Signals and Systems

9.2 The Region of Convergence (ROC)

Imron Rosyadi

The Region of Convergence (ROC) for Laplace Transforms

Unlocking Signal Properties from the s-Plane

In the previous section, we established that the Region of Convergence (ROC) is vital for a complete Laplace Transform specification. The ROC helps distinguish between signals that might share the same algebraic \(X(s)\) expression.

This section explores specific properties of the ROC, which allow us to:

• Implicitly determine the ROC from \(X(s)\) and signal characteristics.
• Reconstruct the time-domain signal \(x(t)\) from \(X(s)\) and its ROC.
• Understand the time-domain features of \(x(t)\) (e.g., causality, stability) directly from the ROC.

We’ve seen that two different signals can have the same algebraic form for their Laplace Transform. The ROC is the key differentiator. Today, we’ll dive into eight fundamental properties of the ROC. These properties are not just theoretical; they are practical tools that bridge the gap between the complex s-plane and the time-domain characteristics of our signals and systems. Understanding these will allow you to infer a lot about a signal just by looking at its Laplace Transform and its ROC.

Property 1: ROC as Strips Parallel to the \(j\omega\)-axis

Property: The ROC of \(X(s)\) consists of strips parallel to the \(j\omega\)-axis in the \(s\)-plane.

Explanation: The ROC consists of values of \(s = \sigma + j\omega\) for which the Fourier Transform of \(x(t)e^{-\sigma t}\) converges. This means \(x(t)e^{-\sigma t}\) must be absolutely integrable:

\[ \int_{-\infty}^{+\infty}|x(t)| e^{-\sigma t} dt < \infty \quad \text{(9.36)} \]

This condition depends only on \(\sigma\) (the real part of \(s\)), not on \(j\omega\). Therefore, if a vertical line \(\operatorname{Re}\{s\} = \sigma_0\) is in the ROC, then the entire strip containing that line, for all imaginary values, is also in the ROC.

Property 2: ROC Does Not Contain Any Poles

Property: For rational Laplace transforms, the ROC does not contain any poles.

Explanation: At a pole, the Laplace Transform \(X(s)\) is infinite. By definition, for \(s\) to be in the ROC, the integral \(\int_{-\infty}^{+\infty} x(t) e^{-st} dt\) must converge to a finite value. Since \(X(s)\) is infinite at a pole, the integral clearly does not converge at a pole. Therefore, poles always lie on the boundaries of the ROC, never within it.

Property 3: Finite-Duration Signals

Property: If \(x(t)\) is of finite duration and is absolutely integrable, then the ROC is the entire \(s\)-plane.

Explanation: A finite-duration signal is zero outside a finite interval, say \(T_1 < t < T_2\). For \(s=\sigma+j\omega\), we need to check the convergence of \(\int_{T_1}^{T_2}|x(t)| e^{-\sigma t} dt\).

• Since \(x(t)\) is absolutely integrable over \([T_1, T_2]\), \(\int_{T_1}^{T_2}|x(t)| dt < \infty\).
• For any \(\sigma\), \(e^{-\sigma t}\) is bounded over the finite interval \([T_1, T_2]\).
  • If \(\sigma > 0\), \(e^{-\sigma t}\) is bounded by \(e^{-\sigma T_1}\).
  • If \(\sigma < 0\), \(e^{-\sigma t}\) is bounded by \(e^{-\sigma T_2}\).
• Thus, \(|x(t)|e^{-\sigma t}\) remains absolutely integrable for all \(\sigma\).

Tip

Finite-duration signals are “well-behaved” enough that the exponential weighting \(e^{-\sigma t}\) never causes divergence within their non-zero interval.

This property is a bit of a special case but important. If a signal only exists for a finite amount of time, its Laplace Transform will always converge, no matter what value \(s\) takes. This is because the exponential weighting function \(e^{-\sigma t}\) can’t grow unboundedly over a finite interval. Think of a short pulse or a single sine wave cycle. Its transform will converge everywhere.

Example 9.6: Finite Duration Signal

Let \(x(t) = e^{-at}\) for \(0 < t < T\), and \(0\) otherwise.

The Laplace Transform is:

\[ X(s) = \int_{0}^{T} e^{-at} e^{-st} dt = \int_{0}^{T} e^{-(s+a)t} dt = \frac{1}{s+a}\left[1-e^{-(s+a)T}\right] \quad \text{(9.42)} \]

Apparent Pole vs. Actual Behavior

If we look at the expression, it seems like there is a pole at \(s=-a\). However, for a finite-duration signal, the ROC should be the entire \(s\)-plane (Property 3). This implies no poles.

