store version.
Hello. Been wardriving for a couple years. I was looking forward to getting into the Top 100 club soon, but ran into an issue. Sometime around spring/early summer I noticed a sharp drop in the amount of WiFi signals I was receiving credit for.
As far as I can tell everything is working on my end. The number of Bluetooth signals I have been racking up indicated that I am continuing to hit new areas. The proportion of WiFi vs Bluetooth signals I'm picking up is roughly the same as always. Checked some random csv files and the WiFi signals all have coordinates.
So either there's a new guy running around Chiefs Kingdom with only WiFi enabled and I just happen to be driving through the same random unmapped hoods they are right after them (unlikely), or there's a bug somewhere in the process that is misreading the data
I ran some tests, occasionally using up to 3 phones at a time (you should be able to tell when because the start/upload times will be right after each other). This issue persists regardless of device or app version. Persisted during a recent trip where I drove from KC to south of Houston (and I drove through some ghetto areas in DFW and Houston metros I know nobody wardrives).
I made sure route logging was enabled, though it seems to only show 10%-20% of where I actually drove with Wigle running.
That's about it
Device list:
Samsung Galaxy S-5, rooted, running (I think) LOS 15/Nougat 7.1, Configured for Nethunter. No longer using due to damaged sd card slot.
Samsung Galaxy A-12, NOT rooted, running Android 12 stock
Samsung Galaxy Note 20 Ultra, NOT rooted, running Android 13 stock
Moto G7 Play, rooted, running Android 10, Configured for Nethunter
Moto G Stylus (2021), rooted, running Android 11, configured for Nethunter
Cricket Dream, rooted, running Android 11, configured for Nethunter
I was running different versions of the app without realizing it. Wasn't aware the version in the Nethunter store was different that the F-Droid version which was different than the Play store version. The Moto G Stylus and Note 20 I know are using the latest Play store version.
First, congrats on the amazing coverage!
Now for the bad news:
If the problem were on our end, we'd see a generalized drop-off - as you can see from the stats graphs (https://wigle.net/graph-large.html) , that hasn't happened on a global scale. This suggests that the problem is local to your area or equipment.
Since you're running lots of phones, it's unlikely to be related to a specific OS upgrade/etc. It IS worth noting that lots of folks use other stumbling packages - many people are cruising around with serious rigs based on Kismet, which is likely only configured to scan WiFi - so don't rule other drivers out based on that. Using bluetooth as a barometer of new-coverage isn't necessarily reliable for a few reasons:
1. bluetooth LE uses MAC address randomization extensively - you're not necessarily seeing new devices; it may be the same devices showing new addresses
2. bluetooth is often mobile - you can drive past an empty lot one week (and see nothing) and then drive past it the next week when the state fair is there and see a bazillion BT nets. (exaggerated example, but you get the idea. Applies to schools, commute-intensive areas, etc)
If you're seeing an unusually higher bluetooth to WiFi ratio in an area (according to your device) that usually means you've already covered it. If you're seeing more bluetooth to less WiFi after uploading (via the site uploads page but NOT on your device) that usually means SOMEONE ELSE covered it.
The easiest way to check these hypotheses is through the maps, both on the site and on in the app - you can select "only mine" or "only others" to get a sense of who's been where, and you can bound the time by year. If anyone has already covered the area recently, you're likely to get a much lower yield. Stumblers in major metros often set the "most-recent bound" downward to find areas that have been covered, but not recently.
Now for the bad news:
If the problem were on our end, we'd see a generalized drop-off - as you can see from the stats graphs (https://wigle.net/graph-large.html) , that hasn't happened on a global scale. This suggests that the problem is local to your area or equipment.
Since you're running lots of phones, it's unlikely to be related to a specific OS upgrade/etc. It IS worth noting that lots of folks use other stumbling packages - many people are cruising around with serious rigs based on Kismet, which is likely only configured to scan WiFi - so don't rule other drivers out based on that. Using bluetooth as a barometer of new-coverage isn't necessarily reliable for a few reasons:
1. bluetooth LE uses MAC address randomization extensively - you're not necessarily seeing new devices; it may be the same devices showing new addresses
2. bluetooth is often mobile - you can drive past an empty lot one week (and see nothing) and then drive past it the next week when the state fair is there and see a bazillion BT nets. (exaggerated example, but you get the idea. Applies to schools, commute-intensive areas, etc)
If you're seeing an unusually higher bluetooth to WiFi ratio in an area (according to your device) that usually means you've already covered it. If you're seeing more bluetooth to less WiFi after uploading (via the site uploads page but NOT on your device) that usually means SOMEONE ELSE covered it.
The easiest way to check these hypotheses is through the maps, both on the site and on in the app - you can select "only mine" or "only others" to get a sense of who's been where, and you can bound the time by year. If anyone has already covered the area recently, you're likely to get a much lower yield. Stumblers in major metros often set the "most-recent bound" downward to find areas that have been covered, but not recently.
A couple of other notes:
- You didn't mention the scan throttling setting - have you made sure to turn OFF scan throttling (via Developer mode) on your Android 10-and-up devices?
- We don't support or maintain whatever that Nethunter release is, and it was done without our knowledge or blessing. That's 3 years old, and should be treated as suspect.
- You didn't mention the scan throttling setting - have you made sure to turn OFF scan throttling (via Developer mode) on your Android 10-and-up devices?
- We don't support or maintain whatever that Nethunter release is, and it was done without our knowledge or blessing. That's 3 years old, and should be treated as suspect.
Return to “WiGLE WiFi Wardriving Bugs”
Who is online
Users browsing this forum: No registered users and 60 guests