Installation guide

Must have been a temporary issue, I have no problems accessing it now:

http://www.lkn.ei.tum.de/personen/jan/imtaphy/install.php

Here is the info (copy & paste from the homepage):

How to install IMTAphy from launchpad.net

This assumes a 64 bit (x86-64) Linux system running a recent Ubuntu version (tested on 10.04 to 12.04alpha) and most other Linux distributions should also work. For compilation on a 32 bit system, see note below.
Building the simulator
Make sure all prerequisites needed by openWNS are installed (bzr, scons, cppunit, ). On recent Ubuntu systems do:
apt-get install g++ bzr wget libitpp-dev libcppunit-dev libboost-all-dev scons python-matplotlib
IMTAphy itself requires Intel's Math Kernel Library (MKL). Intel provides free versions for non-commercial use and trial versions on its website.
Currently, the open-source IT++ library of mathematical, signal processing and communication routines is also required. On Ubuntu systems it is available as a pre-compiled package (see above).
Please make sure that the environment variables $MKLROOT, $INCLUDE, and $LIBRARY_PATH are set appropriately before compiling IMTAphy. This is done by sourcing a script provided by MKL (see the MKL user's guide):
. /opt/intel/mkl/bin/mklvars.sh intel64
It is also a good idea to get an overview of openWNS by visiting the launchpad page at https://launchpad.net/openwns or by looking at the Developer's Guide and User's Guide
Use the bzr utility to check out the openWNS SDK from launchpad:
bzr co lp:openwns-sdk
In the newly checked-out openwns-sdk directory, replace the projects configuration openwns-sdk/config/projects.py.template by this projects.py file which you can copy to openwns-sdk/config/projects.py
~/openwns-sdk/config$ wget http://www.lkn.ei.tum.de/personen/jan/imtaphy/projects.py
Download all remaining openWNS projects by "./playground.py upgrade" inside your openwns-sdk directory:
openwns-sdk$ ./playground.py upgrade
You should now find all IMTAphy code in your openwns-sdk/modules/phy/imtaphy directory.
When all prerequisites (see above, check that $MKLROOT is set appropriately) are fulfilled, debug and release ("opt") versions of openWNS including the IMTAphy module can be built with:
openwns-sdk$ ./playground.py install && ./playground.py install --flavour=opt
Testing the build by running the unit test suite
After following the above steps, you should have an executable openwns simulator application binary in openwns-sdk/sandbox/dbg/bin/ (compiled in debug mode) and in openwns-sdk/sandbox/opt/bin/ (compiled in release/"opt" mode). openWNS loads additional functionality from dynamic libraries that can be found in openwns-sdk/sandbox/dbg/lib/ and openwns-sdk/sandbox/opt/lib/ for the debug and release version, respectively. The IMTAphy functionality is installed in one such library. To test if openWNS and IMTAphy compiled and installed correctly on your system, you can run a suite of unit tests by starting openwns in the unit test directory as:

openwns-sdk/tests/unit/unitTests$ ./openwns -tv
You should see the simulator execute a couple of pre-compiled tests, some of them are IMTAphy unit tests. After each test is completed you should see an [OK] indicator and after a while the simulator should exit gracefully with a summary of successfully completed tests.

Troubleshooting

Here we list a few common problems people have encountered when compiling and running the simulator. If you don't see your problem mentioned here, please search the Question & Answers section on https://launchpad.net/imtaphy before posting a question there.
Compilation on 32 bit systems is possible by setting compile32Bit = True in the openwns-sdk/modules/phy/imtaphy/SConscript file. In addition, the appropriate MKL environment variables have to be set, e.g., by . /opt/intel/mkl/bin/mklvars.sh ia32 . Note that 32 bit operation has /not been as extensively tested as the standard 64 bit mode. However, it seems /to be working fine.
If the built process fails with a "file not found" or similar error message, please re-check all the above requirements. Especially, make sure the environment variables needed by MKL (see above) are set in the shell from which the build process is started. For example, check the output of echo $MKLROOT.
With gcc version 4.6 and beyond the compilation of certain source files (mostly scheduler-related files) consumes a lot of time and memory when compiling in "opt"-mode. On machines with few memory this might cause the build process to fail if the memory is exhausted. When this occurs, try to resume the build process with just a single compile job at a time by building with openwns-sdk$ ./playground install --flavour=opt -j1. This issue does not occur in with older gcc versions or when compiling in "dbg" mode.