Farnaz Gharibian
  • MCS, 2006 - 2008
    University of New Brunswick, Canada
  • BCS, 2000
    Shahid Beheshti University, Iran


Research Topic: Hardware Resource Estimation, Digital System Design

Thesis Research:
Investigating Resource Estimation for a High-Level Language
Compression and encryption algorithms are widely studied in transferring data over the networks to satisfy the security and the speed of the data communication. However, the overheads of the compression and encryption algorithms on data transformation have negative acts on real-time data communication. We propose DecRO, a decryption/decompression engine that can be fit in one FPGA. AES and LZ77 are used as decryption and decompression algorithms, respectively. The implementation language for DecRO engine is Handel-C which supports the techniques such as parallelism and pipelining. The implementation results show the equivciency of the engine in using small number of resources while achieving real time performance. A resource estimation framework for Handel-C is also proposed based on our taxonomy in resource estimation algorithms. The proposed framework helps high level programmers improve their design performances and decrease their design development time. Our framework consists of two modules: Global Estimation and Local Estimation. Global Estimation module focuses on the accuracy of the whole design process, while Local Estimation is fast to facilitate the optimization process.

E-mail: f.gharibian@unb.ca
Hobbies: making fun of Joey :)
Website: here

Thesis Supervisor(s):
Ken Kent