Jacky Ko


Lincoln Laboratory

In 2012, I was accepted into the Lincoln Laboratory Radar Introduction for Student Engineers (LLRISE) program. The program was a two week introduction course to radar design where we worked in teams of three to build a low cost Doppler radar system capable of distance detection, velocity detection, and SAR imaging. At the end of the course, I presented the project at a temporary exhibit in the MIT Museum.

Radar Design

The schematic of the project to the right shows the simple electronic system that our radars were created from. For the two antennas, we used coffee cans and a wire attached to a microwave connector. For the rest of the parts, we used commercial off-the-shelf (COTS) components and a custom PCB.

The electrical system can be broken into two different parts: transmission and reception. For the transmitting chain, we had a voltage controlled oscillator to generate the wave that we would need to transmit. This then went through both an attenuator and an amplifier to reduce noise before heading to a splitter. For the transmit chain, the input into the splitter went directly to the transmission antenna. The receiving chain combines both the signal from the receiving antenna and the input from the splitter. This is necessary because we wanted to design a Doppler radar which required both the initial wave and the received wave to generate any meaningful data. Both streams of data then went through a low pass filter before being sent into our laptop via a 3.5mm auxiliary cable. The completed Doppler radar system mounted on a wooden plank is shown above center left.

Data Processing

Once the data was received by the laptop, I converted it to a .wav file via a free audio editing program called Audacity. These .wav files could then be sent into MATLAB for processing. The first problem that I ran into was bias. To remove the sensor bias, I subtracted the mean signal from the signal to get axis centered signals. I then took the absolute value of the signals in order to remove negative amplitudes. A Fast Fourier Transformation was later implemented to determine the frequency of signals more accurately.

By using both the transmitted and received frequencies, we were able to calculate the Doppler frequency in MATLAB. Once we had this shift, we could write an equation to calculate velocity using the speed of light. Using pulses, we were able to determine ranges of objects assuming a stationary background. For SAR imaging, we manually moved the Doppler radar and created points using range information. The image resolution was not very good but that was to be expected with the antennas we used. An example of data that we collected, from a swinging pendulum, can be seen to the right.