频域中的离散时间信号表示题目

文章目录

CTFT

证明若 x a ( t ) x_a(t) 是有限的,式
X a ( j Ω ) = x a ( t ) e j Ω t d t X_a(j\Omega)=\int_{-\infty}^{\infty}x_a(t)e^{-j\Omega t}dt
定义的 X a ( j Ω ) X_a(j\Omega) 的绝对值绝对可积。
解:
x a ( t ) d t < \int_{-\infty}^{\infty}\vert x_a(t)\vert dt < \infty
X a ( j Ω ) = x a ( t ) e j Ω t d t x a ( t ) e j Ω t d t = x a ( t ) d t < \vert X_a(j\Omega)\vert =\vert \int_{-\infty}^{\infty}x_a(t)e^{-j\Omega t}dt\vert \leq \int_{-\infty}^{\infty}\vert x_a(t)\vert \vert e^{-j\Omega t} \vert dt=\int_{-\infty}^{\infty}\vert x_a(t)\vert dt < \infty


求定义在 < t < -\infty < t < \infty 的下列连续时间函数的CTFT:
(a) y a ( t ) = s i n ( Ω a t ) y_a(t)=sin(\Omega _at)
(b) u a ( t ) = e α t u_a(t)=e^{-\alpha\vert t\vert}
(c) v a ( t ) = e j Ω 0 t v_a(t)=e^{j\Omega_0t}
(d) p a ( t ) = l = δ ( t l T ) p_a(t)=\sum_{l=-\infty}^{\infty}\delta(t-lT)
(e) g a ( t ) = e a t 2 g_a(t)=e^{-at^2}
解:
(a)
s i n ( Ω 0 t ) = 1 2 j ( e j Ω 0 t e j Ω 0 t ) sin(\Omega_0t)=\frac{1}{2j}(e^{j\Omega_0t}-e^{-j\Omega_0t})

1 C T F T 2 π δ ( Ω ) 1\xrightarrow{CTFT}2\pi \delta(\Omega)
所以
1 e j Ω 0 t C T F T 2 π δ ( Ω Ω 0 ) 1 e j Ω 0 t C T F T 2 π δ ( Ω + Ω 0 ) 1\cdot e^{j\Omega_0t}\xrightarrow{CTFT}2\pi \delta(\Omega - \Omega_0) \\ 1 \cdot e^{-j\Omega_0t}\xrightarrow{CTFT}2\pi \delta(\Omega + \Omega_0)

s i n ( Ω 0 t ) C T F T π j ( δ ( Ω + Ω 0 ) δ ( Ω Ω 0 ) ) \color{red}sin(\Omega_0t)\xrightarrow{CTFT}\pi j(\delta(\Omega+\Omega_0)-\delta(\Omega-\Omega_0))
(b)
U ( j Ω ) = e α t e j Ω t d t = 0 e α t e j Ω t d t + 0 e α t e j Ω t d t U(j\Omega)=\int_{-\infty}^{\infty}e^{-\alpha\vert t \vert}e^{-j\Omega t}dt=\int_{-\infty}^{0}e^{\alpha t }e^{-j\Omega t}dt+\int_{0}^{\infty}e^{-\alpha t }e^{-j\Omega t}dt
0 e α t d t = 1 α e α t 0 = 1 α \int_{0}^{\infty}e^{-\alpha t}dt=\frac{1}{-\alpha}e^{-\alpha t}\vert_{0}^{\infty}=\frac{1}{\alpha}

U ( j Ω ) = 0 e α t e j Ω t d t + 0 e α t e j Ω t d t = 0 e α t e j Ω t d t + 0 e α t e j Ω t d t = 1 α j Ω + 1 α + j Ω = 2 α Ω 2 + α 2 U(j\Omega)=\int_{-\infty}^{0}e^{\alpha t }e^{-j\Omega t}dt+\int_{0}^{\infty}e^{-\alpha t }e^{-j\Omega t}dt=\int_{0}^{\infty}e^{-\alpha t}e^{j\Omega t}dt + \int_{0}^{\infty}e^{-\alpha t }e^{-j\Omega t}dt \\ =\frac{1}{\alpha-j\Omega}+\frac{1}{\alpha+j\Omega}=\frac{2\alpha}{\Omega^2+\alpha^2}

e α t C T F T 2 α Ω 2 + α 2 \color{red}e^{-\alpha \vert t \vert} \xrightarrow{CTFT}\frac{2\alpha}{\Omega^2+\alpha^2}
(c)
e j Ω 0 t C T F T 2 π δ ( Ω Ω 0 ) \color{red}e^{j\Omega_0t}\xrightarrow{CTFT}2\pi\delta(\Omega-\Omega_0)
(d)
   p a ( t ) p_a(t) 是周期为 l l 的周期函数,所以可以将其进行傅里叶级数展开,其主周期的傅里叶变换为
F ( j Ω ) = C T F T [ δ ( t ) ] = 1 F(j\Omega)=CTFT[\delta(t)]=1
故其傅里叶级数的系数为
c n = 1 T F ( j Ω ) Ω = n Ω 0 = 1 T = 1 l c_n=\frac{1}{T}F(j\Omega)\vert_{\Omega=n\Omega_0}=\frac{1}{T}=\frac{1}{l}

