Complex random vector

In probability theory and statistics, a complex random vector is typically a tuple of complex-valued random variables, and generally is a random variable taking values in a vector space over the field of complex numbers. If $Z_{1}, \dots, Z_{n}$ are complex-valued random variables, then the n-tuple $(Z_{1}, \dots, Z_{n})$ is a complex random vector. Complex random variables can always be considered as pairs of real random vectors: their real and imaginary parts.

Some concepts of real random vectors have a straightforward generalization to complex random vectors. For example, the definition of the mean of a complex random vector. Other concepts are unique to complex random vectors.

Applications of complex random vectors are found in digital signal processing.

Template:Probability fundamentals

Definition

A complex random vector $𝐙 = (Z_{1}, \dots, Z_{n})^{T}$ on the probability space $(Ω, ℱ, P)$ is a function $𝐙 : Ω \to ℂ^{n}$ such that the vector $(ℜ (Z_{1}), ℑ (Z_{1}), \dots, ℜ (Z_{n}), ℑ (Z_{n}))^{T}$ is a real random vector on $(Ω, ℱ, P)$ where $ℜ (z)$ denotes the real part of $z$ and $ℑ (z)$ denotes the imaginary part of $z$ .^[1]Template:Rp

Cumulative distribution function

The generalization of the cumulative distribution function from real to complex random variables is not obvious because expressions of the form $P (Z \leq 1 + 3 i)$ make no sense. However expressions of the form $P (ℜ (Z) \leq 1, ℑ (Z) \leq 3)$ make sense. Therefore, the cumulative distribution function $F_{𝐙} : ℂ^{n} \mapsto [0, 1]$ of a random vector $𝐙 = (Z_{1}, ..., Z_{n})^{T}$ is defined as

Template:NumBlk

where $𝐳 = (z_{1}, ..., z_{n})^{T}$ .

Expectation

As in the real case the expectation (also called expected value) of a complex random vector is taken component-wise.^[1]Template:Rp

Template:NumBlk

Covariance matrix and pseudo-covariance matrix

Template:See also

The covariance matrix (also called second central moment) $K_{𝐙 𝐙}$ contains the covariances between all pairs of components. The covariance matrix of an $n \times 1$ random vector is an $n \times n$ matrix whose $(i, j)$ ^th element is the covariance between the i^th and the j^th random variables.^[2]Template:Rp Unlike in the case of real random variables, the covariance between two random variables involves the complex conjugate of one of the two. Thus the covariance matrix is a Hermitian matrix.^[1]Template:Rp

Template:NumBlk

K_{𝐙 𝐙} = [\begin{matrix} E [(Z_{1} - E [Z_{1}]) \overline{(Z_{1} - E [Z_{1}])}] & E [(Z_{1} - E [Z_{1}]) \overline{(Z_{2} - E [Z_{2}])}] & \dots & E [(Z_{1} - E [Z_{1}]) \overline{(Z_{n} - E [Z_{n}])}] \\ E [(Z_{2} - E [Z_{2}]) \overline{(Z_{1} - E [Z_{1}])}] & E [(Z_{2} - E [Z_{2}]) \overline{(Z_{2} - E [Z_{2}])}] & \dots & E [(Z_{2} - E [Z_{2}]) \overline{(Z_{n} - E [Z_{n}])}] \\ ⋮ & ⋮ & ⋱ & ⋮ \\ E [(Z_{n} - E [Z_{n}]) \overline{(Z_{1} - E [Z_{1}])}] & E [(Z_{n} - E [Z_{n}]) \overline{(Z_{2} - E [Z_{2}])}] & \dots & E [(Z_{n} - E [Z_{n}]) \overline{(Z_{n} - E [Z_{n}])}] \end{matrix}]

The pseudo-covariance matrix (also called relation matrix) is defined replacing Hermitian transposition by transposition in the definition above.

