Jeff Haoxuan Zheng
Graduate Student in Physics at M.I.T.
LATEST PROJECTS
How to build a radio telescope in 2 minutes, compiled by Max and Meia Chita-Tegmark.
I worked on almost every aspect of this instrument: writing data collection and analysis software, programming FPGAs, validating analogue components, and planning the deployment you see in the video. Using the data we collected on this trip, we developed a new software package for monitoring and calibration that is now being used on other instruments such as PAPER and HERA.






Project |01
Project |01
Automated real time calibration
I'm creating a software package designed to perform real-time calibration and diagnosis with unpreceded speed and precision. It intelligently detects antenna problems, solves for complex gains, and calibrates array alignment. Majority of the package is in Python, whereas the core algorithm is in C++ using the Python/C API. In addition to our own instrument, the package has been deployed on the PAPER array, and will be used on the HERA array. The code is open source at github.com/jeffzhen/omnical


Project |02
Project |02
Foreground mapping in low frequency radio band
I'm working on analyzing the 4TB of data we took in July 2013 to make the best calibrated foreground maps in 120MHz-180MHz band to date. While traditional wisdom says that it's impossible to make maps with our instrument, I have successfully developed a new algorithm that makes great maps, which has never been done at this wavelength.






Project |03
Project |03 (DONE)
FPGA(Field-programmable gate array) design for 128 ant-pol 8 bit correlator
I'm working on the design and implementation of the state of the art correlator using CASPER's ROACH2 boards powered by Xilinx FPGAs. In addition to difficulties in meeting timing constraints due to very high FPGA component usage rate, the design challenges center around the addition of cross communication between 4 F-engines and 4 X-engines as well as the shift from ROACH to ROACH2 boards.


Project |04
Project |04 (DONE)
Correlator control and data collection software package
I am creating a python software package that controls all F-engines and X-engines through various protocols, collects data from 4 X-engines, and stores data in ODF format. Our current data rate is about 2GB/minute.
Just a sample of my work. To see more or discuss possible work >>