$Id: README,v 1.4 2006/11/21 21:11:07 rbeverly Exp $

IP Spoofing Tester v0.5 (http://spoofer.csail.mit.edu)
-------------------------------------------
1. Running a binary

This is the preferred method of running the spoofer.  
  a. On Windows, please download and run the installer.

  b. On *unix systems, uncompress and untar with:
      $ gzip -dc spoofer-xxx-0.5.tar | tar -xvf -
     You must run the spoofer as root (in order to create 
     the raw socket) with no arguments, e.g.
       # ./spoofer


2. Compiling

Makefiles for different operating systems are included.  Copy the
appropriate Makefile over and run "make" (assumes sane, gnu-like compiler
and make).


3. How it works

Our methodology is simple. We are making software to test spoofing
publicly available and asking the community to run it from as many sites
as possible.

The spoofer program attempts to send a series of spoofed UDP packets to a
server on our campus. These may not reach the server if a filter or other
policy blocks the packets. Alternatively, the packets may arrive with the
true source address instead of the spoofed address if a network address
translation (NAT) device rewrote the header. After sending the spoofed UDP
packets, the program establishes a TCP connection with the server
informing it of the spoofed packets it tried to send. Based on this
transaction, the server records the success or failure of this machine's
ability to send spoofed packets in a database. With sufficient network
coverage, we can better understand the current state of filtering in the
Internet.


4. Background

The Internet architecture provides no explicit mechanism to prevent
packets with forged headers from traversing the network.  By "spoofing"
the source address of an IP packet, a malicious user can send packets
toward a victim anonymously.

Techniques such as ingress and egress address filtering (RFC2267, RFC1918)
and unicast reverse path forwarding (uRPF) are used to prevent spoofing,
but these are typically useful only at the edge of the network and are
often sporadically applied. While attacks emanating from zombie farms
often don't bother spoofing their source address, spoofing is still
prevalent on the Internet from analysis of backscatter.

Previous work (Snoeren, et al.) has suggested efficient means to trace
spoofed packets back to their origin, but today finding the source of
spoofed packets remains an operationally difficult problem. This project
seeks to determine the extent to which spoofing is currently possible on
the Internet.


5. Frequency of testing

The server which the spoofer client interacts with only permits
up to three sessions within the last seven days from any single
client.  We limit the number of tests to prevent abuse on our
servers and remote networks.  However, periodically running our 
client, even from the same point in the network, gives us additional
data.  For instance, we are very interested in how ISP policies
change over time.  If you wish to setup the spoofer client as
a cron job, please schedule it to run monthly.


6. Feedback

Written by Rob Beverly <rbeverly at mit dot edu>.  Feedback, flames and
bugfixes welcome.