Template:NumBlk

J_{𝐙 𝐙} = [\begin{matrix} E [(Z_{1} - E [Z_{1}]) (Z_{1} - E [Z_{1}])] & E [(Z_{1} - E [Z_{1}]) (Z_{2} - E [Z_{2}])] & \dots & E [(Z_{1} - E [Z_{1}]) (Z_{n} - E [Z_{n}])] \\ E [(Z_{2} - E [Z_{2}]) (Z_{1} - E [Z_{1}])] & E [(Z_{2} - E [Z_{2}]) (Z_{2} - E [Z_{2}])] & \dots & E [(Z_{2} - E [Z_{2}]) (Z_{n} - E [Z_{n}])] \\ ⋮ & ⋮ & ⋱ & ⋮ \\ E [(Z_{n} - E [Z_{n}]) (Z_{1} - E [Z_{1}])] & E [(Z_{n} - E [Z_{n}]) (Z_{2} - E [Z_{2}])] & \dots & E [(Z_{n} - E [Z_{n}]) (Z_{n} - E [Z_{n}])] \end{matrix}]

Properties

The covariance matrix is a hermitian matrix, i.e.^[1]Template:Rp

K_{𝐙 𝐙}^{H} = K_{𝐙 𝐙}

.

The pseudo-covariance matrix is a symmetric matrix, i.e.

J_{𝐙 𝐙}^{T} = J_{𝐙 𝐙}

.

The covariance matrix is a positive semidefinite matrix, i.e.

𝐚^{H} K_{𝐙 𝐙} 𝐚 \geq 0 for all 𝐚 \in ℂ^{n}

.

Covariance matrices of real and imaginary parts

Template:See also

By decomposing the random vector $𝐙$ into its real part $𝐗 = ℜ (𝐙)$ and imaginary part $𝐘 = ℑ (𝐙)$ (i.e. $𝐙 = 𝐗 + i 𝐘$ ), the pair $(𝐗, 𝐘)$ has a covariance matrix of the form:

[\begin{matrix} K_{𝐗 𝐗} & K_{𝐗 𝐘} \\ K_{𝐘 𝐗} & K_{𝐘 𝐘} \end{matrix}]

The matrices $K_{𝐙 𝐙}$ and $J_{𝐙 𝐙}$ can be related to the covariance matrices of $𝐗$ and $𝐘$ via the following expressions:

\begin{matrix} K_{𝐗 𝐗} = E [(𝐗 - E [𝐗]) (𝐗 - E [𝐗])^{T}] = \frac{1}{2} Re (K_{𝐙 𝐙} + J_{𝐙 𝐙}) \\ K_{𝐘 𝐘} = E [(𝐘 - E [𝐘]) (𝐘 - E [𝐘])^{T}] = \frac{1}{2} Re (K_{𝐙 𝐙} - J_{𝐙 𝐙}) \\ K_{𝐘 𝐗} = E [(𝐘 - E [𝐘]) (𝐗 - E [𝐗])^{T}] = \frac{1}{2} Im (J_{𝐙 𝐙} + K_{𝐙 𝐙}) \\ K_{𝐗 𝐘} = E [(𝐗 - E [𝐗]) (𝐘 - E [𝐘])^{T}] = \frac{1}{2} Im (J_{𝐙 𝐙} - K_{𝐙 𝐙}) \end{matrix}

Conversely:

\begin{matrix} K_{𝐙 𝐙} = K_{𝐗 𝐗} + K_{𝐘 𝐘} + i (K_{𝐘 𝐗} - K_{𝐗 𝐘}) \\ J_{𝐙 𝐙} = K_{𝐗 𝐗} - K_{𝐘 𝐘} + i (K_{𝐘 𝐗} + K_{𝐗 𝐘}) \end{matrix}

Cross-covariance matrix and pseudo-cross-covariance matrix

The cross-covariance matrix between two complex random vectors $𝐙, 𝐖$ is defined as:

Template:NumBlk

K_{𝐙 𝐖} = [\begin{matrix} E [(Z_{1} - E [Z_{1}]) \overline{(W_{1} - E [W_{1}])}] & E [(Z_{1} - E [Z_{1}]) \overline{(W_{2} - E [W_{2}])}] & \dots & E [(Z_{1} - E [Z_{1}]) \overline{(W_{n} - E [W_{n}])}] \\ E [(Z_{2} - E [Z_{2}]) \overline{(W_{1} - E [W_{1}])}] & E [(Z_{2} - E [Z_{2}]) \overline{(W_{2} - E [W_{2}])}] & \dots & E [(Z_{2} - E [Z_{2}]) \overline{(W_{n} - E [W_{n}])}] \\ ⋮ & ⋮ & ⋱ & ⋮ \\ E [(Z_{n} - E [Z_{n}]) \overline{(W_{1} - E [W_{1}])}] & E [(Z_{n} - E [Z_{n}]) \overline{(W_{2} - E [W_{2}])}] & \dots & E [(Z_{n} - E [Z_{n}]) \overline{(W_{n} - E [W_{n}])}] \end{matrix}]

And the pseudo-cross-covariance matrix is defined as:

Template:NumBlk

J_{𝐙 𝐖} = [\begin{matrix} E [(Z_{1} - E [Z_{1}]) (W_{1} - E [W_{1}])] & E [(Z_{1} - E [Z_{1}]) (W_{2} - E [W_{2}])] & \dots & E [(Z_{1} - E [Z_{1}]) (W_{n} - E [W_{n}])] \\ E [(Z_{2} - E [Z_{2}]) (W_{1} - E [W_{1}])] & E [(Z_{2} - E [Z_{2}]) (W_{2} - E [W_{2}])] & \dots & E [(Z_{2} - E [Z_{2}]) (W_{n} - E [W_{n}])] \\ ⋮ & ⋮ & ⋱ & ⋮ \\ E [(Z_{n} - E [Z_{n}]) (W_{1} - E [W_{1}])] & E [(Z_{n} - E [Z_{n}]) (W_{2} - E [W_{2}])] & \dots & E [(Z_{n} - E [Z_{n}]) (W_{n} - E [W_{n}])] \end{matrix}]

Two complex random vectors $𝐙$ and $𝐖$ are called uncorrelated if

K_{𝐙 𝐖} = J_{𝐙 𝐖} = 0

.

Independence

Template:Main Two complex random vectors $𝐙 = (Z_{1}, ..., Z_{m})^{T}$ and $𝐖 = (W_{1}, ..., W_{n})^{T}$ are called independent if

Template:NumBlk

where $F_{𝐙} (𝐳)$ and $F_{𝐖} (𝐰)$ denote the cumulative distribution functions of $𝐙$ and $𝐖$ as defined in Template:EquationNote and $F_{𝐙, 𝐖} (𝐳, 𝐰)$ denotes their joint cumulative distribution function. Independence of $𝐙$ and $𝐖$ is often denoted by $𝐙 ⊥ ⊥ 𝐖$ . Written component-wise, $𝐙$ and $𝐖$ are called independent if

F_{Z_{1}, \dots, Z_{m}, W_{1}, \dots, W_{n}} (z_{1}, \dots, z_{m}, w_{1}, \dots, w_{n}) = F_{Z_{1}, \dots, Z_{m}} (z_{1}, \dots, z_{m}) \cdot F_{W_{1}, \dots, W_{n}} (w_{1}, \dots, w_{n}) for all z_{1}, \dots, z_{m}, w_{1}, \dots, w_{n}

.

