DSP-based design service

Our company is able to provide full range of DSP-based system design services, from mere DSP programming to algorithm development, component selection and system-level design.

Digital signal processor is a microprocessor device optimized for signal processing tasks. In some sense it combines the simplicity of conventional processor programming with the processing power of specialized devices (FPGAs and ASICs).

Given the low cost of DSP programming, using DSP-based design can be an ideal solution for many classes of problems connected with signal processing. DSP is a heart of many widely used multimedia and communications products.

Digital signal processors used to serve as a core of the systems that work with relatively low speed signals (a few MSps) but demanding a sophisticated processing algorithms. Examples include audio processing, voice encoding/decoding, low bandwidth communications (on public telephone networks) and control systems.

However, modern DSPs have enough processing power to process high bandwidth signals. They are often used now in video processing and encoding/decoding, RF signal processing and similar areas.

Another advantage of modern DSPs (such as Texas Instruments TMS320 series or Analog Devices TigerSHARC) is a large quantity of integrated peripherals, including high-speed serial links.

DSP applications

Digital signal processor-based designs are currently widely used in the following areas:

• Audio and Video
• Communications
• Automatic control systems
• Security systems

Typical DSP algorithms

The architecture of digital signal processors is optimized to solve a class of problems including:

• Calculating dot products;
• FIR (finite impulse response) and IIR (infinite impulse response) filters;
• FFT (Fast Fourier Transform) calculation;
• Decoding convolutional codes using Viterbi algorithm.

DSP versus analog

Digital signal processing has undisputable advantages over conventional analog filtering.

1. Less noise and distortion. When digital signal processing is used, the signal is filtered once - before the analog-to-digital converter (ADC) - and quantized. All further processing is made in digital form, thus avoiding the constant signal quality reduction typical for analog systems.
2. Uniformity. Once developed, digital processing system will work equal in every copy of the product. On the contrary, analog system will often need a costly fine-tuning process for every copy because its properties depend heavily on component values that are subject to inevitable dispersion.
3. Flexibility. Digital signal processing system can be easily reprogrammed at any time.

There is a widespread delusion that DSP design is intrinsically inexact because it is based on processing of a discrete signal samples and not the continuous signal. It is not true. In fact, the sampling theorem states that any signal with finite spectrum discretized with the correct sampling frequency can be exactly reconstructed. The only inevitable source of discrepancy is quantization, which can be made as little as needed.

DSP versus FPGA

Compared with FPGAs DSPs can offer simpler design and higher performance in sequential tasks (because of high clock frequency - up to 1200 MHz). FPGAs, conversely, are more suited for problems where paralleling and pipelining can be utilized.

How to make an order?

For any questions concerning our DSP design services you can contact us by e-mail, phone or by using a webform below. You can send us your requirements specification or simply leave your contact information (our managers will contact you as soon as possible). Of course, by contacting us you assume no obligations.

Still have questions?