傅里叶变换的基本性质
1. 对称性
若\(F(\omega)=\mathscr{F}[f(t)]\),那么\(\mathscr{F}[F(t)]=2\pi f(-\omega)\)
证明:
\[\begin{aligned}
f(t)&=\frac{1}{2\pi}\int_{-\infty}^{\infty}F(\omega)e^{j\omega t}d\omega \\
f(-t)&=\frac{1}{2\pi}\int_{-\infty}^{\infty}F(\omega)e^{-j\omega t}d\omega \\
2\pi f(-\omega)&=\int_{-\infty}^{\infty}F(t)e^{-j\omega t}dt
\end{aligned} \tag{1}
\]
显然上式就是傅里叶正变换的定义形式。
2. 线性(叠加性)
若\(\mathscr{F}[f_{i}(t)]=F_{i}(\omega)\ (i=1,2,\cdots,n)\),则
\[\mathscr{F}[\sum_{i=1}^{n}a_{i}f_{i}(t)]=\sum_{i=1}^{n}a_{i}F_{i}(\omega) \tag{2}
\]
3. 奇偶虚实性
一般情况下,\(F(\omega)\)是复函数,因而可以把它表示成模与相位或者实部与虚部两部分,即
\[F(\omega)=|F(\omega)|e^{j\varphi(\omega)}=R(\omega)+jX(\omega)
\]
\[\left \{
\begin{aligned}
|F(\omega)| &= \sqrt{R^{2}(\omega)+X^{2}(\omega)} \\
\varphi(\omega) &= \text{arctan}[\frac{X(\omega}{R(\omega)}]
\end{aligned} \tag{3a}
\right.
\]
(1) \(f(t)\)是实函数
\[\begin{aligned}
F(\omega) &= \int_{-\infty}^{\infty}f(t)e^{-j\omega t}dt \\
&= \int_{-\infty}^{\infty}f(t)cos(\omega t)dt-j\int_{-\infty}^{\infty}f(t)sin(\omega t)dt
\end{aligned}
\]
在这种情况下,显然
\[\left \{
\begin{aligned}
R(\omega)&=\int_{-\infty}^{\infty}f(t)cos(\omega t)dt \\
X(\omega) &= -\int_{-\infty}^{\infty}f(t)sin(\omega t)dt
\end{aligned} \tag{3b}
\right.\]
此时,\(R(-\omega)=\int_{-\infty}^{\infty}f(t)cos(-\omega t)dt=R(\omega)\)为偶函数,\(X(-\omega)= \int_{-\infty}^{\infty}f(t)sin(-\omega t)dt=-X(\omega)\)为奇函数。在此基础上,根据式\((3a)\)可以得到\(|F(\omega)|\)为偶函数,\(\varphi(\omega)\)为奇函数。也就是说,实函数的傅里叶变换的幅度谱是偶函数,相位谱是奇函数。
(2) \(f(t)\)是虚函数
令\(f(t)=jg(t)\),则
\[\left \{
\begin{aligned}
R(\omega)&=\int_{-\infty}^{\infty}g(t)sin(\omega t)dt \\
X(\omega) &= \int_{-\infty}^{\infty}g(t)cos(\omega t)dt
\end{aligned}
\right.\]
在这种情况下,\(R(\omega)\)是奇函数,\(X(\omega)\)是偶函数,而\(|F(\omega)|\)仍然为偶函数,\(\varphi(\omega)\)仍然为奇函数。
4. 尺度变换特性
若\(\mathscr{F}[f(t)]=F(\omega)\),则
\[\mathscr{F}[f(at)]=\frac{1}{|a|}F(\frac{\omega}{a}) \tag{4}
\]
其中\(a\)为非零实常数。
证明:
\[\begin{aligned}
\mathscr{F}[f(at)]=\int_{-\infty}^{\infty}f(at)e^{-j\omega t}dt
\end{aligned}
\]
令\(x=at\)
当\(a>0\)
\[\begin{aligned}
\mathscr{F}[f(at)]&=\frac{1}{a}\int_{-\infty}^{\infty}f(at)e^{-j\omega\frac{x}{a}}dx \\
&=\frac{1}{a}F(\frac{\omega}{a})
\end{aligned}
\]
当\(a<0\)
\[\begin{aligned}
\mathscr{F}[f(at)]&=\frac{1}{a}\int_{+\infty}^{-\infty}f(at)e^{-j\omega\frac{x}{a}}dx \\
&=\frac{-1}{a}\int_{-\infty}^{\infty}f(x)e^{-j\omega\frac{x}{a}}dx \\
&=\frac{-1}{a}F(\frac{\omega}{a})
\end{aligned}
\]
综合上述两种情况,便可以证明式\((4)\)。
对于\(a=-1\)这种特殊情况,\(\mathscr{F}[f(-t)]=F(-\omega)\)。
由上可见,信号在时域中压缩(a>1)等效于在频域中扩展;反之,信号在时域中扩展(a<1)则等效于在频域中压缩。对于\(a=-1\)的情况,它说明信号在时域中反褶等效于在频域中也反褶。上述结论是不难理解的,因为信号的波形压缩\(a\)倍,信号随时间变化加快\(a\)倍,所以它所包含的频率分量增加\(a\)倍,也就是说频谱展宽\(a\)倍,根据能量守恒原理,各频率分量的大小必然减小\(a\)倍。
因为\(F(\omega)=\int_{-\infty}^{\infty}f(t)e^{-j\omega t}dt\),所以
\[F(0)=\int_{-\infty}^{\infty}f(t)dt \tag{4a}
\]
