According to wikipedia, a computer trojan is a malicious piece of software, which is used to hack into a computer by misleading a user of its true intent. Hardware trojans are no different in this regard; they are malicious circuits instead of programs, capable of causing much damage or leaking lots of information, if they manage to avoid detection. In this project, we will focus on a new type of trojan: trojans in hardware modules (IP cores) of reconfigurable computers.
Reconfigurable computers are malleable hardware, whose function may be changed repeatedly at runtime, or put in another way, it is hardware that is reprogramable like software. One example of such reconfigurable hardware devices are FPGAs (field programmable gate arrays). FPGAs can load configuration files to reconfigure their hardware functions, much like processors can load programs. With todays complex devices, creating a complete FPGA configuration from scratch is a daunting task, which is usually too costly to perform, so that nowadays FPGA designs usually are created using existing hardware modules as building blocks, so called IP cores (intellectual property cores), much like you would use libraries to reduce the effort of creating a new software program. And just as a trojan infected library could undermine a sofware program, an infected IP core is a large threat that could be lurking in FPGA configurations. Detecting such trojans before they are able to deliver their malicious payload, is a very hot research topic.