l = δ ( t l T ) = n = 1 l e j n Ω 0 t \sum_{l=-\infty}^{\infty}\delta(t-lT)=\sum_{n=-\infty}^{\infty}\frac{1}{l}e^{jn\Omega_0 t}
其中 Ω 0 = 2 π l \Omega_0=\frac{2\pi}{l} ,将上式两边同时进行傅里叶变换
C T F T [ l = δ ( t l T ) ] = 1 l n = 2 π δ ( Ω n Ω 0 ) = Ω 0 n = δ ( Ω n Ω 0 ) CTFT[\sum_{l=-\infty}^{\infty}\delta(t-lT)]=\frac{1}{l}\sum_{n=-\infty}^{\infty}2\pi\delta(\Omega - n\Omega_0)=\Omega_0\sum_{n=-\infty}^{\infty}\delta(\Omega - n\Omega_0)

l = δ ( t l T ) C T F T Ω 0 n = δ ( Ω n Ω 0 ) ,   Ω 0 = 2 π l \color{red}\sum_{l=-\infty}^{\infty}\delta(t-lT)\xrightarrow{CTFT}\Omega_0\sum_{n=-\infty}^{\infty}\delta(\Omega - n\Omega_0),\, \Omega_0=\frac{2\pi}{l}
(e)
e a t 2 e j Ω t d t = e a ( t 2 + 2 j Ω 2 a t + ( j Ω 2 a ) 2 ) + a ( j Ω 2 a ) 2 d t = e Ω 2 4 a e a ( t + j Ω 2 a ) 2 d t \int_{-\infty}^{\infty}e^{-at^2}e^{-j\Omega t}dt=\int_{-\infty}^{\infty}e^{-a(t^2+2\frac{j\Omega}{2a}t+(\frac{j\Omega}{2a})^2)+a(\frac{j\Omega}{2a})^2}dt \\ \\ =e^{-\frac{\Omega^2}{4a}}\int_{-\infty}^{\infty}e^{-a(t+\frac{j\Omega}{2a})^2}dt
考虑一高斯分布为
1 2 π 1 2 a e a ( t + j Ω 2 a ) 2 a π e a ( t + j Ω 2 a ) 2 d t = 1 e a ( t + j Ω 2 a ) 2 d t = π a \frac{1}{\sqrt{2\pi}\frac{1}{\sqrt{2a}}}e^{-a(t+\frac{j\Omega}{2a})^2} \Rightarrow\int_{-\infty}^{\infty}\frac{\sqrt{a}}{\sqrt{\pi}}e^{-a(t+\frac{j\Omega}{2a})^2}dt = 1 \Rightarrow\int_{-\infty}^{\infty}e^{-a(t+\frac{j\Omega}{2a})^2}dt=\frac{\sqrt{\pi}}{\sqrt{a}}
所以
e a t 2 e j Ω t d t = π a e Ω 2 4 a \int_{-\infty}^{\infty}e^{-at^2}e^{-j\Omega t}dt=\sqrt{\frac{\pi}{a}}e^{-\frac{\Omega^2}{4a}}
所以
e a t 2 C T F T π a e Ω 2 4 a \color{red}e^{-at^2} \xrightarrow{CTFT}\sqrt{\frac{\pi}{a}}e^{-\frac{\Omega^2}{4a}}


