其他分享
首页 > 其他分享> > Random variables and Random Process/随机变量和随机过程

Random variables and Random Process/随机变量和随机过程

作者:互联网

Random Variables:

(随机变量)

1.Bernoulli Random Variable:

(伯努利随机变量)

It is a discrete binary-valued that takes 0 and 1 with probability 1-p and p respectively.

(它是一个离散的二进制值,取0和1的概率分别为1-p和p)

P[x=1]=p,P[x=0]=1-p

E[x]=\sum x*P[X=x]=1*p+0*(1-p)=p

Var[x]=E[(x-E[x])^{2}]=p-p^{2}

2.Binomial Random Variable:

(二项随机变量)

This random varible models the number of heads when a coin is flipped n times and the probability of head is p,noted as:

(该随机变量是一个硬币抛n次,p次得到正面的概率的模型,记为)

X\sim B(n,p)

P[x=k]=\begin{pmatrix} n\\p \end{pmatrix}*p^{k}(1-p)^{n-k}

E[x]=\sum k*P[x=k]=n*p

Var[x]=n*p*(1-p)

3.Uniform Random Variable

(均匀随机变量)

It is continuous random variable with pdf(probability density function)

(它是一个连续的随机变量,其概率密度函数如下:)

 p(x)=\begin{cases} 1/b-a & \text{ if } a<x<b \\ 0& \text{ otherwise} \end{cases}

E[x]=(a+b)/2

Var[x]=(b-a)^{2}/12

4.Gaussian Random Varriable:

(高斯随机变量)

It is continuous random variable defined as X\sim N(m,\sigma ^{2})

(它是一个连续的随机变量,记为:X\sim N(m,\sigma ^{2})

where m is the mean that E[x]=m

(在这里,m是它的平均值,E[x]=m)

σ is the standard giving the pdf p(x)=e^{-(x-m)^{2}/2\sigma ^{2}}/\sqrt{2*\pi *\sigma^{2}}

(使用σ规定标准的概率密度函数为 p(x)=e^{-(x-m)^{2}/2\sigma ^{2}}/\sqrt{2*\pi *\sigma^{2}}

Var[x]=\sigma ^{2}

In Gaussian random variable,there are some definition to notice:

(在高斯随机变量中,需要注意以下几个定义:)

I.Q function:

(Q 函数)

Q(x)=P[N(0,1)>x]=\int_{x}^{\infty }e^{-t^{2}/2}/\sqrt{2*\pi} dt

II.CDF:cumulative distribution function

(累积分布函数)

F(x)=P[X<x]=\int_{-\infty}^{x }e^{-t^{2}/2}/\sqrt{2*\pi} dt=1-\int_{x}^{\infty}e^{-t^{2}/2}/\sqrt{2*\pi} dt=1-\int_{(x-m)/\sigma}^{\infty }e^{-u^{2}/2}/\sqrt{2*\pi} du=1-Q((x-m)/\sigma)

where u=(t-m)/σ

III.Complementary Error Function:

(余误差函数)

erfc(x)=\int_{x}^{\infty }2*e^{-t^{2}}/\sqrt{\pi} dt

with the definitions above,we have:

(根据这些定义,我们可以推出:)

P(x>a)=1-F(a)=Q((a-m)/σ)

P(x<a)=F(a)=Q((m-a)/σ)

Q(x)=0.5*erfc(x/1.414)

Limit theorems of sums of random variable:

(随机变量和的极限定理)

Given a set of random variable x_{i},where i=1~n,there are two limit theorems governing their running average:

(有两个极限定理控制它们的滑动平均值)

1.The law of large number:If n becomes sufficiently large,then \frac{1}{n}\sum_{i=1}^{n}X_{i}converges to E[X_{i}]

(大数定理:如果n足够大,那么\frac{1}{n}\sum_{i=1}^{n}X_{i}收敛于E[X_{i}]

2.Central limit:If n becomes sufficiently large,then \frac{\frac{1}{n}\sum_{i=1}^{n}X_{i}-m}{\sigma /\sqrt{n}}converges to N(0,1)

(中心极限法则:如果n足够大,那么\frac{\frac{1}{n}\sum_{i=1}^{n}X_{i}-m}{\sigma /\sqrt{n}}收敛于N(0,1))

Random Process:

(随机过程)

A random process is a collection of infinite random variables{x(t)} defined over a common probability space.