Let’s check the behavior at \(s=-a\) using L’Hôpital’s Rule:

\[ \lim _{s \rightarrow-a} X(s) = \lim _{s \rightarrow-a}\left[\frac{\frac{d}{d s}\left(1-e^{-(s+a) T)}\right.}{\frac{d}{d s}(s+a)}\right] = \lim _{s \rightarrow-a} T e^{-a T} e^{-s T} \]

So, \(X(-a) = T e^{-aT} e^{-(-a)T} = T e^{-aT} e^{aT} = T\).

Since \(X(-a)\) is a finite value (\(T\)), there is no pole at \(s=-a\). The apparent pole is canceled by a zero at the same location.

Example 9.6: Finite Duration Signal

Interactive Visualization: Magnitude of \(X(s)\)

Observe that \(X(s)\) remains finite even when \(s\) approaches \(-a\).

viewof a_ex6 = Inputs.range([-2, 2], {value: 1, step: 0.1, label: "Parameter 'a'"});
viewof T_ex6 = Inputs.range([0.1, 5], {value: 2, step: 0.1, label: "Duration T"});

Example 9.6 beautifully illustrates Property 3. Although the algebraic form of \(X(s)\) might suggest a pole, the fact that the signal is of finite duration means this “pole” must be removable. The interactive plot shows that the magnitude of \(X(s)\) is perfectly finite at \(s=-a\), confirming that it’s not a true pole. This cancellation happens because the numerator also becomes zero at \(s=-a\), leading to an indeterminate form that evaluates to a finite value.

Property 4: Right-Sided Signals

Property: If \(x(t)\) is right-sided, and if the line \(\operatorname{Re}\{s\}=\sigma_0\) is in the ROC, then all values of \(s\) for which \(\operatorname{Re}\{s\}>\sigma_0\) will also be in the ROC.

Definition: A right-sided signal is zero prior to some finite time \(T_1\) (i.e., \(x(t)=0\) for \(t < T_1\)).

Explanation: If \(x(t)e^{-\sigma_0 t}\) is absolutely integrable, then for any \(\sigma_1 > \sigma_0\):

\[ \int_{T_1}^{\infty}|x(t)| e^{-\sigma_1 t} dt = \int_{T_1}^{\infty}|x(t)| e^{-\sigma_0 t} e^{-(\sigma_1-\sigma_0)t} dt \]

Since \(\sigma_1 > \sigma_0\), the term \(e^{-(\sigma_1-\sigma_0)t}\) is a decaying exponential for \(t>T_1\). This means \(e^{-(\sigma_1-\sigma_0)t}\) decays faster than \(e^{-(\sigma_0-\sigma_0)t} = 1\) as \(t \to \infty\). Therefore, if the integral converges for \(\sigma_0\), it will also converge for any \(\sigma_1 > \sigma_0\).

Tip

For a right-sided signal, the ROC is always a right-half plane, extending to the right from some vertical line. If \(X(s)\) is rational, the ROC is to the right of the rightmost pole.

Property 5: Left-Sided Signals

Property: If \(x(t)\) is left-sided, and if the line \(\operatorname{Re}\{s\}=\sigma_0\) is in the ROC, then all values of \(s\) for which \(\operatorname{Re}\{s\}<\sigma_0\) will also be in the ROC.

Definition: A left-sided signal is zero after some finite time \(T_2\) (i.e., \(x(t)=0\) for \(t > T_2\)).

Explanation: The argument is analogous to Property 4. If \(x(t)e^{-\sigma_0 t}\) is absolutely integrable, then for any \(\sigma_1 < \sigma_0\):

\[ \int_{-\infty}^{T_2}|x(t)| e^{-\sigma_1 t} dt = \int_{-\infty}^{T_2}|x(t)| e^{-\sigma_0 t} e^{-(\sigma_1-\sigma_0)t} dt \]

Since \(\sigma_1 < \sigma_0\), the term \(e^{-(\sigma_1-\sigma_0)t}\) is a decaying exponential for \(t<T_2\) (as \(t \to -\infty\)). This means \(e^{-(\sigma_1-\sigma_0)t}\) decays faster than \(1\) as \(t \to -\infty\). Therefore, if the integral converges for \(\sigma_0\), it will also converge for any \(\sigma_1 < \sigma_0\).

Tip

For a left-sided signal, the ROC is always a left-half plane, extending to the left from some vertical line. If \(X(s)\) is rational, the ROC is to the left of the leftmost pole.

Property 6: Two-Sided Signals

Property: If \(x(t)\) is two-sided, and if the line \(\operatorname{Re}\{s\}=\sigma_0\) is in the ROC, then the ROC will consist of a strip in the \(s\)-plane that includes the line \(\operatorname{Re}\{s\}=\sigma_0\).

Definition: A two-sided signal is of infinite extent for both \(t>0\) and \(t<0\).