求定义在 < t < -\infty < t <\infty 的下列连续时间函数的CTFT:
(a) v a ( t ) = 1 v_a(t)=1
(b) μ a ( t ) = { 1 , t 0 0 , t < 0 \mu_a(t)= \begin{cases} 1, \quad t \geq 0 \\ 0, \quad t < 0 \end{cases}
(c) x a ( t ) = { 1 , t < 1 2 1 2 , t = 1 2 0 , t > 1 2 x_a(t)= \begin{cases} 1, \quad &\vert t \vert < \frac{1}{2}\\ \frac{1}{2}, \quad &\vert t \vert = \frac{1}{2} \\ 0, \quad &\vert t \vert > \frac{1}{2} \end{cases}
(d) y a ( t ) = { 1 2 t ,   t < 1 2 0 ,   t 1 2 y_a(t)= \begin{cases} 1 - 2\vert t \vert, \, &\vert t \vert < \frac{1}{2} \\ 0, \, &\vert t \vert \geq \frac{1}{2} \end{cases}
解:
(a)
由于 1 1 既不绝对可和,也不平方可和,所以不能从定义直接得到,不过可以考虑
1 = lim α 0 e α t 1=\lim\limits_{\alpha \to 0}e^{-\alpha \vert t \vert}
由于 e α t C T F T 2 α Ω 2 + α 2 e^{-\alpha \vert t \vert}\xrightarrow{CTFT}\frac{2\alpha}{\Omega^2+\alpha^2}
所以
lim α 0 2 α Ω 2 + α 2 = { ,   Ω = 0 0 ,   Ω ̸ = 0 \lim\limits_{\alpha \to 0}\frac{2\alpha}{\Omega^2+\alpha^2}= \begin{cases} \infty, \, &\Omega = 0 \\ 0, \, &\Omega \not = 0 \end{cases}
并且
lim α 0 2 α Ω 2 + α 2 d Ω = lim α 0 2 1 1 + Ω 2 α 2 d ( Ω α ) = lim α 0 2 a r c t a n ( Ω α ) = 2 π \lim\limits_{\alpha \to 0}\int_{-\infty}^{\infty}\frac{2\alpha}{\Omega^2+\alpha^2}d\Omega=\lim\limits_{\alpha \to 0}2\int_{\infty}^{\infty}\frac{1}{1+\frac{\Omega^2}{\alpha^2}}d(\frac{\Omega}{\alpha})=\lim\limits_{\alpha \to 0}2arctan(\frac{\Omega}{\alpha})\vert_{-\infty}^{\infty}=2\pi
所以
1 C T F T 2 π δ ( Ω ) \color{red}1\xrightarrow{CTFT}2\pi\delta(\Omega)
(b)
μ ( t ) \mu(t) 既不是平方可和,也不是平方可和的,所以还得用别的办法:
μ ( t ) = 1 2 ( 1 + s g n ( t ) ) \mu(t)=\frac{1}{2}(1+sgn(t))
C T F T [ μ ( t ) ] = C T F T [ 1 2 ( 1 + s g n ( t ) ) ] = π δ ( Ω ) + 1 j Ω CTFT[\mu(t)]=CTFT[\frac{1}{2}(1+sgn(t))]=\pi\delta(\Omega)+\frac{1}{j\Omega}
(c)
这道题暂时没想到怎么做
(d)
该函数为 Λ ( 2 t ) \Lambda(2t) ,由于
Λ ( t ) C T F T ( s i n ( Ω 2 ) Ω 2 ) 2 = S a 2 ( Ω 2 ) \Lambda(t)\xrightarrow{CTFT}(\frac{sin(\frac{\Omega}{2})}{\frac{\Omega}{2}})^2=Sa^2(\frac{\Omega}{2})
所以
Λ ( 2 t ) C T F T 1 4 S a 2 ( Ω 4 ) \Lambda(2t)\xrightarrow{CTFT}\frac{1}{4}Sa^2(\frac{\Omega}{4})


为了方便,式定义的高斯密度函数重写如下:
h ( t ) = 1 σ 2 π e ( t μ ) 2 2 σ 2 h(t)=\frac{1}{\sigma\sqrt{2\pi}}e^{-\frac{(t-\mu)^2}{2\sigma^2}}
其中, σ \sigma μ \mu 分别是该密度函数的方差和均值。具有上式给出的零均值的冲激响应的连续时间的滤波器称为高斯滤波器。证明 h ( t ) h(t) C T F T CTFT 也是 Ω \Omega 的高斯函数。
解:考虑零均值的高斯函数
h ( t ) = 1 σ 2 π e t 2 2 σ 2 h(t)=\frac{1}{\sigma\sqrt{2\pi}}e^{-\frac{t^2}{2\sigma^2}}
a = 1 2 σ 2 a=\frac{1}{2\sigma^2} ,则
h ( t ) = 1 σ 2 π e a t 2 h(t)=\frac{1}{\sigma\sqrt{2\pi}}e^{-at^{2}}
之前已经证明
e a t 2 C T F T π a e Ω 2 4 a e^{-at^2} \xrightarrow{CTFT}\sqrt{\frac{\pi}{a}}e^{-\frac{\Omega^2}{4a}}

C T F T [ h ( t ) ] = 1 σ 2 π π a e Ω 2 4 a = e σ 2 Ω 2 2 CTFT[h(t)]=\frac{1}{\sigma\sqrt{2\pi}}\sqrt{\frac{\pi}{a}}e^{-\frac{\Omega^2}{4a}}=e^{-\frac{\sigma^2\Omega^2}{2}}
服从高斯分布。


