I'm doing some work for my theses in networking, and have stumbled into a little problem. One of the first steps in the work I must do consists on having a computer working as an AP (I am using hostapd for this) and with it, detect all the devices in the room which currently have wifi turned on (do not need to be associated with any AP).
I have found a thread that pretty much asks the same ( discover mobil devices using wifi ), and I understand the answers that were given, but they don't give any hint as to how this can be done. The post ends saying that the person was able to do this using Kismet, however I can only seem to use Kismet to discover clients already associated with an AP.
Can someone point me in the right direction here please? If not using Kismet, then maybe suggest a different tool that works with Ubuntu.
Ps. I will need to run a continuous scan of the "room" to find any new devices and then send this information to an event manager written in Java.
I guess you could have a WiFi card, in monitor mode, scanning every channel for beacons. On Linux, aircrack-ng is the tool suite you are looking for. airodump-ng is the tool that shows you a list of devices present around your location. It is designed to display first the hotspots with the potentials clients, but also shows all the devices that are connected to an AP or trying to probe to an AP.
However, you won't be able to scan devices having their WiFi connection turned down. I'm not sure about devices not associated to an AP, my guess is you will be able to detect them if they send beacons one way or another (for example, to detect WiFi hotspots).
If you need this in Java, you can write a wrapper to airodump-ng, or you can launch airodump-ng as a service outputing to a file and read this file from a Java app.
No concrete answer I'm afraid, but I hope these will help you figure a way to solve your problem.
Related
I have a rc drone. I want to create an Android app to control it but I do not know where to begin and how. I browsed the internet and found only one useful tutorial but it required use of Adruino and was through Bluetooth but I want to create my own Android app and not use Adruino or Raspberry Pi. Please help
First you need to connect via bluetooth (If you have no documentation of your RC drone it will be had to figure out the commands etc).
Maybe you can find a bluetooth sniffer program on the www to record what is going on when using the original hardware.
But I think its not worth the effort. To figure all this out will most likely take a lot of time and might not lead to success at all. So I recommend: next time by a drone where the communication protocol is documented or better a SDK is provided with it.
First, you have to figure out, if this drone uses Wifi or Bluetooth for control, or, some custom- build wireless device. If it use custom adapter, working on another frequency, you won't be able to do anything from software, only reverse- engineer control schematics.
I am working on an android application and I want to make this app be able to tell me how many devices are connected to the current access point I am attached to. What is the best possible solution for this. I don't want the details of devices instead I just need the number of devices connected to a particular access point. I want to use this data for my web GUI which will show the utilization trend for each access point. Please suggest the way how to do this.
I'm writing an Android app that I would like to eventually port to iOS and Windows Mobile (though I know absolutely nothing about them right now). I'd like my app to be able to find other phones within a certain radius (maybe 20-30 feet) that also have the game installed so that the users can interact with others in their current physical space. Is this possible and, if so, how?
I plan on the app requiring an internet connection so my first thought was to use GPS like described at iOS Find Other Nearby Devices (GPS locations), but that would require continually transmitting data to the server to update the user's location using both bandwidth and battery (not to mention the server resources). GPS also seems to take a while to get any sort of accurate fix, and even though I wouldn't store historical data I wouldn't expect users to trust me with that sort of private information (as well they shouldn't!)
I know there are proprietary technologies, such as Sony's NFC, but that's far too narrow a focus to use for an app that I'd like to be device and OS agnostic. What other options are there? Assuming the user is willing to give the app the appropriate permissions, what other methods could allow one device to find another nearby one, within a 10-15 meter range, that is practical for a large (million+) userbase? Is HTML5 Geolocation a better path? Or is GPS my only real option?
I did find Use Android GPS to detect and connect with other phones, but was hoping the past few years have opened up further options.
UPDATE
Thank you for all the responses so far. To answer some questions posed in the comments and answers, I thought I'd describe how I expect the apps/devices to interact.
