Digital signal processor

A digital signal processor is a programmable microprocessor optimized to perform mathematical operations on sampled signals. Its architecture supports fast, repeated calculations used in audio, video, communications, and sensor processing.

Multiply–accumulate operation: A computation that multiplies two values and adds the product to an accumulated total. Repeated multiply–accumulate operations power many digital filters and transforms.
Digital signal: A sequence of discrete numerical values representing a varying quantity over time or space. DSPs operate on digital signals rather than directly on continuous physical quantities.
Audio signal processing: The analysis and transformation of digital representations of sound. DSPs handle equalization, noise reduction, effects, and audio coding.
Microcontroller: An integrated computer containing a processor core, memory, and peripherals for embedded control. Microcontrollers prioritize control and integration, while DSPs emphasize high-throughput signal arithmetic.
Harvard architecture: A computer architecture with separate instruction and data memories or buses. Separate paths let a DSP fetch instructions and data in parallel.
Analog-to-digital converter: An electronic circuit that converts an analog input into digital samples. It supplies numerical samples when a DSP processes sound or sensor measurements.
Digital image processing: The computational manipulation and analysis of images represented as discrete data. Image and video systems use DSP-like operations for filtering, enhancement, and compression.
Central processing unit: The general-purpose processor that executes instructions and coordinates operations in a computer. A CPU handles varied workloads; a DSP is optimized for recurring signal calculations.
Circular buffer: A fixed-size memory region that reuses positions in a loop as data is added and removed. Circular buffers efficiently retain recent samples for filters and other streaming algorithms.
Sampling theorem: A result specifying the conditions under which a band-limited signal can be reconstructed from samples. It sets the sampling-rate requirements for faithfully digitizing signals before DSP processing.

Linked from 9 pages