有限能量的函数 x a ( t ) = s i n ( t ) / π t x_a(t)=sin(t)/\pi t 不是绝对可和的。证明其 C T F T CTFT
X a ( j Ω ) = { 1 ,   Ω 1 0 ,   Ω > 1 X_a(j\Omega)= \begin{cases} 1, \, & \vert \Omega \vert \leq 1 \\ 0, \, & \vert \Omega \vert > 1 \end{cases}
解:
X a ( j Ω ) = x a ( t ) e j Ω t d t = s i n ( t ) π t e j Ω t d t = s i n ( t ) π t ( c o s ( Ω t ) j s i n ( Ω t ) ) d t = s i n ( t ) c o s ( Ω t ) π t d t = 0 s i n ( Ω + 1 ) t + s i n ( Ω 1 ) t π t d t = Ω + 1 π 0 s i n ( Ω + 1 ) t ( Ω + 1 ) t Ω 1 π 0 s i n ( Ω 1 ) t ( Ω 1 ) t = Ω + 1 2 Ω + 1 Ω 1 2 Ω 1 \begin{aligned} X_a(j\Omega)&=\int_{-\infty}^{\infty}x_a(t)e^{-j\Omega t}dt=\int_{-\infty}^{\infty}\frac{sin(t)}{\pi t}e^{-j\Omega t}dt \\ &=\int_{-\infty}^{\infty}\frac{sin(t)}{\pi t}(cos(\Omega t)-jsin(\Omega t))dt \\ &=\int_{-\infty}^{\infty}\frac{sin(t)cos(\Omega t)}{\pi t}dt =\int_{0}^{\infty}\frac{sin(\Omega + 1)t + sin(\Omega - 1)t}{\pi t}dt\\ &=\frac{\Omega+1}{\pi}\int_{0}^{\infty}\frac{sin(\Omega + 1)t}{(\Omega+1)t}-\frac{\Omega-1}{\pi}\int_{0}^{\infty}\frac{sin(\Omega - 1)t}{(\Omega-1)t} \\ &=\frac{\Omega+1}{2\vert \Omega + 1\vert} -\frac{\Omega-1}{2\vert \Omega - 1\vert} \end{aligned}
容易验证
X ( j Ω ) = { 1 ,   Ω 1 0 ,   Ω > 1 X(j\Omega)= \begin{cases} 1, \, &\vert \Omega \vert \leq 1 \\ 0, \, &\vert \Omega \vert > 1 \end{cases}


考虑 C T F T CTFT
x a ( t ) C T F T X a ( j Ω ) x_a(t)\xleftrightarrow{CTFT}X_a(j\Omega)
证明下面的定理:
(a)时移定理: x a ( t t 0 ) C T F T X a ( j Ω ) e j Ω t 0 x_a(t-t_0)\xleftrightarrow{CTFT}X_a(j\Omega)e^{-j\Omega t_0}
(b)频移定理: x a ( t ) e j Ω 0 t C T F T X a ( j ( Ω Ω 0 ) ) x_a(t)e^{j\Omega_0t}\xleftrightarrow{CTFT}X_a(j(\Omega-\Omega_0))
(c)对称定理: X a ( t ) C T F T 2 π x a ( j Ω ) X_a(t)\xleftrightarrow{CTFT}2\pi x_a(-j\Omega)
(d)尺度缩放定理: x a ( a t ) C T F T 1 a X a ( j Ω a ) x_a(at)\xleftrightarrow{CTFT}\frac{1}{\vert a\vert}X_a(j\frac{\Omega}{a})
(e)时间微分定理: d x a ( t ) d t C T F T j Ω X a ( j Ω ) \frac{dx_a(t)}{dt}\xleftrightarrow{CTFT}j\Omega X_a(j\Omega)
解:
(a)
x a ( t t 0 ) e j Ω t d t m = t t 0 x a ( m ) e j Ω m e j Ω t 0 d t = X a ( j Ω ) e j Ω t 0 \int_{-\infty}^{\infty}x_a(t-t_0)e^{-j\Omega t}dt\xrightarrow{m=t-t_0}\int_{-\infty}^{\infty}x_a(m)e^{-j \Omega m}e^{-j \Omega t_0}dt=X_a(j\Omega)e^{-j\Omega t_0}
(b)
x a ( t ) e j Ω 0 t e j Ω t d t = x a ( t ) e j ( Ω Ω 0 ) t d t = X a ( j ( Ω Ω 0 ) ) \int_{-\infty}^{\infty}x_a(t)e^{j\Omega_0t}e^{-j\Omega t}dt=\int_{-\infty}^{\infty}x_a(t)e^{-j(\Omega-\Omega_0)t}dt=X_a(j(\Omega-\Omega_0))
(c)
x a ( t ) = 1 2 π X ( j Ω ) e j Ω t d Ω X ( j Ω ) e j Ω t d Ω = 2 π x a ( t ) x_a(t)=\frac{1}{2\pi}\int_{-\infty}^{\infty}X(j\Omega)e^{j\Omega t}d\Omega\Rightarrow\int_{-\infty}^{\infty}X(j\Omega)e^{j\Omega t}d\Omega=2\pi x_a(t)
X ( t ) e j Ω t d t = X ( t ) e j ( Ω ) t d t = 2 π x a ( j Ω ) \int_{-\infty}^{\infty}X(t)e^{-j\Omega t}dt=\int_{-\infty}^{\infty}X(t)e^{j(-\Omega) t}dt=2\pi x_a(-j \Omega)
(d)
x a ( a t ) e j Ω t d t \int_{-\infty}^{\infty}x_a(at)e^{-j\Omega t}dt
a > 0 , l = a t a>0,令l=at
x a ( l ) e j Ω a l d ( l a ) = 1 a x a ( l ) e j Ω a l d l = 1 a X a ( j Ω a ) \int_{-\infty}^{\infty}x_a(l)e^{-j\frac{\Omega}{a} l}d(\frac{l}{a})=\frac{1}{a}\int_{-\infty}^{\infty}x_a(l)e^{-j\frac{\Omega}{a} l}dl=\frac{1}{a}X_a(\frac{j \Omega}{a})
a < 0 , l = a t a<0,令l=at
x a ( l ) e j Ω a l d ( l a ) = 1 a x a ( l ) e j Ω a l d l = 1 a X a ( j Ω a ) \int_{\infty}^{-\infty}x_a(l)e^{-j\frac{\Omega}{a} l}d(\frac{l}{a})=-\frac{1}{a}\int_{-\infty}^{\infty}x_a(l)e^{-j\frac{\Omega}{a} l}dl=-\frac{1}{a}X_a(\frac{j \Omega}{a})
注意这里因为 a < 0 a<0 ,所以变量替换的时候积分上下限的正负性发生变化
综上
x a ( a t ) C T F T 1 a X a ( j Ω a ) x_a(at)\xleftrightarrow{CTFT}\frac{1}{\vert a \vert}X_a(\frac{j\Omega}{a})
(e)
x a ( t ) = 1 2 π X ( j Ω ) e j Ω t d Ω x_a(t)=\frac{1}{2\pi}\int_{-\infty}^{\infty}X(j\Omega)e^{j\Omega t}d\Omega
左右两边同时对 t t 进行积分
d x a ( t ) d t = 1 2 π j Ω X ( j Ω ) e j Ω t d Ω \frac{dx_a(t)}{dt}=\frac{1}{2\pi}\int_{-\infty}^{\infty}j\Omega X(j\Omega)e^{j\Omega t}d\Omega
所以
d x a ( t ) d t C T F T j Ω X a ( j Ω ) \frac{dx_a(t)}{dt}\xleftrightarrow{CTFT}j\Omega X_a(j\Omega)


