Lecture 11: Properties of Discrete Systems.

Prof. Vikram M Gadre

EE210.1X
Indian Institute of Technology Bombay
January 29, 2015

IIT Bombay, 2015

This work is licensed under a Creative Commons
Attribution-ShareAlike 4.0 International License.
IITBombayX EE210.1X

1 Introduction
Previously, we were introduced to a new form of systems that is, Discrete In-
dependent Variable Systems. We had also established that the properties of
Discrete Systems are more or less similar to those of the Continuous Systems.
In this session, we look some of the properties of Discrete Systems.

2 Additivity
Definition. A system is said to be additive if the output corresponding to the
sum of any two inputs is the sum of the two outputs.
Thus if we give the input X[n] = X1 [n]+X2 [n] to a system which is additive,
the output of the system will be Y [n] = Y1 [n] + Y2 [n]. (Y1 [n] and Y2 [n] are the
outputs of the system when the inputs are X1 [n] and X2 [n] This is very much
like the continuous time system.
We must note that we are adding the entire signals and not just certain
points of the signal. Also, the signals in the Discrete case are sequences, instead
of functions.

2.1 The Tax Example

We could now look at the tax example which we had previously seen.
Example 1. As we know, the tax collected is given by the equation :

Y [n] = αX[n] + βX[n − 1]

Here α and β are the rates of taxation at the given instances.
In order to illustrate the additivity of this taxation system , let us consider
two states S1 and S2 having the same tax rules. X1 [n] and X2 [n] are the
populations of the two states in the nth interval respectively.
Let S be the system that collects the tax. If we consider both the states
together, we need to find the tax that is collected together.

X[n] = X1 [n] + X2 [n]

Y [n] = αX[n] + βX[n − 1]
Y [n] = αX1 [n] + αX2 [n] + βX1 [n − 1] + βX2 [n − 1]
Y [n] = {αX1 [n] + βX1 [n − 1]} + {αX2 [n] + βX2 [n − 1]}
Y [n] = Y1 [n] + Y2 [n]
The above will be true for any given X1 [n] and X2 [n]. We thus conclude that
the above system is additive.

2
IITBombayX EE210.1X

3 Homogeneity
The properties of continuous as well as discrete systems seem to be more or
less similar. However, there are minor differences in the way we define these
properties.
Definition. A Homogeneous system is defined as one where if the input is
scaled by a given constant, the output is also scaled by the same constant.
Consider a system S. It is given an input X[n], which gives an output Y [n].
When we scale the given input signal by a constant, say α, for the system to be
homogeneous, the output should be scaled by the same constant, α. And this
must be true for every possible input X[n] and every possible constant α.

System is additive ⇐⇒ {X[n] → Y [n] =⇒ αX[n] → αY [n]} (1)

Example 2. We can now illustrate the tax system example taken a few sessions
before.
If we multiply the number of people in the state by α, the tax collected will
also be multiplied by α. Clearly, the system is homogeneous. (Homogeneity ⇐⇒
output is scaled by the same constant with which we scale the input.)

4 Shift Invariance
Shift Invariant system is defined as the one where if we shift the input by a
given amount of time (or given number of samples in this case), the output will
be shifted by the same amount of time (or number of samples).1
Note that in discrete systems, we can only shift the input by integral values
unlike in continuous systems.

5 A Shift Invariant System

Consider a discrete system, with an input X[n], producing an output Y [n].
Now, if we shift the given input by an integer D, we have the input as
X[n − D].
For the system to be shift invariant, it should produce an output which will
be shifted by the same integer D.
X[n] → Y [n]
X[n − D] → Y [n − D]
Where D ∈ Integer and the above holds true for every D and every X[n].

5.1 Tax example – Shift Invariant System

Going back to the taxation example, what Shift invariance implies is that the
rate of the tax remains the same over several years.
So as we jump by a certain amount of time, the tax or the output Y [n]
should continue to remain the same.
1 ShiftInvariance demands a lot from the system. Shift Invariant Systems are very rare in
real life and probably don’t exist. However, we can have systems that are shift invariant over
reasonably large intervals.

3
IITBombayX EE210.1X

6 Causality
Definition. A system is said to be Causal if the output at any time depends
only on the current input and all the past inputs.
Thus, if two inputs to a Causal System are identical up to some value of
n = n0 , the corresponding outputs must also be equal up to this point.
A causal System is also referred to as a non-anticipative system since the
system output does not anticipate future values of the input.
Example 3. A binary adder (digital) circuit is a causal system. You can think
about it and argue that that is indeed true.

Mathematically, a system is said to be causal iff

X1 [n] = X2 [n] ∀n ≤ n0 =⇒ Y1 [n] = Y2 [n] ∀n ≤ n0 (2)

7 Stability
Definition. A signal is said to be bounded iff ∃Mx < ∞ such that X[n] <
Mx ∀n.
Definition. A system is said to be stable if every bounded input to the system
results in a bounded output over the interval [−∞, ∞].
This must hold for all initial samples. What this really says is that as long
as we don’t input ∞ to our system, we won’t get ∞ output. In other words, a
Stable System is one for which reasonable inputs lead to reasonable outputs.

Example 4. Consider the system with a description:

T
1 X
Y [n] = ( ) X[n − k]
2T + 1
k=−T

We can see that as X[n] is a bounded value for all values of n by certain
number β, then the largest possible magnitude of Y [n] will also be β.
This is because Y [n] is the average of a finite set of values of input.

8 Conclusion
We have looked at a few important properties of Discrete Independent Variable
Systems in this Lecture. It is very clear that their properties are very similar to
those of Continuous systems. We shall study more about discrete systems soon.