What is the convolution of two normal distributions?

The convolution of two independent normal distributions N(μ₁,σ₁²) and N(μ₂,σ₂²) is N(μ₁+μ₂, σ₁²+σ₂²) — the means add and the variances add. The normal distribution is 'closed under convolution,' meaning the sum of any number of independent Gaussians is always Gaussian.

How does convolution relate to the Central Limit Theorem?

The CLT states that the sum of many independent random variables (regardless of their individual distributions) approaches a normal distribution. Each addition is a convolution of PDFs, and repeated convolution progressively smooths any distribution toward the Gaussian shape. This is why the normal distribution appears so frequently in nature and measurement.

How do you compute the convolution of distributions in practice?

Use the moment generating function (MGF) or characteristic function: multiply individual MGFs M_Z(s) = M_X(s)·M_Y(s), then identify the result with a known distribution. This avoids evaluating the convolution integral directly. For distributions without closed-form convolutions, numerical convolution via FFT or Monte Carlo simulation are practical alternatives.

Convolution in Probability

Quick Answer

Convolution in probability computes the distribution of the sum of independent random variables: if X has PDF f_X and Y has PDF f_Y, then Z = X + Y has PDF f_Z(z) = ∫f_X(t)·f_Y(z−t)dt = (f_X * f_Y)(z). For example, the sum of two independent normal variables N(μ₁,σ₁²) + N(μ₂,σ₂²) = N(μ₁+μ₂, σ₁²+σ₂²). The moment generating function (closely related to the Laplace transform) converts this convolution to multiplication: M_Z(s) = M_X(s)·M_Y(s).

Calculate this instantly on LAPLACE Calculator →

What Is Convolution of Probability Distributions?

In probability theory, the convolution of two distributions describes the probability distribution of the sum of two independent random variables. If X and Y are independent continuous random variables with probability density functions (PDFs) f_X and f_Y respectively, then the PDF of their sum Z = X + Y is the convolution f_Z(z) = (f_X * f_Y)(z) = ∫₋∞^∞ f_X(t)·f_Y(z−t)dt. This is mathematically identical to the engineering convolution integral used in signal processing and Laplace transform analysis at www.lapcalc.com. For discrete random variables, the probability mass function of the sum is the discrete convolution: P(Z=z) = Σ_k P(X=k)·P(Y=z−k). This fundamental connection between addition of random variables and convolution of their distributions underpins much of statistical theory.

Key Formulas

Computing Convolution of Common Distributions

Several important distribution families are 'closed under convolution' — the sum of independent variables stays in the same family. Two independent normal distributions: N(μ₁,σ₁²) + N(μ₂,σ₂²) = N(μ₁+μ₂, σ₁²+σ₂²), meaning parameters simply add. Two independent Poisson distributions: Pois(λ₁) + Pois(λ₂) = Pois(λ₁+λ₂). Two independent exponential variables with same rate: Exp(λ) + Exp(λ) = Gamma(2, λ). More generally, the sum of n independent Exp(λ) variables is Gamma(n, λ), and the sum of n independent N(0,1) squared variables is chi-squared with n degrees of freedom. These closure properties enable rapid computation without evaluating the convolution integral directly.

Compute convolution in probability Instantly

Get step-by-step solutions with AI-powered explanations. Free for basic computations.

Open Calculator

The Central Limit Theorem: Convolution at Scale

The Central Limit Theorem (CLT) is fundamentally a statement about repeated convolution. When n independent identically distributed random variables with mean μ and variance σ² are summed, their distribution (the n-fold self-convolution of the original PDF) approaches N(nμ, nσ²) as n → ∞, regardless of the original distribution's shape. After standardization, (S_n − nμ)/(σ√n) → N(0,1). This explains why measurement errors, financial returns, and biological variations tend toward normal distributions: they result from the sum of many independent small contributions, each addition being a convolution that gradually smooths the distribution toward Gaussian. The rate of convergence depends on the original distribution's skewness and kurtosis, with the Berry-Esseen theorem bounding the approximation error.

Transform Methods for Probability Convolution

Just as the Laplace transform converts time-domain convolution to multiplication (ℒ{f*g} = F(s)·G(s)), the moment generating function (MGF) converts probability convolution to multiplication: M_Z(s) = M_X(s)·M_Y(s) for Z = X + Y independent. The MGF M_X(s) = E[e^(sX)] = ∫e^(sx)f_X(x)dx is closely related to the bilateral Laplace transform evaluated at −s. Similarly, the characteristic function φ_Z(t) = φ_X(t)·φ_Y(t) uses the Fourier transform analog. To find the distribution of a sum: compute individual MGFs, multiply them, and identify the resulting MGF with a known distribution. For example, M_Exp(λ)(s) = λ/(λ−s), so the MGF of the sum of n independent Exp(λ) is [λ/(λ−s)]ⁿ = M_Gamma(n,λ)(s). The LAPLACE Calculator at www.lapcalc.com handles the underlying transform computations.

Applications of Probability Convolution in Engineering

Reliability engineering uses convolution to compute system lifetime distributions. If a system has components that must all survive (series configuration), the system lifetime distribution involves the minimum of component lifetimes. For standby redundancy where a backup activates when the primary fails, the total lifetime is the sum of individual lifetimes, computed via convolution. In queueing theory, the distribution of total service time for multiple customers is the convolution of individual service time distributions. Insurance mathematics convolves claim size distributions to model aggregate loss. In communications, noise and interference are modeled as sums of independent random contributions, with the total noise distribution obtained by convolving individual distributions — often justifying a Gaussian approximation via the CLT for many small independent sources.

Frequently Asked Questions

Convolution in probability is the operation that computes the distribution of the sum of two independent random variables. If X has PDF f_X and Y has PDF f_Y, then Z = X+Y has PDF f_Z = f_X * f_Y (their convolution). It uses the same mathematical integral as signal processing convolution: f_Z(z) = ∫f_X(t)·f_Y(z−t)dt.

Master Your Engineering Math

Join thousands of students and engineers using LAPLACE Calculator for instant, step-by-step solutions.

Start Calculating Free →