X a ( j Ω ) X_a(j\Omega) 表示实值连续时间函数 x a ( t ) x_a(t) C T F T CTFT 。证明其幅度谱 X a ( j Ω ) \vert X_a(j\Omega) \vert Ω \Omega 的偶函数,而相位谱 θ ( Ω ) = a r g { X a ( j Ω ) } \theta(\Omega)=arg\{X_a(j\Omega)\} Ω \Omega 的奇函数。
解:由于 x a ( t ) x_a(t) 为实值函数,所以 x a ( t ) = x a ( t ) x_a(t)=x^{*}_a(t)
x a ( t ) e j Ω t d t = ( x a ( t ) e j ( Ω ) t d t ) = X a ( j Ω ) \int_{\infty}^{-\infty}x^{*}_a(t)e^{-j\Omega t}dt=(\int_{\infty}^{-\infty}x_a(t)e^{-j(-\Omega) t}dt)^{*}=X^{*}_a(-j\Omega)
X a ( j Ω ) = X a ( j Ω ) X a ( j Ω ) = X a ( j Ω ) \Rightarrow X_a(j\Omega)=X^{*}_a(-j\Omega) \Rightarrow X_a(-j\Omega)=X^{*}_a(j\Omega)

X a ( j Ω ) = X a ( j Ω ) = X ( j Ω ) \vert X_a(j\Omega) \vert = \vert X^{*}_a(j\Omega)\vert = \vert X(-j\Omega) \vert
a r g { X a ( j Ω ) } = a r g { X a ( j Ω ) } = a r g { X a ( j Ω ) } arg\{X_a(j\Omega)\}=-arg\{X^{*}_a(j\Omega)\}=-arg\{X_a(-j\Omega)\}
所以其幅度谱为偶函数,其相位谱为奇函数。


证明式
h H T ( t ) = 1 π t h_{HT}(t)=\frac{1}{\pi t}
定义的希尔伯特变换的 C T F T CTFT
H H T ( j Ω ) = { j ,   Ω > 0 j ,   Ω < 0 H_{HT}(j\Omega)= \begin{cases} -j, \, &\Omega > 0 \\ j, \, &\Omega < 0 \end{cases}
解:
H H T ( j Ω ) = 1 π t e j Ω t d t = 1 π c o s Ω t j s i n Ω t t d t = 2 j Ω π 0 s i n Ω t Ω t d t = j Ω Ω = { j ,   Ω > 0 j ,   Ω < 0 \begin{aligned} H_{HT}(j\Omega)&=\int_{-\infty}^{\infty}\frac{1}{\pi t}e^{-j\Omega t}dt \\ &=\frac{1}{\pi}\int_{-\infty}^{\infty}\frac{cos\Omega t -jsin\Omega t}{t}dt \\ &=\frac{-2j\Omega}{\pi}\int_{0}^{\infty}\frac{sin\Omega t}{\Omega t}dt \\ &=-j\frac{\Omega}{\vert \Omega \vert} \\ &=\begin{cases} -j, \, &\Omega > 0 \\ j, \, &\Omega < 0 \end{cases} \end{aligned}
这里使用了一个结论
0 s i n Ω t Ω t d t = π 2 Ω \int_{0}^{\infty}\frac{sin\Omega t}{\Omega t}dt=\frac{\pi}{2\vert \Omega \vert}


