Graphics Processing Unit Based High Speed Demodulation

Demodulation of high-speed, wide-bandwidth wave forms has typically been performed by ASICs or large CPU clusters. However, development times for these technologies can be long and expensive, especially for ASICs. While general purpose processors are flexible and easier to program, they often lack the throughput required to handle high-speed waveform demodulation. GPUs open up new avenues for flexible demodulators that can be developed quickly and modified and maintained easily. This project seeks to harness the power of GPU processing for high-speed demodulation. To do this, the design uses an industry-leading GPU development environment to optimize the
repetitive processes during demodulation. The system can exploit these processes through parallelization while maintaining a modest implementation loss.