What is Finite Impulse Response (FIR) ?

A Finite Impulse Response of FIR filter, as the name suggests, has a finite duration impulse response
as opposed to an Infinite Impulse Response (IIR) filter which has an infinite duration impulse
response. Mathematically, this means the output of an FIR filter y(n) in response to an impulse input
x(n) = δ(n) dies out within a finite time. The impulse response of an FIR filter is represented as h(n)
which is non-zero only for a finite number of samples. This gives FIR filters an advantage over IIR
filters in terms of stability which we will explore more in the later sections.

In simple terms, an FIR filter is a digital filter whose impulse response is of finite duration because it
settles to zero in finite time. This makes the mathematical analysis and implementation of FIR filters
relatively simple compared to IIR filters. The filter coefficients of an FIR filter represent the impulse
response of the filter. FIR filters are widely used in digital signal processing applications such as
filtering, correlation, convolution, etc. due to their stability and ease of design advantages which will
be covered in detail ahead.

Properties of Finite Impulse Response (FIR)

o Stability: Since the impulse response of an FIR filter is of finite duration, it always results in a
stable filter implementation. IIR filters on the other hand can become unstable based on the
choice of coefficients.
o Linear Phase: FIR filters can be designed to have a linear phase characteristic which means
the phase shift introduced by the filter is constant across its frequency band. This preserves
waveform shapes.
o Ease of design: The design methods for FIR filters like windowing, convolution, and least
squares are simpler to implement compared to IIR filters.
o Non-recursive implementation: FIR filters can be implemented without feedback which
simplifies their hardware and software realizations compared to IIR filters that use feedback.
o Exact linearity: FIR filters demonstrate exact linearity and do not suffer from problems like
saturation which may occur in IIR filter implementations.
o Superior approximation: FIR filters can approximate any desired frequency response more
accurately compared to IIR filters of the same order especially in transition regions of the
filter.

These properties make FIR filters suitable for many applications requiring stability, linear phase
response, and simpler implementation structures. However, their main drawback is higher filter order
requirements compared to IIR filters.
Frequency Response of FIR Filters
FIR filters have a frequency response that can be easily tailored to specific applications by adjusting
the filter coefficients. The response is typically described in terms of magnitude and phase:

o Magnitude Response: The magnitude response determines how much of each frequency
component in the input signal is passed through the filter. By carefully selecting the filter's
coefficients, it is possible to attenuate certain frequencies while allowing others to pass,
resulting in the desired filtering effect.
o Phase Response: One of the most significant advantages of FIR filters is their linear phase
response, which ensures that all frequency components of the input signal are delayed by the
same amount of time. This property eliminates phase distortion, making FIR filters ideal for
applications where preserving the waveform is critical.

FIR Filter Design

o Windowing Method: This is one of the simplest methods for FIR filter design. It involves
multiplying the ideal filter impulse response by a window function to obtain a finite-length
filter. Commonly used windows include the Hamming, Hanning, and Blackman windows.
o Frequency Sampling Method: This technique designs FIR filters by specifying the desired
frequency response at a set of sample points. The filter coefficients are then computed using
an inverse discrete Fourier transform (IDFT).
o Optimization Techniques: Advanced optimization algorithms, such as the Parks-McClellan
algorithm, are used to design FIR filters that meet strict performance criteria, such as
minimizing the maximum error between the actual and desired frequency response.

Characteristics of FIR Digital Filters

o Exact linear phase response: FIR filters can be designed as a linear phase filter resulting in
constant group delay which translates input signals without distortion in the passband. This
is useful in applications like pulse shaping.
o Narrow transition bands: FIR filters are better able to approximate sharp transitions
compared to IIR filters leading to narrow transition regions between passband and stopband.
o No overshoot: Since FIR filters have an inherently stable structure, they do not suffer from
problems like overshoot which may occur during startup or in case of improper IIR filter
designs.
o Exact representation of arbitrary magnitude responses: FIR filters can accurately replicate
any theoretically ideal magnitude responses using design techniques like frequency sampling
compared to approximated IIR magnitude responses.
 o Easier implementation: Absence of feedback terms makes FIR filters simpler to implement
in hardware as well as software without complicated multiplier-accumulator structures
needed for IIR filters.
o Insensitivity to quantization errors: Quantization effects have minimal impact on FIR filters
owing to their non-recursive nature compared to IIR filters which may become unstable due
to quantization noise.

These characteristics make FIR filters suitable for a wide range of DSP applications requiring linear
phase response, stability, and simplicity in structures.

Advantages and Disadvantages of Finite Impulse

Advantages of FIR Filter

o Stability: As mentioned earlier, FIR filters are inherently stable due to their non-recursive
nature.
o Exact linear phase response: FIR filters can be designed to have linear phase thus
preserving signal waveforms.
o Ease of design: Windowing, frequency sampling methods are simple to implement for FIR
filter design.
o Ease of implementation: Non-recursive structure makes FIR filters simpler to realize
compared to IIR filters.
o Superior control of transitions: FIR filters can control passband-stopband transitions better
than IIR especially in multiband cases.

Disadvantages of FIR Filter

o Higher order required: For same filtering performance, FIR filters require higher orders than
IIR filters leading to more complex designs and computations.
o Higher delay: Due to their higher order, FIR filters introduce more delay compared to lower
order IIR filters of the same specifications.
o Greater computational complexity: Higher order FIR filters require more multiplication-
accumulation operations compared to equivalent IIR filters.
o More coefficient precision required: Owing to non-recursive nature, FIR filters demand
higher coefficient precision compared to IIR filters.