In seeking to design a digital filter, it is important to follow a step-by-step guide from specification to implementation. Every option in the design process is open to all designers. Most of the experts in digital signal processing apply some digital filter theory that especially reflects the considerable research effort, the approach to find a useful method to calculate the filter coefficient.

What is digital filter?

It is a network or a precise system by which we can change the waveform, amplitude, phase, frequency and other characteristics of the signal in the desired way. The main goals of using the filter in digital signal processing are to improve signal quality by reducing noise, extracting or combining the desired portion of the signal. Basically, in signal processing technology, it is an algorithm that is implemented in software and hardware to generate a digital output signal from the input signal, obviously, in the desired way. It acts on a routine filtering algorithm that represents some variables, numbers or digitized analog signals in specific hardware or software.

There are few steps in real time digital filter with analog input. First, the bandlimiter limits the analog signal to take a periodic sample and convert it to a series of binary samples. Second, a processor is used to process these input streams into output streams with some kind of computational mathematical algorithm. After that, the processed signal is converted to analog signal by reducing some unwanted components.

Advantages

• It has linear phase response.
• Thermal and environmental variation cannot change performance.
• It is the so-called adaptive filter because the frequency response can be adjusted automatically with the implementation of a programmable processor.
• It is possible to filter multiple input streams without any hardware replication.
• All data can be stored
• It has repeatable unit-to-unit performance.
• Because its level of operation is low frequency, it is used where the use of the analog system is not practical.

Disadvantages

• In reality, the signal bandwidth of the digital stream is much less than that of the analog stream. Signal processing speed is one of the key factors in calculating overall device performance. In reality, the operation speed is totally dependent on the number of arithmetic operations in the processor.
• The finite word length effect, which produces quantization noise and rounding noise, is another major drawback during computation.
• Much more time is needed to design and develop the digital sequences, although it can be used on other tasks or applications once developed. Good computer aided design support can usually turn these into enjoyable tasks.