About Digital Signal Processing

Today, signals i.e. quantities that fluctuate over a period of time with usually with high frequency have acquired a great amount of importance and are being used in many fields especially in communication. Digital signal processing involves conversion of digital data into signals thus, making its transfer easier and subsequently converting these signals back into the original form.

A signal has many characteristics or domains such as time domain, spatial domain, frequency, wavelet domain etc. Any one among these can be used to process a respective signal. From among these, the engineer usually selects the one that best represents the characteristics of the signal concerned or in other words, the one from which data can be obtained easily. In order to ascertain the required characteristic, the engineer may try out many among these properties.

The use of signals has gone up especially with the use of computers. Computers are capable of analyzing and processing only digital (discrete) data and cannot handle analogue (continuous) data. Thus, conversion of the signal from analogue to the digital form becomes necessary. In order that the digital signal is exactly similar to the analogue signal that it has been obtained from, some mathematical techniques such as the Nyquist-Shannon sampling theorem are used. Usually after analysis or transformation, the output signal is converted back to the analogue form.

Usually, the time or the space domain of the signal is used for its processing. The signal that is received is usually enhanced by filtering. This process transforms some of the surrounding signals to the one that is under consideration. Among this filtering technique, there are many types:

  • In case of linear filtering, the input signal is transformed linearly to obtain the output signal. In case of non-linear filtering, other transformations take place as well.
  • Filtering can also be classified as casual and non-casual. Casual filtering makes use of only the previous input signal samples whereas; non-casual filtering makes use of the future samples as well.
  • The filters may also be time variant or invariant. Those that are invariant do not change over a period of time while time-variant or adaptive filters change over a period of time.
  • Finite impulse response filters make use of only the current input signals as samples while filtering. Whereas, infinite impulse response filters make use of the output signals that have been generated previously as well.


In order to study the different properties of the signal, it is first converted into the frequency domain from the space or time domain and then, its power spectrum is subsequently obtained from it. Fourier transforms are usually used for this purpose. By the study of the frequency domain, an engineer can find the missing frequencies in a particular signal.

Digital signal processing is likely to acquire more amount of importance in the future along with the advancements in technology. They are used in some type of scientific research as well, apart from the industrial applications of the same.

Leave a Reply

Your email address will not be published. Required fields are marked *