All users who install this app will have an agent running in the background to update their current geolocation and to listen for pings. Any user may open the full program and send a quick request for all other users nearby. If I go the central server route, then the server responds to that request with a list of nearby usernames. The user can then pick one of those to initiate a connection, in which case the target device makes a sound to notify of a request to connect.
From there it's easy enough to handle the interaction between two devices. Finding those nearby users is the sticking point for me.
Options I can see
Bluetooth (BT). I don't know, if I'm not too skeptical about it, but I see couple of problems. Short range. The biggest energy consumption. Maximum 7 peers connected at the same time (but maybe it's enough...). The big advantage is that a lot of devices support BT.
BT LE (low energy). Ok, it helps energy consumption problem. But, requires Android 4.3.
WiFi Direct (aka P2P). I gave it a shot, I'm a bit disappointed. The problem is that it needs Android 4.0 or higher. Even worse, some devices don't have required hardware. I launched WiFiDirectDemo from Android samples on Samsung Xperia Miro ST23i. Clicked on "search" and the only thing I got was a toast "enable WiFi P2P - in your settings". I clicked on everything in that settings, there is just normal WiFi and other stuff, no mention about Direct or P2P. Xperio Miro just doesn't support WiFi Direct.
WiFi connection with a server through internet + device geolocation. Server keeps positions of users and answers questions like "who is 20 feet in my range?". The problem can be the precision of geolocation. If the walls are thick or you are in a place on earth with few satellites - it can be very inaccurate.
To sum up
There is greater probability that somebody will have WiFi turned on all the time than BT. BT also have some time limits after which it is turned off (they make it this way to probably save battery). So, WiFi seems a better choice than BT.
If WiFi Direct is popular enough for you - I would give it a try. If not, option number 4 seems the best. It's not ideal, I know, but at this point it's hard to come with something better (we still need technology development, WiFi Direct looks promising).
I would use the central server approach. Each phone sends position data to the server. The server notifies the other users when they are in range. To limit used data, only send positions when the device reports movement more then X meters.
Also, consider using a Ping button. The user presses it to send the position when they wish to find someone nearby. Position data would only be sent on this ping. This would use very little power and provide good privacy.
It might be possible to detect details about the connected Wifi access point. If you could get the AP or Gateway's MAC address, you could send that as a kind of position. All the players on the same Gateway would be close by.
I'm writing an application that needs access to a serial port (it's a credit card reader). I'd like to be able to read and write to the port. The following seems like it would let me do exactly that: http://www.activexperts.com/serial-port-component/howto/html/. However, this deals with ActiveX/COM which means I'm tied to Windows.
Any better solutions for me? Would be nice if the solution would work on multiple platforms, not just Windows.
I'm using PHP.
Does the device emulate a keyboard? You'd be much better off getting the user to position the cursor inside a textbox (or putting the cursor there programmatically) and swiping the card if that's the case.
I spent a lot of time writing a userspace driver in C for a Magtek Mini-Mag card reader only to realize years later that it was academic more than anything. I mean we used the code, only if we'd have gone with the keyboard solution there would have been tons of time to focus on other things.
For the Mini-Mag the USB/HID support in the kernel would recognize the device, hopefully there's something similar for the device you're getting.
Also, Windows has it's own variant of USB/HID so the 'treat it as a keyboard' solution works pretty much out of the box for devices that act that way.
I have an application (however I'm testing using the sample bluetooth chat from the SDK) where two android devices connects each other, and exchange data.
I already have part of the pairing process hardcoded using the BT address, problem is: the dialog it appears to confirm the pin. Is there a way to make that work programmatically? The only thing that "appeared" to solve my problem was in this question, but that API is completely abandoned, has no documentation and many issues.
As fair as I know the problem is that if the device is not the device database with paired devices, without root access the standard API is always going to prompt the user, but I'm still wondering if there's a way.
Bluetooth autopairing is defined only for some devices. To change this you should rewrite Android framework. So, from the application, it seems to me, this is impossible to do.