(随机过程是由公共概率空间上定义的无限个随机变量组成的)

The mean m_{x}(t) of a random process x(t) is defined as m_{x}(t)=E[x(t)]

(其平均值m_{x}(t)定义为:m_{x}(t)=E[x(t)]

The autocorrelation of x(t) is: R_{x}(t_{1},t_{2})=E[x(t_{1}),x^{*}(t_{2})]

(其自相关函数定义为:R_{x}(t_{1},t_{2})=E[x(t_{1}),x^{*}(t_{2})]

The cross-correlation of x(t) and y(t) is:R_{xy}(t_{1},t_{2})=E[x(t_{1}),y^{*}(t_{2})]

(其互相关函数定义为:R_{xy}(t_{1},t_{2})=E[x(t_{1}),y^{*}(t_{2})]

Wide-Sense Stationary Random Process:

(广义平稳随机过程)

x(t) is Wide-Sense Stationary if its mean is constant and :R_{x}(t_{1},t_{2})=R_{x}(\tau ) where \tau =t_{1}-t_{2}

(如果x(t)的平均值是个常数并且其自相关函数R_{x}(t_{1},t_{2})=R_{x}(\tau )\tau =t_{1}-t_{2},那么它就是广义平稳的)

x(t) and y(t) are Wide-Sense Stationary if they are both Wide-Sense Stationary and R_{xy}(t_{1},t_{2})=R_{xy}(\tau ),where\tau =t_{1}-t_{2}

(如果x(t),y(t)分别广义平稳并且其互相关函数R_{xy}(t_{1},t_{2})=R_{xy}(\tau )\tau =t_{1}-t_{2},那么它们互相广义平稳)

Wiener-khinchin theorem:

(维纳-辛钦定理)

PSD:Power spectral density

The PSD of Wide-Sense Stationary Random Process S_{x}(f)=F[R_{x}(\tau )]

(广义平稳随机过程的功率谱密度函数S_{x}(f)=F[R_{x}(\tau )]

The power of  x(t) is P_{x}=E[\left | x(t) \right |^{2}]=\int_{-\infty }^{\infty }S_{x}(f)df

(其能量为P_{x}=E[\left | x(t) \right |^{2}]=\int_{-\infty }^{\infty }S_{x}(f)df

When a Wide-Sense Stationary signal x(t) pass through an LTI system,the output y(t) and x(t) are Wide-Sense Stationary.

In this situation,there are equations:

(当一个广义平稳的信号x(t)通过线性时不变系统时,输出y(t)和输入x(t)是广义平稳的,在这情况下,有以下的公式需注意:)

1.m_{y}=m_{x}*\int_{-\infty }^{\infty }h(t)dt=m_{x}*H(0)

2.R_{xy}(\tau)=R_{x}(\tau )*h(-\tau)

3.R_{y}(\tau)=R_{x}(\tau )*h(\tau)*h^{*}(-\tau)

4.Cross spectral density: S_{xy}(f)=S_{x}(f)*H^{*}(f)

(交叉谱密度)

5.PSD:S_{y}(f)=S_{x}(f)*\left | H(f) \right |^{2}

Gaussian Random Process:

(高斯随机过程)

A real random process x(t) is Gaussian if all random variable x_{i}(t) for any i>0 are jointly gaussian.

(如果对于任何的i>0,所有随机变量x_{i}(t)都是联合高斯的,则实随机过程x(t)是高斯的。)

Linear filtering of Gaussian random process results in a Gaussian random process.

(对一个高斯随机过程线性滤波,产生的结果依然是高斯随机过程)

White Random Process:

(白随机过程)

A random process is white if its PSD is constant for all frequencies.

(如果一个随机过程的功率谱密度在所有功率上都是常数,那么它是一个白随即过程)

标签:random,Process,Stationary,variables,Random,随机,Sense,随机变量
来源: https://blog.csdn.net/Katsura95/article/details/87965944