In this paper we propose an FPGA-based hardware platform to accelerate an audio tracking method. Our tracking approach is inspired from the problem of molecular sequence alignment and adopts a well-known dynamic programming algorithm (Smith-Waterman algorithm) from the area of bioinformatics. However, the high computational complexity of such algorithms imposes a significant barrier to their adoption by audio tracking systems. To alleviate the time-consuming problem and achieve realistic response times we propose the acceleration of computationally intensive parts of our tracking method using an FPGA-based platform. Our FPGA accelerator is actually based on the systolization of the Smith-Waterman algorithm proposed in previous approaches for the acceleration of bio-sequence scanning but the special requirements of the audio tracking method impose significant design challenges in the accelerator architecture.