Explanation: A two-sided signal \(x(t)\) can be decomposed into a right-sided part \(x_R(t)\) and a left-sided part \(x_L(t)\): \(x(t) = x_R(t) + x_L(t)\), where \(x_R(t)\) is \(x(t)u(t-T_0)\) and \(x_L(t)\) is \(x(t)u(T_0-t)\).

• The ROC for \(x_R(t)\) is a right-half plane: \(\operatorname{Re}\{s\} > \sigma_R\).
• The ROC for \(x_L(t)\) is a left-half plane: \(\operatorname{Re}\{s\} < \sigma_L\).

The ROC for \(x(t)\) is the intersection of these two ROCs. This intersection forms a vertical strip: \(\sigma_R < \operatorname{Re}\{s\} < \sigma_L\).

Note

For a two-sided signal, a ROC exists only if \(\sigma_R < \sigma_L\). If \(\sigma_R \ge \sigma_L\), there is no overlap, and thus no Laplace Transform for \(x(t)\).

Example 9.7: Two-Sided Exponential

Let \(x(t) = e^{-b|t|}\). We can write this as \(x(t) = e^{-bt}u(t) + e^{bt}u(-t)\).

1. Right-sided part: \(x_R(t) = e^{-bt}u(t)\) \(\mathcal{L}\{x_R(t)\} = \frac{1}{s+b}\), with ROC: \(\operatorname{Re}\{s\} > -b\).

2. Left-sided part: \(x_L(t) = e^{bt}u(-t)\) \(\mathcal{L}\{x_L(t)\} = \frac{-1}{s-b}\), with ROC: \(\operatorname{Re}\{s\} < b\).

For \(X(s)\) to exist, the ROCs must overlap: \(\operatorname{Re}\{s\} > -b\) AND \(\operatorname{Re}\{s\} < b\). This defines a strip: \(-b < \operatorname{Re}\{s\} < b\).

This overlap exists only if \(b > 0\). If \(b \le 0\), there is no common ROC, and \(x(t)\) has no Laplace Transform.

For \(b > 0\), the combined Laplace Transform is:

\[ X(s) = \frac{1}{s+b} - \frac{1}{s-b} = \frac{(s-b) - (s+b)}{(s+b)(s-b)} = \frac{-2b}{s^2-b^2} \]

\[ X(s) = \frac{-2b}{s^2-b^2}, \quad -b < \operatorname{Re}\{s\} < b \quad \text{(9.51)} \]

Example 9.7: Two-Sided Exponential

Pole-Zero Plot & ROC

• Poles: \(s = b, s = -b\) (roots of \(s^2-b^2=0\))
• Zeros: None in the finite s-plane (numerator is a constant).

Example 9.7 is a classic illustration of a two-sided signal and its strip ROC. The signal \(e^{-b|t|}\) can be seen as a sum of a causal exponential and an anti-causal exponential. For its Laplace Transform to exist, the ROCs of these two components must overlap. This leads to the condition \(b > 0\) and a strip ROC between \(-b\) and \(b\). If \(b\) were negative or zero, these two half-planes would not overlap, and the signal would not have a Laplace Transform. This example clearly shows how the ROC is bounded by the poles of \(X(s)\).

Property 7: Rational \(X(s)\) ROC Bounded by Poles

Property: If the Laplace transform \(X(s)\) of \(x(t)\) is rational, then its ROC is bounded by poles or extends to infinity. In addition, no poles of \(X(s)\) are contained in the ROC.

Explanation: This property combines insights from previous ones:

• Poles as Boundaries: As established by Property 2, poles define the edges of the ROC.
• Structure of Rational Transforms: Rational Laplace Transforms arise from signals composed of sums of exponentials. Each exponential has an ROC that is a half-plane.
• Intersection of Half-Planes: The overall ROC for a sum of exponentials is the intersection of these half-planes, resulting in a strip or a single half-plane bounded by the poles.

Important

This property reinforces that poles are critical in defining the exact shape and location of the ROC for rational transforms.

Property 7 essentially formalizes what we’ve been observing: for rational Laplace Transforms, poles are the geographic markers of the ROC. They define its boundaries. This is a direct consequence of rational transforms arising from sums of exponentials, each of which has a half-plane ROC. The combined ROC is simply the intersection of these half-planes.

Property 8: ROC for Rational \(X(s)\) and Sidedness

Property: If the Laplace transform \(X(s)\) of \(x(t)\) is rational, then:

1. If \(x(t)\) is right-sided, the ROC is the region in the \(s\)-plane to the right of the rightmost pole.
2. If \(x(t)\) is left-sided, the ROC is the region in the \(s\)-plane to the left of the leftmost pole.

Explanation: This property is a direct consequence of Properties 4, 5, and 7.

