project-memoria-detector

The project-memoria-detector tool aims to determine whether a target network device runs a specific embedded TCP/IP stack.

Currently, the tool supports fingerprints for 16 embedded TCP/IP stacks (and their variations):

Several of the above stacks were found vulnerable during the Project Memoria research.

How does it work?

The script identifies the use of the 16 TCP/IP stacks on a target device via the following active fingerprinting methods:

• ICMP probing: the script performs a malformed ICMP echo request and checks for characteristics of the reply, including changes in the Time-to-live (TTL) value and specific payload content, which varies per stack.
• TCP options signatures: the script sends a TCP SYN packet and monitors the TCP SYN ACK response for the format of the TCP options field. Each stack replies with different values for the options, such as a Maximum Segment Size (MSS) and window scale.
• TCP Urgent flag handling: the script sends a TCP packet with the Urgent flag set and monitors the response. Each stack replies with a different set of TCP flags and a different TCP Window size value.
• HTTP banners and error messages: the script performs HTTP requests to a webserver hosted on a device under the test and checks for specific HTTP headers and application-specific error messages.
• SSH banners and error messages: the script performs requests to an SSH server hosted on a device under the test and checks for specific SSH banners.
• FTP banners and error messages: the script performs requests to an FTP server hosted on a device under the test and checks for specific FTP banners.

Although the script has been tested in a lab environment, we cannot guarantee its use to be safe against every possible device. Malformed ICMP packets, for instance, could crash a device that is running a different stack. THEREFORE, WE DO NOT RECOMMEND ITS USE DIRECTLY ON LIVE ENVIRONMENTS WITH MISSION-CRITICAL DEVICES (SUCH AS HOSPITALS WITH PATIENT-CONNECTED DEVICES OR SAFETY-CRITICAL INDUSTRIAL CONTROL SYSTEMS). An ideal approach is to test devices in a lab setting or during a maintenance window.

Dependencies

Please note that the tool has been tested only in a Linux test environment (any modern Linux distribution should work). It is assumed that the test environment is using iptables (https://linux.die.net/man/8/iptables) as a basic firewall.

The tool requires a recent version of Python 3.x (see https://www.python.org/, version >= 3.7.7 is preferable), as well as Scapy (see https://scapy.net/, version >= 2.4.3 is preferable).

To install Scapy together with Python and for general troubleshooting, please follow the documentation here: https://scapy.readthedocs.io/en/latest/installation.html

If python is already installed on your machine, to install the required dependencies you can use pip and the provided requirements.txt file:

$ pip install -r requirements.txt

Usage

As Scapy requires root privileges to run, the tool must be run with sudo, e.g.:

$ sudo -E python project-memoria-detector.py [options]

To see the available options for running the script, run it with the -h option:

$ python project-memoria-detector.py -h

In general, the script requires at least the following three options: (1) the IP address of the target device (-ip_dst), an open TCP port (-p or --tcp-port), and the target network interface (-i). For example:

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 -i eth1

The script uses default values for ports to fingerprint different protocols such as SSH (22), FTP (21) and HTTP (80). In order to overwrite these default values, you are able to do so by setting the flags like:

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 --http-port 80 --ssh-port 22 --ftp-port 21 -i eth1

Interpreting the results

By default, the script will output only the name of a matched TCP/IP stack and the level of certainty of the match (high, medium, low). In this case, the tool will assess all available fingerprint matches and make a decision about which stack it is, based on the fingerprint that has the highest level of confidence. For example:

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 -i eth1
    Host 192.168.212.42 runs uIP/Contiki TCP/IP stack (High level of confidence)

However, if you wish to see the matches for individual fingerprints (ICMP, TCP, and HTTP), add the -v flag (or --verbose). This will produce the output like this:

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 --http-port 80 -i eth1 -v
Host IP: 192.168.212.42
        ICMP fingerprint => uIP/Contiki (High level of confidence)
        TCP fingerprint => uIP/Contiki (Medium level of confidence)
        HTTP fingerprint => uIP/Contiki (High level of confidence)
        SSH fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        FTP fingerprint => failed to determine the TCP/IP stack (reason: No reply)

In cases when some of the fingerprints do not match, the tool produces an output similar to the following (in this example the target device did not match any of the TCP and HTTP fingerprints, but there was a match with the ICMP fingerprints):

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 -i eth1
    Host 192.168.212.42 runs uIP/Contiki TCP/IP stack (High level of confidence)

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 -i eth1 -v
Host IP: 192.168.212.42
        ICMP fingerprint => uIP/Contiki (High level of confidence)
        TCP fingerprint => failed to determine the TCP/IP stack (reason: No match)
        HTTP fingerprint => failed to determine the TCP/IP stack (reason: No match)
        SSH fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        FTP fingerprint => failed to determine the TCP/IP stack (reason: No reply)

In case there is no reply to ICMP and/or TCP messages (e.g., the device is offline, and/or it does not respond to ICMP echo requests, and/or there are no open TCP ports), the output may look like this:

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 -i eth1
    Failed to determine the TCP/IP stack for host 192.168.212.42 (reason: No reply)

$ sudo -E python project-memoria-detector.py -ip_dst 192.168.212.42 -p 80 --http-port 80 -i eth1 -v
Host IP: 192.168.212.42
        ICMP fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        TCP fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        HTTP fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        SSH fingerprint => failed to determine the TCP/IP stack (reason: No reply)
        FTP fingerprint => failed to determine the TCP/IP stack (reason: No reply)

License

Copyright (C) 2020 Forescout Technologies, Inc.

Program License

"The Program" refers to any copyrightable work licensed under this License. Each licensee is addressed as "you."

All rights granted under this License are granted for the term of copyright on the Program, and are irrevocable provided the stated conditions are met. This License explicitly affirms your unlimited permission to run the unmodified Program for personal, governmental, business or non-profit use. You are prohibited from using the Program in derivative works for commercial purposes. You are prohibited from modifying the Program to be used in a commercial product or service, either alone or in conjunction with other code, either downloadable or accessed as a service. "Derivative works" shall mean any work, whether in source or object form, that is based on (or derived from) the Program and for which the editorial revisions, annotations, elaborations, or other modifications represent, as a whole, an original work of authorship.

You may convey verbatim copies of the Program's source code as you receive it, in any medium, provided that you conspicuously and appropriately publish on each copy an appropriate copyright notice; keep intact all notices stating that this License applies to the code; keep intact all notices of the absence of any warranty; give all recipients a copy of this License along with the Program; and do not financially benefit from the sale or other conveyance of the Program either alone or in conjunction with other code, downloaded or accessed as a service.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Program and reproducing the content of the copyright notice.