x ( t ) x(t) 是实值输入信号,其 C T F T CTFT X ( j Ω ) = X p ( j Ω ) + X n ( j Ω ) X(j\Omega)=X_p(j\Omega)+X_n(j\Omega) ,其中 X p ( j Ω ) X_p(j\Omega) 是占据 X ( j Ω ) X(j\Omega) 正频率范围的分量, X n ( j Ω ) X_n(j\Omega) 是占据 X ( j Ω ) X(j\Omega) 负频率范围的分量。令 x ^ ( t ) \hat{x}(t) 表示 x ( t ) x(t) 的希尔伯特变换。证明:复值信号 y ( t ) = x ( t ) + j x ^ ( t ) y(t)=x(t)+j\hat{x}(t) C T F T   Y ( j Ω ) CTFT\, Y(j\Omega) Y ( j Ω ) = 2 X p ( j Ω ) Y(j\Omega)=2X_p(j\Omega) ,即 y ( t ) y(t) 的谱只包含正频率范围的分量。
解:
由希尔伯特变换的定义
X ^ ( j Ω ) = j X p ( j Ω ) + j X n ( j Ω ) \hat{X}(j\Omega)=-jX_p(j\Omega)+jX_n(j\Omega)
则信号 y ( t ) = x ( t ) + j x ^ ( t ) y(t)=x(t)+j\hat{x}(t) 的傅里叶变换为
Y ( j Ω ) = X ( j Ω ) + j X ^ ( j Ω ) = X p ( j Ω ) + X n ( j Ω ) + j ( j X p ( j Ω ) + j X n ( j Ω ) ) = 2 X p ( j Ω ) \begin{aligned} Y(j\Omega)&=X(j\Omega)+j\hat{X}(j\Omega)\\ &=X_p(j\Omega)+X_n(j\Omega)+j(-jX_p(j\Omega)+jX_n(j\Omega))\\ &=2X_p(j\Omega) \end{aligned}


计算式
x a ( t ) = { e α t ,   t 0 0 ,   t < 0 x_a(t)= \begin{cases} e^{-\alpha t}, \, &t \geq 0 \\ 0, \, & t < 0 \end{cases}
中连续时间信号在 α = 0.6 \alpha=0.6 时的总能量,并计算其 75 % 75\% 带宽。
解:
ε 2 = x 2 ( t ) d t = e 2 α t d t = 1 2 α = 5 6 \varepsilon^2=\int_{-\infty}^\infty x^2(t)dt=\int_{-\infty}^\infty e^{-2\alpha t}dt = \frac{1}{2\alpha}=\frac{5}{6}
X ( j Ω ) = 1 α j Ω X(j\Omega)=\frac{1}{\alpha-j\Omega}

1 2 π Ω c Ω c X ( j Ω ) X ( j Ω ) d Ω = 1 2 π Ω c Ω c 1 α 2 + Ω 2 d Ω = 0.75 1 2 α = 3 8 α \frac{1}{2\pi}\int_{-\Omega_c}^{\Omega_c}X(j\Omega)X^{*}(j\Omega)d\Omega=\frac{1}{2\pi}\int_{-\Omega_c}^{\Omega_c}\frac{1}{\alpha^2 + \Omega^2}d\Omega=0.75 \cdot \frac{1}{2\alpha}=\frac{3}{8\alpha}

1 2 π Ω c Ω c 1 α 2 + Ω 2 d Ω = 1 2 π Ω c Ω c 1 α 1 1 + ( Ω α ) 2 d Ω α = 1 α π a r c t a n ( Ω c α ) \frac{1}{2\pi}\int_{-\Omega_c}^{\Omega_c}\frac{1}{\alpha^2 + \Omega^2}d\Omega=\frac{1}{2\pi}\int_{-\Omega_c}^{\Omega_c}\frac{1}{\alpha}\frac{1}{1+(\frac{\Omega}{\alpha})^2}d\frac{\Omega}{\alpha}=\frac{1}{\alpha\pi}arctan(\frac{\Omega_c}{\alpha})
a r c t a n ( Ω c α ) = 3 π 8 \Rightarrow arctan(\frac{\Omega_c}{\alpha})=\frac{3\pi}{8}