• Right-sided signals: Their ROCs are right-half planes (Property 4). For rational \(X(s)\), the right-half plane must be bounded by the pole with the largest real part (the rightmost pole) to ensure all terms converge.
• Left-sided signals: Their ROCs are left-half planes (Property 5). For rational \(X(s)\), the left-half plane must be bounded by the pole with the smallest real part (the leftmost pole) to ensure all terms converge.

Tip

This property provides a quick and practical way to determine the ROC for many common signals in ECE, especially when dealing with system impulse responses.

Property 8 is incredibly useful for practical application. Once you have the algebraic form of a rational Laplace Transform and you know whether the corresponding time-domain signal is right-sided or left-sided, you can immediately sketch its ROC. This rule simplifies the process significantly and is a key tool in system analysis, particularly for determining causality and stability from pole locations.

Example 9.8: Multiple ROCs for Same \(X(s)\)

Consider \(X(s) = \frac{1}{(s+1)(s+2)}\). This \(X(s)\) has poles at \(s=-1\) and \(s=-2\).

Based on the ROC properties and the time-domain characteristics of \(x(t)\), there are three possible ROCs for this algebraic expression, each corresponding to a distinct \(x(t)\):

1. Right-Sided Signal (Causal)

• ROC: \(\operatorname{Re}\{s\} > -1\) (right of the rightmost pole).
• Time-domain: \(x(t) = (e^{-t} - e^{-2t})u(t)\).
• Fourier Transform: Exists, as the ROC includes the \(j\omega\)-axis.

2. Left-Sided Signal (Anti-Causal)

• ROC: \(\operatorname{Re}\{s\} < -2\) (left of the leftmost pole).
• Time-domain: \(x(t) = (-e^{-t} + e^{-2t})u(-t)\).
• Fourier Transform: Does not exist, as the ROC does not include the \(j\omega\)-axis.

Example 9.8: Multiple ROCs for Same \(X(s)\) (cont.)

3. Two-Sided Signal

• ROC: \(-2 < \operatorname{Re}\{s\} < -1\) (strip between poles).
• Time-domain: \(x(t) = -e^{-t}u(-t) + e^{-2t}u(t)\).
• Fourier Transform: Does not exist, as the ROC does not include the \(j\omega\)-axis.

This example truly highlights why the ROC is essential for unique signal specification!

Example 9.8 is a cornerstone for understanding the ROC. It shows that an identical algebraic expression for \(X(s)\) can represent three entirely different time-domain signals, depending solely on the specified ROC. Each ROC implies a different type of signal (causal, anti-causal, two-sided) and has different implications for the existence of the Fourier Transform. This is why you must always state the ROC along with the algebraic expression of \(X(s)\).

Interactive ROC and Time-Domain Visualization

Explore the relationship between the ROC and the corresponding time-domain signal for \(X(s) = \frac{1}{(s+1)(s+2)}\).

viewof roc_choice = Inputs.select(
  ["Re{s} > -1 (Right-Sided)", "Re{s} < -2 (Left-Sided)", "-2 < Re{s} < -1 (Two-Sided)"],
  {label: "Select ROC Type"}
);

This interactive element is the culmination of our discussion on ROC properties. By selecting different ROC types, you can directly observe how the same algebraic \(X(s)\) corresponds to radically different time-domain signals. Pay close attention to the causality (whether the signal starts at \(t=0\)) and the behavior for \(t<0\). Also, notice whether the \(j\omega\)-axis is included in the ROC, which tells you if the Fourier Transform exists. This hands-on visualization should solidify your understanding of the profound impact of the ROC.

Key Takeaways: ROC Properties

1. Vertical Strips: ROCs are always vertical strips in the \(s\)-plane.
2. No Poles in ROC: Poles define the boundaries but are never part of the ROC itself.
3. Finite Duration: ROC is the entire \(s\)-plane.
4. Right-Sided: ROC is a right-half plane ($ {s} > _{max_pole} $).
5. Left-Sided: ROC is a left-half plane ($ {s} < _{min_pole} $).
6. Two-Sided: ROC is a strip between poles ($ _L < {s} < _R $).
7. Rational \(X(s)\): ROC is always bounded by poles or extends to infinity.
8. Sidedness Rule: For rational \(X(s)\), sidedness directly dictates the ROC relative to the poles.

Important

The ROC is as crucial as the algebraic expression of \(X(s)\) for uniquely defining a signal and understanding its time-domain characteristics, including causality and stability.

Tip

Always specify the ROC when dealing with Laplace Transforms! It’s not just a mathematical detail; it’s fundamental engineering information.

To summarize, these eight properties are your go-to rules for understanding and working with the ROC. Remember, poles are the critical points that define the ROC boundaries. The type of signal (finite duration, right-sided, left-sided, or two-sided) directly determines the shape of the ROC. This knowledge is not just academic; it’s essential for correctly interpreting system behavior and designing filters and controllers in ECE.