Hardware Design Considerations
This chapter discusses digital signal processing (DSP) hardware applied to noise reduction. It describes the processing of real-world signals that are converted and represented by sequences of numbers. DSP architectures are designed to process complex signal processing algorithms efficiently, including those used to perform advanced noise reduction processing. The chapter reviews DSP architectures. Many DSP architectures are available today. Each architecture has unique features to address specific problem domains. DSP architectures can be classified by the following attributes: instruction set architecture specialization, component specialization, processors, memory, system, load distribution, and component interaction. The chapter also describes some of the main differences between general-purpose processors and digital signal processors. Filtering and transform algorithms required to perform various types of noise reduction map well to the specialized architectures of modern DSPs. Noise reduction systems that use advanced signal processing implemented on a low-cost, high-performance DSP are an emerging new technology.