fingerprinting traffic from smartphone mobile hotspot - tcp

Is it possible to detect whether certain TCP or UDP traffic is generated and processed directly by an Android mobile phone itself, or is being tethered to a different machine?
With HTTP, detection is relatively simple -- just look at the User-Agent string in the requests, and there is a certain likelihood to see some sort of a revealing pattern.
What about the discreet protocols, like https, ssh, IPsec, OpenVPN etc? Is it possible to detect the originator OS based on the TCP/UDP headers or content?
How would one masquerade the traffic to avoid detection through the passive fingerprinting of the encrypted protocols?
For example, if the mobile phone is running Android 2.2, and is tethered to OS X or OpenBSD, which TCP/UDP and https/ssh/IPsec/OpenVPN parameters would reveal the non-Android OS?

you can do analysis based on TTL, TCP ISN etc. http://www.cse.sc.edu/~wyxu/papers/wisec66-schulz.pdf‎ contains a detailed analysis

You appear to be looking for TCP/IP stack fingerprinting software.
P0f, which stands for "passive Operating System fingerprinting," observes packets at the TCP level and makes assumptions about the OS based on the network stack implementation. P0f does this passively, meaning that it does not need to interact with the client.
The p0f site has an example with which you can check your own fingerprint. Note that many mobile web providers run through proxies, which tends to mean that the proxy is what is fingerprinted rather than the client.
See also the EFF's Panopticlick project, which shows that pretty much every web browser is uniquely fingerprintable, even when blocking cookies. Please do not abuse fingerprinting to invade users' privacy.

Related

Is there something I should be concerned about before port-forwarding my server?

I'm setting up my first server on a Raspberry Pi 4 but after reading some articles online I was wondering whether my server is ready to be open to the internet or not. I premise I'm just an individual who would like to publish some programming projects on a site that is accessible on a browser.
After some concerns I designed a PHP page which checks the client IP and returns a 403 header until i give that user the permission to access. Is it enough? Is it necessary?
And also, are there ports that are more safe to open than others?
You "can" open ports 80 and/or 443 for displaying webpages - depending on SSL certificates
I do it myself (not for web hosting) and restrict the open ports to certain IPs - my friends (not smart enough to levy an attack 😂). Though IPs are likely to change every so often and your firewall will need updating.
It's a key thing to remember that anything is open to exploitation if it's not properly maintained/set up. Also displaying a 403 isn't a silver bullet.
Port 25 would give a user access to the files on your device if proper authorisation isn't set up. Opening ports 80 and 443 will give users access to webpages but makes your device/network exposed to DoS attacks or platform level attacks. If there's a known exploit for your version of PHP or your firewall/router or possibly the device itself then an attacker will exploit it.
Hosting providers have layers upon layers of security and are constantly updating devices throughout their network. Keeping your device and platform up to date will help - but it may be worth instead investing a little in a host (from about £4 a month).
There are loads more things I can touch on but will leave it at that for now
Edit after comment:
my website is just a little project i mean who could casually target it?
Strictly speaking, anyone. "Who would want to?" Again, anyone. Sure you're a small target that wouldn't provide any useful data. But your device, once hacked, can be used as a DoS zombie or as a crypto-miner and you probably wouldn't even realise.
And also can't I use whatever port like 6969 or 45688?
Yes, strictly speaking, you can. You could tell your device to listen on that port and reply with the website data. To do this you would also need to provide the port number on the end of the URL in the format www.example.com:6969. Though, again, this isn't a silver bullet. Most security issues aren't with port-forwarding but with poor management/security and bugs in the components themselves. All a port forwarder is doing is saying "oh, device X wants data on this port... here you go".
Another point is, data sent on "Well-known ports" (1-1023) tend to have their headers checked for irregularities by the firewall - which can dispose of any irregular packets. By using a custom port the firewall doesn't really know what to expect, so it sends it anyway. Also, steer away from "Private ports" (49152-65535) these are used as source ports, not destination ports.

What options are there to "push" to internet connected devices?

I am working on a project that involves triggering actions on various internet connected devices such as phones, computers, home automation devices, etc. What technologies/general actions are available to connect devices like that? From my limited knowledge I can think of constant poling by all the devices, or making use of a prolonged HTTP connection (I believe that is somewhere along the lines of how Facebook does their notifications).
What other options do I have? What frameworks are there out there? Which option is the best when it comes to time from when the server wants to contact the device to when the device is actually contacted? Which option is easiest on the battery life of the devices?
For battery life, there's not much that beats a plain TCP socket connection. If you can do all the device's polling over a single connection, there is very little overhead besides a few bytes every x seconds for a keepalive packet (which you probably should either enable in the TCP stack or generate as a part of your protocol) This is afaik similar to what Apple uses for iOS notifications.
If you're in an environment where socket communication may be blocked by a firewall, I'd go for websockets with a fallback of long polling (which you mentioned in the question). Combined with proxy support, this should be able to traverse just about any firewall without much complication.

Why some internet providers close certain ports?

