CUDA-Enabled Multiple Symbol Detection for PCM/FM Demodulation

Abstract

PCM/FM has been widely used in telemetry system. Multiple Symbol Detection (MSD) based signal demodulation methods can achieve lower BER (Bit Error Rate) than other methods. Due to its high computational complexity, current MSD algorithms are implemented in specialized signal processing devices, such as FPGAs (Field Programmable Gate Arrays). As the rapid development of CUDA, GPU has successfully accelerated applications in a variety of domains. In this paper, we explore to utilize CUDA-enabled GPU to accelerate MSD-based signal demodulation method. The computation core of MSD, sliding correlation problem, is formulated and an efficient parallelization scheme is proposed. CU-MSD (CUDA-enabled MSD) algorithm is implemented by adapting CUDA-enabled sliding correlation. Various optimization techniques are used to achieve the maximum performance. We evaluate our implementation by using data sets from a real aerospace PCM/FM integrated baseband system. The experimental results demonstrate up to 52.8x speedup.