DTFT

证明 μ [ n ] \mu[n] D T F T DTFT X ( e j w ) = 1 1 e j w + k = π δ ( w + 2 π k ) X(e^{jw})=\frac{1}{1-e^{-jw}}+\sum_{k=-\infty}^{\infty}\pi\delta(w+2\pi k)
解:
  如果直接按照定义去求的话,会发现是无穷级数不是收敛的,所以该级数一定有狄拉克函数的形式。
  将 μ [ n ] \mu[n] 分为奇偶两部分,则
x e v [ n ] = 1 2 ( u [ n ] + u [ n ] ) = 1 2 + 1 2 δ [ n ] x_{ev}[n]=\frac{1}{2}(u[n]+u[-n])=\frac{1}{2}+\frac{1}{2}\delta[n]
x o d [ n ] = 1 2 ( u [ n ] u [ n ] ) = 1 2 ( 2 μ [ n ] ( μ [ n ] + μ [ n ] ) ) = μ [ n ] 1 2 1 2 δ [ n ] x_{od}[n]=\frac{1}{2}(u[n]-u[-n])=\frac{1}{2}(2\mu[n]-(\mu[n]+\mu[-n]))=\mu[n]-\frac{1}{2}-\frac{1}{2}\delta[n]
则偶数部分的DTFT为
X e v ( e j w ) = k = π δ ( w + 2 π k ) + 1 2 X_{ev}(e^{jw})=\sum_{k=-\infty}^{\infty}\pi\delta(w+2\pi k)+\frac{1}{2}
因为奇数部分中含 μ [ n ] \mu[n] ,稍微变一下形
x o d [ n ] x o d [ n 1 ] = 1 2 ( δ [ n ] + δ [ n 1 ] ) ( 1 e j w ) X o d ( e j w ) = 1 2 ( 1 + e j w ) X o d ( e j w ) = 1 2 1 + e j w 1 e j w = 1 2 + 1 1 e j w x_{od}[n]-x_{od}[n-1]=\frac{1}{2}(\delta[n]+\delta[n-1]) \\ \Rightarrow (1-e^{-jw})X_{od}(e^{-jw})=\frac{1}{2}(1+e^{-jw}) \\ \Rightarrow X_{od}(e^{jw})=\frac{1}{2}\frac{1+e^{-jw}}{1-e^{-jw}}=-\frac{1}{2}+\frac{1}{1-e^{-jw}}

X ( e j w ) = X e v ( e j w ) + X o d ( e j w ) = 1 1 e j w + k = π δ ( w + 2 π k ) \color{red}X(e^{jw})=X_{ev}(e^{jw})+X_{od}(e^{jw})=\frac{1}{1-e^{-jw}}+\sum_{k=-\infty}^{\infty}\pi\delta(w+2\pi k)


证明序列 x [ n ] = 1 x[n]=1 D T F T DTFT X ( e j w ) = k = 2 π δ ( w + 2 π k ) ,   < k < X(e^{jw})=\sum_{k=-\infty}^{\infty}2\pi\delta(w+2\pi k),\, -\infty < k < \infty
解:
  求 X ( e j w ) X(e^{jw}) 的反变换
x [ n ] = 1 2 π π π X ( e j w ) e j w t d w = π π k = δ ( w + 2 π k ) d w = 1 x[n]=\frac{1}{2\pi}\int_{-\pi}^\pi X(e^{jw})e^{jwt}dw=\int_{-\pi}^{\pi}\sum_{k=-\infty}^{\infty}\delta(w+2\pi k)dw=1
所以
1 D T F T k = 2 π δ ( w + 2 π k ) ,   < k < \color{red}1\xrightarrow{DTFT}\sum_{k=-\infty}^{\infty}2\pi\delta(w+2\pi k),\, -\infty < k < \infty


求双边序列 y [ n ] = α n , α < 1 y[n]=\alpha^{\vert n \vert},\vert \alpha \vert < 1 D T F T DTFT
解:
X ( e j w ) = n = α n e j w n = 1 α n e j w n + n = 0 α n e j w n = m = 1 α m e j w m + n = 0 α n e j w n = α e j w 1 α e j w + 1 1 α e j w \begin{aligned}X(e^{jw})&=\sum_{n=-\infty}^{\infty}\alpha^{\vert n \vert}e^{-jwn}\\ &=\sum_{-\infty}^{-1}\alpha^{-n}e^{-jwn}+\sum_{n=0}^{\infty}\alpha^{n}e^{-jwn}\\ &=\sum_{m=1}^{\infty}\alpha^{m}e^{jwm}+\sum_{n=0}^{\infty}\alpha^{n}e^{-jwn}\\ &=\frac{\alpha e^{jw}}{1-\alpha e^{jw}}+\frac{1}{1-\alpha e^{-jw}} \end{aligned}


X ( e j w ) X(e^{jw}) 表示实序列 x [ n ] x[n] D T F T DTFT
(a)证明:若 x [ n ] x[n] 是偶序列,则它可以用 x [ n ] = 1 π 0 π X ( e j w ) c o s ( n w ) d w x[n]=\frac{1}{\pi}\int_{0}^{\pi}X(e^{jw})cos(nw)dw X ( e j w ) X(e^{jw}) 计算。
(b)证明:若 x [ n ] x[n] 是奇序列,则它可以用 x [ n ] = j π 0 π X ( e j w ) s i n ( n w ) d w x[n]=\frac{j}{\pi}\int_{0}^{\pi}X(e^{jw})sin(nw)dw X ( e j w ) X(e^{jw}) 计算。
解:
D T F T [ x [ n ] ] = n = x [ n ] e j w n m = n m = x [ m ] e j ( w ) m = X ( e j w ) DTFT[x[-n]]=\sum_{n=-\infty}^{\infty}x[-n]e^{-jwn}\xrightarrow{m=-n}\sum_{m=-\infty}^{\infty}x[m]e^{-j(-w)m}=X(e^{-jw})

