A digital signal processor (DSP) is a single IC-microcomputer device and its functionality mirrors that of a microcontroller however DSPs are normally better suited for high-speed floating point calculations. In other words, a DSP is a processor core or chip for the applications that process digital signals for instance, filtering, noise cancellations, echo eliminations, compression and encryption applications.
A DSP is a key element of an embedded system in most applications that require the processing of signals. DSPs are used in communication, audio/video, and control applications where fast calculation of digital filters and weighted sums is important for fast cycle times. Other applications of a DSP include image processing in multimedia, HDTV, image pattern recognition in DNA sequence, etc.
A DSP as ASIC (Application Specific IC) is a single chip in a VLSI unit. It incorporates the computational capabilities of a microprocessor and multiply and accumulate (MAC) units.
The DSP system executes discrete-time, signal processing instructions. It has very large instruction word (VLIW) processing capabilities. It employs the SIMD multiple processor architecture in executing its instructions. DSP processes discrete cosine transformations (DCT) and inverse DCT (IDCT) functions. The IDCT functions are used in algorithms for signal analyzing, coding, filtering, noise cancellations, echo elimination, compressing and decompressing, and so forth.
Related articles:
- Applications of Embedded Systems
- Embedded Processor vs. Embedded Microcontroller
- Embedded System-On-Chip (SoC)
