Laboratory for High Performance DSP Research (HPDSPR)

Advances in high performance digital signal processing and technical computing, low-latency internetworking, and RAM based storage of data with real-time access have been transforming R&D activities of almost every discipline. IT centric and data intensive scientific discovery offers unmatched opportunities with outstanding potential impact, and demanding cutting edge High Performance Technical Computing & Internet Engineering skill sets. Many industries have already built state of the art information processing infrastructure enabling themselves to reinvent for further growth.

HPDSPR is an industry agnostic, inter-disciplinary Research & Development laboratory of excellence to advance theory and implementation of high performance DSP & technical computing for new technology and applications development.

HPDSPR activities are highlighted below.

GRADUATE COURSES

• ECE 641 Laboratory for High Performance Digital Signal Processing (offered every semester)
  
  This course first introduces today's FPGA and GPU technologies, VHDL and CUDA coding, the design tools for the state-of-the-art DSP algorithms and systems. It focuses on computer arithmetic including possible number representations for DSP with FPGA like distributed arithmetic (DA) and CORDIC algorithm. Then, it introduces CUDA development tools for GPUs. Finally, there is a set of high performance DSP implementations spanning from finite impulse response and infinite impulse response filters to wavelet processors with two-channel filter banks and others. Each student is also assigned a term project for the course to be implemented on FPGA or GPU.

ECE 641 STUDENT PROJECT PRESENTATIONS, Spring 2011, Instructor: Mustafa U. Torun


• ECE 640 Digital Signal Processing (offered every semester)
  
  
• ECE 740 Advanced Digital Signal Processing (offered every year)
  
  
• ECE 747 Signal Decomposition Techniques: Transforms, Subbands and Wavelets (offered every year)
  

  TECHNOLOGY PARTNERS

• ALTERA
• JUNIPER NETWORKS
• NVIDIA

  THESIS STUDENTS

  Ph.D. Students

• Mustafa U. Torun, High Performance Signal Processing: Theory, Design, and Applications in Finance. Ph.D. Candidate, 2008-Present.
• Yuewen Wang, Generalized DFT with Nonlinear Phase for MIMO Radar and Communications. Ph.D. Candidate, 2009-Present. (Co-Advisor: A. Haimovich)
• Yanjia Sun, On Automated Face Reading. Ph.D. Candidate, 2009-Present. (Co-Advisor: J. Carpinelli)
• Onur Yilmaz, Ph.D. Candidate, 2011-Present.
• Irfan Lateef, Ph.D. Candidate, 2011-Present.

M.S. Students

• Boyan Zhang, FPGA and GPU Implementation of Block Transforms. M.S. Thesis. 2011-.
• Di Mu, Single-Carrier Frequency Domain Equalization Using Subband Decomposition For Optical Wireless Communications. M.S. Thesis. 2011. (Co-Advisor: Y. Bar-Ness)
• Christopher N. Njoku, A Simple Trading Algorithm. M.S. Project. 2011.

ANNOUNCEMENTS

• Special Issue on Signal Processing Methods in Finance and Electronic Trading, IEEE Journal of Selected Topics in Signal Processing, 2012.

PLENARY TALKS, TUTORIALS AND SPECIAL SESSIONS

• A.N. Akansu, "Financial Signal Processing and High Frequency Trading: A Killer App for Smart Clouds," The 13th IEEE International Workshop on Signal Processing Advances in Wireless Communications, Cesme, Turkey, June 2012.
• A.N. Akansu, "Electronic Trading: A Data Intensive Discovery," International Conference on Mathematical Finance and Economics, Istanbul, Turkey, July 6, 2011.
• A.N. Akansu, I. Pollak, and F. Rubio, "Electronic Trading and Portfolio Optimization: A Signal Processing Perspective," EUSIPCO European Signal Processing Conference, Barcelona, Spain, Aug. 29, 2011.
• A.N. Akansu and M. Avellaneda (Organizers), Special Session: "Signal Processing Methods for Finance Applications," IEEE ICASSP, Prague, Czech Republic, May 24, 2011.

PAPERS

• A.N. Akansu and M.U. Torun, "Toeplitz Approximation to Empirical Correlation Matrix of Asset Returns: A Signal Processing Perspective," IEEE Journal of Selected Topics in Signal Processing, to appear, 2012.


• M.U. Torun, O. Yilmaz and A.N. Akansu, "Novel GPU Implementation of Jacobi Algorithm for Karhunen-Loève Transform of Dense Matrices," IEEE 46th Annual Conference on Information Sciences and Systems (CISS), March 2012.


• A.N. Akansu and M.U. Torun, "On Toeplitz Approximation to Empirical Correlation Matrix of Financial Asset Returns," IEEE 46th Annual Conference on Information Sciences and Systems (CISS), March 2012.


• Y. Wang and A.N. Akansu, "Generalized DFT Based Partial Matched Filter Bank for Doppler Estimation," IEEE 46th Annual Conference on Information Sciences and Systems (CISS), March 2012.


• W.P. Weydig, M.U. Torun and A.N. Akansu, "Implementation of Generalized DFT on Field Programmable Gate Array," IEEE ICASSP, March 2012.


• M.U. Torun and A.N. Akansu, "On Epps Effect and Rebalancing of Hedged Portfolio in Multiple Frequencies," The Fourth International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, San Juan, Puerto Rico, Dec. 2011.


• M.U. Torun, A.N. Akansu and M. Avellaneda, "Portfolio Risk in Multiple Frequencies," IEEE Signal Processing Magazine, vol. 28, no. 5, pp. 61-71, Sept. 2011.


• M.U. Torun and A.N. Akansu, "On Basic Price Model and Volatility in Multiple Frequencies," Proc. IEEE Statistical Signal Processing Workshop, Nice, France, June 2011.


• M.U. Torun, A.N. Akansu and M. Avellaneda, "Risk Management for Trading in Multiple Frequencies," Proc. IEEE ICASSP, Prague, Czech Republic, May 2011.



LINKS

The Fourth Paradigm: Data-Intensive Scientific Discovery

Click here to go back to the home page .