x [ n ] D T F T X ( e j w ) \color{red}x[-n]\xrightarrow{DTFT}X(e^{-jw})
x [ n ] = 1 2 π π π X ( e j w ) e j w n d w = 1 2 π π π X ( e j w ) ( c o s ( w n ) + j s i n ( w n ) ) d w = 1 2 π π π X ( e j w ) c o s ( w n ) d w + j 2 π π π X ( e j w ) s i n ( w n ) d w \begin{aligned} x[n]&=\frac{1}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})e^{jwn}dw \\ &=\frac{1}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})(cos(wn)+jsin(wn))dw\\ &=\frac{1}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})cos(wn)dw+\frac{j}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})sin(wn)dw \end{aligned}
(a)
由于 x [ n ] x[n] 是实序列且为偶序列,所以 x [ n ] = x [ n ] x[n]=x[-n] ,所以 X ( e j w ) = X ( e j w ) X(e^{jw})=X(e^{-jw}) ,即 X ( e j w ) X(e^{jw}) 为偶函数,所以 X ( e j w ) c o s ( w n ) X(e^{jw})cos(wn) 为偶函数, X ( e j w ) s i n ( w n ) X(e^{jw})sin(wn) 为奇函数,所以
x [ n ] = 1 2 π π π X ( e j w ) c o s ( w n ) d w + j 2 π π π X ( e j w ) s i n ( w n ) d w = 1 π 0 π X ( e j w ) c o s ( w n ) d w \begin{aligned} x[n]&=\frac{1}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})cos(wn)dw+\frac{j}{2\pi}\int_{-\pi}^{\pi}X(e^{jw})sin(wn)dw\\ &=\frac{1}{\pi}\int_{0}^{\pi}X(e^{jw})cos(wn)dw \end{aligned}
(b)同理,仿照上面即可得出
x [ n ] = j π 0 π X ( e j w ) s i n ( w n ) d w x[n]=\frac{j}{\pi}\int_{0}^{\pi}X(e^{jw})sin(wn)dw


求因果序列 x [ n ] = A α n c o s ( w 0 n + ϕ ) μ [ n ] x[n]=A\alpha^ncos(w_0n+\phi)\mu[n] D T F T DTFT ,其中 A α w 0 A、\alpha、w_0 ϕ \phi 是实数, α < 1 \vert \alpha \vert <1
解:
x [ n ] = A α n c o s ( w 0 n + ϕ ) μ [ n ] = A 2 α n ( e j ( w 0 n + ϕ ) + e j ( w 0 n + ϕ ) ) μ [ n ] = A 2 α n e j ϕ e j w 0 n μ [ n ] + A 2 α n e j ϕ e j w 0 n μ [ n ] x[n]=A\alpha^{n}cos(w_0n+\phi)\mu[n]=\frac{A}{2}\alpha^n(e^{j(w_0n+\phi)}+e^{-j(w_0n+\phi)})\mu[n] \\ =\frac{A}{2}\alpha^ne^{j\phi}e^{jw_0n}\mu[n]+\frac{A}{2}\alpha^ne^{-j\phi}e^{-jw_0n}\mu[n]
α n μ [ n ] D T F T 1 1 α e j w \alpha^n\mu[n]\xrightarrow{DTFT}\frac{1}{1-\alpha e^{-jw}}
所以
α n e j w 0 n μ [ n ] D T F T 1 1 α e j ( w w 0 ) \alpha^ne^{jw_0n}\mu[n]\xrightarrow{DTFT}\frac{1}{1-\alpha e^{-j(w-w_0)}}
α n e j w 0 n μ [ n ] D T F T 1 1 α e j ( w + w 0 ) \alpha^ne^{-jw_0n}\mu[n]\xrightarrow{DTFT}\frac{1}{1-\alpha e^{-j(w+w_0)}}
所以
x [ n ] D T F T A 2 e j ϕ 1 1 α e j ( w w 0 ) + A 2 e j ϕ 1 1 α e j ( w + w 0 ) x[n]\xrightarrow{DTFT}\frac{A}{2}e^{j\phi}\frac{1}{1-\alpha e^{-j(w-w_0)}}+\frac{A}{2}e^{-j\phi}\frac{1}{1-\alpha e^{-j(w+w_0)}}


求下面每个序列的 D T F T DTFT :
(a) x 1 [ n ] = α n μ [ n 1 ] , α < 1 x_1[n]=\alpha^n\mu[n-1],\vert \alpha \vert<1
(b) x 2 [ n ] = n α n μ [ n ] , α < 1 x_2[n]=n\alpha^n\mu[n],\vert \alpha \vert<1
(c) x 3 [ n ] = α n μ [ n + 1 ] , α < 1 x_3[n]=\alpha^n\mu[n+1],\vert \alpha \vert<1
(d) x 4 [ n ] = n α n μ [ n + 2 ] , α < 1 x_4[n]=n\alpha^n\mu[n+2],\vert \alpha \vert<1
(e) x 5 [ n ] = α n μ [ n 1 ] , α > 1 x_5[n]=\alpha^n\mu[-n-1],\vert \alpha \vert>1
(f) x 6 [ n ] = { α n ,   n M 0 ,   x_6[n]= \begin{cases} \alpha^{\vert n \vert}, \, &\vert n \vert \leq M \\ 0, \, &其他 \end{cases}

猜你喜欢

转载自blog.csdn.net/The_last_knight/article/details/84303962