Advanced Audio Coding (AAC) on an FPGA

Student: Ryan Linneman

Supervisor: Peter Sutton

Category: Engineering Thesis Project - Computer Systems

The Internet Audio market is rapidly expanding, and new algorithms for encoding audio data more efficiently are attracting more attention than ever before. MPEG-2 Advanced Audio Coding (AAC) was standardised in 1997 but is only now emerging as the audio coding standard of the future. AAC offers a 30% performance increase over the popular MP3 format and is capable of producing a ‘better than CD quality’ output.

This thesis presents an investigation and partial implementation of an AAC decoder on an FPGA (Field Programmable Gate Array). The hardware implementation includes twoof the five essential coding tools specified by the AAC standard (ISO/IEC 13818-7); the Quantisation tool and the Scalefactor tool. The hardware description was written using the VHDL language and a Xess XSV-300 development board was used for the basis of the hardware implementation environment.

This thesis makes three significant contributions toward a documented solution for an FPGA-based AAC decoder:

1. The VHDL source code for the full implementation of two of the five essential tools used by the AAC decoding algorithm.

2. A detailed description of the conformance bitstream used for the VHDL test bench to aid understanding and testing.

3. The inclusion of suggestions and recommendations for future work and development with the MPEG-2 AAC standard, particularly the Filter Bank tool and the decoder controller.

Thesis Document (PDF)

Other Related Files

Poster Presentation (PDF)