We are developing integrated solutions for information security. Part of our work is focused on low-cost secure key establishment for IoT applications. Any security application uses keys in order to protect or construct security primitives. These keys are generated in a random way, thus, we are researching about fully-synthesized true random number generators (TRNG) for generating high quality keys with a robust performance against process and environmental variations. | Finally, we are interested in state of the art problems such as the vulnerability of DRAM information against row hammer attacks and we are studying a strategy to mitigate this vulnerability. We developed a first core prototype to implement key establishment. The algorithm is based on neural cryptography that allows us to perform a fast key sharing. The prototype has a re-keying characteristic enabled in order to update faster and easier the final shared key. We also are working in developing a fully-synthesized TRNG. We have a first prototype to understand and probe the concept of fully-synthesized entropy sources. One of the major problems of the TRNGs in general is the poor statistical qualities of the randomness sources. Ideal sources can not be constructed, therefore the final data is not really uniform. Hence, we are working in a better entropy source using two independent sources based on a mathematical breakthrough developed by the professor David Zuckerman at UT Austin.

alt text