We published the game on russian server and 1% of people couldn't connect to server on 46xx port through raw TCP while they can load it's HTML page (through HTTP). Most of such people live in Germany, Israel....
Why is it so? What's the politics decisions lay behind it? We discovered that their such ports (which are free on IANA) are closed. Does it mean that such people cannot run Steam (and, then, play all games which you can buy through it), play WoW and many other modern games which use TCP through 4xxx ports?
Thank you.
ISPs have been known to filter certain ports for various reasons. Users should complain loudly to them (or switch) in order to send a signal that such is not to be tolerated. You can encourage them to do so but of course that doesn't solve your problem (or really answer your question).
Common reasons are:
- trying to block bittorrent traffic
- limit bandwidth usage (largely related to previous reason)
- security (mistaken)
- control (companies often don't want employees goofing off)
The easiest thing for you to do is run your game over port 443 (perhaps as an alternate). That's HTTPS and so will not generally be blocked. However, because HTTPS is encrypted, there's no way to inspect the stream to know if its web traffic or something else and thus you can run any data stream (encrypted or not) that you wish over it.
That's precisely correct. In fact every public web site would by default block all ports except the ones they expect to be running some traffic they would want to.
This is the reason many applications often try to encapsulate their programs to use port 80 which can't be blocked as long as some one wants http traffic to run.
They simply don't want any application that they haven't approved to run through their servers. If you have a sensitive server in public you surely won't want any one to use your machine for any apps that you don't allow. A common reason is applications that eat up bandwidth such as bittorent, edonkey, gnutella as well as streaming, voip and other high bandwidth consuming apps

How to watch network traffic? A la packet sniffing or web debugging proxy?

So when I am debugging my web applications and such, I've used the Charles web proxy and debugger and love it. It's so nice to see what's being sent and received via port 80 and 443. I can see all the resources loading, not just from the "browser" per say, but also flash applications. I can also see how the calls are being made, and it pretty easy to reconstruct them. It's a great debugging tool and I love it.
So I'm wondering two things:
First, I'm wondering is if there is something similar I can use to watch traffic that might be coming though on other ports. I guess some desktop applications will use the internet, but not necessarily via http / https requests. I remember looking at some security tools a few years ago - there are a lot of security tools out there, like kismet / etherCap, ethershark, etc - is there one that does what I'm describing in an easy and intuitive way?
Also, I'm wondering if I am using my iPhone / iPad / Android device, how can I set up a proxy through my computer so I can watch the http/https requests that the device makes?
Found the answer to that one here: http://www.ravelrumba.com/blog/ipad-http-debugging/
I'm mostly on a Mac so anything that is Mac friendly would be extra helpful.
Thanks!
I believe you are looking for Wireshark. It allows you to monitor the network interface on your machine and be able to tell you sent/receive packets as well as their protocols. It also has a protocol decoder that can be used to get Layer 7 information about a IP stream. You can also do a "Follow TCP stream" which allows you to view the entire conversation of that connection. It's based on libpcap (Packet capture) which the built in tcpdump also uses.
The only downside for you web developers is that if you're using SSL encrypted sessions, you can't decode it. The endpoints of the SSL session are "above" (using OSI model) the layer at which wireshark (and similar tools) operate.
Here's a good list http://sectools.org/sniffers.html. I used Wireshark back when it was Ethereal. At that time it ran under X11, It looks like that has changed.

What percentage of users are behind symmetric NATs, such that "p2p" traffic needs to be relayed?

We're implementing a SIP-based solution and have configured the setup to work with RTPProxy. Right now, we're routing everything through RTPProxy as we were having some issues with media transport relying on ICE. If we're not mistaken, a central relay server is necessary for relaying streaming data between two clients if they're behind symmetric NATs. In practice, is this a large percentage of all consumer users? How much bandwidth woudl we save if we implemented proper routing to skip the relay server when not necessary. Are there better solutions we're missing?
In falling order of usefulness:
There is a direct connection between the two endpoints in both directions. You just connect and you are essentially done.
There is a direct connection between the two endpoints in one direction. In that case you just connect via the right direction by trying both.
Both parties are behind NATs of some kind.
Luckily, UPnP works in one end, you can then upgrade the connection to the above scheme
UPnP doesn't work, but STUN does. Use it to punch a hole in the NAT. There are a couple of different protocols but the general trick is to negotiate via a middle man that coordinates the NAT-piercing.
You fall back to let another node on the network act as a relaying proxy.
If you implement the full list above, then you have to give up very few connections and don't have to spend much time on bandwidth utilization at proxies. The BitTorrent protocol, of which I am somewhat familiar, usually stops at UPnP, but provides a built-in test to test for connectivity through the NAT.
One really wonders why IPv6 did not get implemented earlier - this is a waste of programmers time.
Real world NAT types survey (not a huge dataset, though):
http://nattest.net.in.tum.de/results.php
According to Google, about 8% of the traffic has to be relayed: http://code.google.com/apis/talk/libjingle/important_concepts.html
A large percentage (if not the majority) of home users uses NAT, as that is what those xDSL/cable routers use to provide network access to the local network.
You can theoretically use UPnP to open ports and set-up forwarding rules on the router to go through the NAT transparently. Unfortunately (or fortunately, depending on who you are) many users disable UPnP as a matter of course on their router and may not appreciate having to add forwarding rules manually.
What you might be able to do (and what Skype does AFAIK) is to have some of the users that have clear network paths and enough bandwidth act as relay nodes. Apart from the routing and QoS issues, you would at least have to find some way to ensure the privacy of any relayed data from anyone, including the owner of the relay node. In addition, there might be legal issues to settle with this approach, apart from the technical ones.

Resources