Circular symmetry

A complex random vector $𝐙$ is called circularly symmetric if for every deterministic $φ \in [- π, π)$ the distribution of $e^{i φ} 𝐙$ equals the distribution of $𝐙$ .^[3]Template:Rp

Properties

The expectation of a circularly symmetric complex random vector is either zero or it is not defined.^[3]Template:Rp
The pseudo-covariance matrix of a circularly symmetric complex random vector is zero.^[3]Template:Rp

Proper complex random vectors

A complex random vector $𝐙$ is called proper if the following three conditions are all satisfied:^[1]Template:Rp

$E [𝐙] = 0$ (zero mean)
$var [Z_{1}] < \infty, \dots, var [Z_{n}] < \infty$ (all components have finite variance)
$E [𝐙 𝐙^{T}] = 0$

Two complex random vectors $𝐙, 𝐖$ are called jointly proper is the composite random vector $(Z_{1}, Z_{2}, \dots, Z_{m}, W_{1}, W_{2}, \dots, W_{n})^{T}$ is proper.

Properties

A complex random vector $𝐙$ is proper if, and only if, for all (deterministic) vectors $𝐜 \in ℂ^{n}$ the complex random variable $𝐜^{T} 𝐙$ is proper.^[1]Template:Rp
Linear transformations of proper complex random vectors are proper, i.e. if $𝐙$ is a proper random vectors with $n$ components and $A$ is a deterministic $m \times n$ matrix, then the complex random vector $A 𝐙$ is also proper.^[1]Template:Rp
Every circularly symmetric complex random vector with finite variance of all its components is proper.^[1]Template:Rp
There are proper complex random vectors that are not circularly symmetric.^[1]Template:Rp
A real random vector is proper if and only if it is constant.
Two jointly proper complex random vectors are uncorrelated if and only if their covariance matrix is zero, i.e. if $K_{𝐙 𝐖} = 0$ .

Cauchy-Schwarz inequality

The Cauchy-Schwarz inequality for complex random vectors is

{| E [𝐙^{H} 𝐖] |}^{2} \leq E [𝐙^{H} 𝐙] E [| 𝐖^{H} 𝐖 |]

.

Characteristic function

The characteristic function of a complex random vector $𝐙$ with $n$ components is a function $ℂ^{n} \to ℂ$ defined by:^[1]Template:Rp

φ_{𝐙} (𝝎) = E [e^{i ℜ (𝝎^{H} 𝐙)}] = E [e^{i (ℜ (ω_{1}) ℜ (Z_{1}) + ℑ (ω_{1}) ℑ (Z_{1}) + \dots + ℜ (ω_{n}) ℜ (Z_{n}) + ℑ (ω_{n}) ℑ (Z_{n}))}]

References

Template:Reflist

↑ ^1.00 ^1.01 ^1.02 ^1.03 ^1.04 ^1.05 ^1.06 ^1.07 ^1.08 ^1.09 Template:Cite book
↑ Template:Cite book
↑ ^3.0 ^3.1 ^3.2 Template:Cite book

[Lapidoth-1] 1.00 ^1.01 ^1.02 ^1.03 ^1.04 ^1.05 ^1.06 ^1.07 ^1.08 ^1.09 Template:Cite book

[Gubner-2] Template:Cite book

[TseViswanath-3] 3.0 ^3.1 ^3.2 Template:Cite book

[1]

[2]

[3]

Complex random vector

Contents

Definition

Cumulative distribution function

Expectation

Covariance matrix and pseudo-covariance matrix

Covariance matrices of real and imaginary parts

Cross-covariance matrix and pseudo-cross-covariance matrix

Independence

Circular symmetry

Proper complex random vectors

Cauchy-Schwarz inequality

Characteristic function

See also

References

Navigation menu

Complex random vector

Definition

Cumulative distribution function

Expectation

Covariance matrix and pseudo-covariance matrix

Covariance matrices of real and imaginary parts

Cross-covariance matrix and pseudo-cross-covariance matrix

Independence

Circular symmetry

Proper complex random vectors

Cauchy-Schwarz inequality

Characteristic function

See also

References

Navigation menu

Search