同样因为:\(f(t)=\frac{1}{2\pi}\int_{-\infty}^{\infty}F(\omega)e^{-j\omega t}d\omega\),所以
\[f(0)=\frac{1}{2\pi}\int_{-\infty}^{\infty}F(\omega)d\omega \tag{4b}
\]
上面两式分别说明\(f(t)\)与\(F(\omega)\)所覆盖的面积等于\(F(\omega)\)与\(2\pi f(t)\)在零点的数值\(F(0)\)与\(2\pi f(0)\)。
如果\(f(0)\)与F(0)各自等于\(f(t)\)与\(F(\omega)\)曲线的最大值,这时定义\(\tau\)和\(B\)分别为\(f(t)\)与\(F(\omega)\)的等效宽度,可写出如下关系式:
\[\begin{aligned}
f(0)\tau &= F(0) \\
F(0)B &= 2\pi f(0)
\end{aligned}
\]
从而可以得到
\[B=\frac{2\pi}{\tau} \tag{4c}
\]
可以看出,信号的等效脉冲宽度与占有的等效带宽成反比,若要压缩信号的持续时间,则不得不以展宽频带作为代价。所以通信系统中,通信速度和占用频带宽度是一对矛盾。
5. 时移特性
若\(\mathscr{F}[f(t)]=F(\omega)\),则
\[\mathscr{F}[f(t-t_{0})]=F(\omega)e^{-j\omega t_{0}} \tag{5}
\]
证明:
因为
\[\mathscr{F}[f(t-t_{0})]=\int_{-\infty}^{\infty}f(t-t_{0})e^{-j\omega t}dt
\]
令
\[x = t-t_{0}
\]
那么
\[\begin{aligned}
\mathscr{F}[f(t-t_{0})]&=\mathscr{F}[f(x)]=\int_{-\infty}^{\infty}f(x)e^{-j\omega(x+t_{0})}dx \\
&= e^{-j\omega t_{0}}\int_{-\infty}^{\infty}f(x)e^{-j\omega x}dx
\end{aligned}
\]
同理可得:
\[\mathscr{F}[f(t+t_{0})]=e^{j\omega t_{0}}\cdot F(\omega)
\]
也就是说,信号移动后,其幅度谱不变,而相位谱产生附件变换\((-\omega t_{0}/\omega t_{0})\)。
6. 频移特性
若\(\mathscr{F}[f(t)]=F(\omega)\),则,
\[\mathscr{F}[f(t)e^{j\omega_{0}t}]=F(\omega-\omega_{0}) \tag{6}
\]
证明:
因为:
\[\begin{aligned} \mathscr{F}[f(t)e^{j\omega_{0}t}]&=\int_{-\infty}^{\infty}f(t)e^{j\omega_{0}t}\cdot e^{-j\omega t}dt \\
&= \int_{-\infty}^{\infty}f(t)e^{-j(\omega-\omega_{0})t}dt\end{aligned}
\]
所以
\[\mathscr{F}[f(t)e^{j\omega_{0}t}]=F(\omega-\omega_{0})
\]
同理
\[\mathscr{F}[f(t)e^{-j\omega_{0}t}]=F(\omega+\omega_{0})
\]
可见,若时间信号\(f(t)\)乘以\(e^{j\omega_{0}t}\),等效于\(f(t)\)的频谱\(F(\omega)\)沿频率轴右移\(\omega_{0}\),或者说在频域中将频谱沿频率轴右移\(\omega_{0}\)等效于在时域中信号乘以因子\(e^{j\omega_{0}t}\)。
频谱搬移技术在通信系统中得到广泛应用,诸如调幅、同步解调、变频等过程都是在频谱搬移的基础上完成的。频谱搬移的实现原理是将信号\(f(t)\)乘以所谓载频信号\(cos(\omega_{0}t)/sin(\omega_{0}t)\)。
因为:
\[\begin{aligned}
cos(\omega_{0}t) &= \frac{1}{2}(e^{j\omega_{0}t}+e^{-j\omega_{0}t}) \\
sin(\omega_{0}t) &= \frac{1}{2j}(e^{j\omega_{0}t}-e^{-j\omega_{0}t})
\end{aligned}
\]
那么,可以导出:
\[\begin{aligned}
\mathscr{F}[f(t)cos(\omega_{0}t)] &= \frac{1}{2}[F(\omega+\omega_{0})+F(\omega-\omega_{0})] \\
\mathscr{F}[f(t)sin(\omega_{0}t)] &= \frac{1}{2}[F(\omega+\omega_{0})-F(\omega-\omega_{0})]
\end{aligned} \tag{6a}
\]
所以,若时间信号\(f(t)\)乘以\(cos(\omega_{0}t)\)或\(sin(\omega_{0}t)\),等效于\(f(t)\)的频谱\(F(\omega)\)一分为二,沿频率轴向左向右各平移\(\omega_{0}\)。
已知直流信号的频谱是位于\(\omega=0\)点的冲激函数,也即:
\[\mathscr{F}[1]=2\pi\delta(\omega)
\]
利用频移定理,可以得到:
\[\mathscr{F}[cos(\omega_{0}t)]=\pi[\delta(\omega+\omega_{0})+\delta(\omega-\omega_{0})]
\]
\[\mathscr{F}[sin(\omega_{0}t)]=j\pi[\delta(\omega+\omega_{0})-\delta(\omega-\omega_{0})]
\]
7. 微分特性
若\(\mathscr{F}[f(t)]=F(\omega)\),则
\[\begin{aligned}
\mathscr{F}[\frac{df(t)}{dt}] &= j\omega F(\omega) \\
\mathscr{F}[\frac{d^{n}f(t)}{dt^{n}}] &= (j\omega)^{n} F(\omega) \\
\end{aligned} \tag{7}
\]
证明:
因为:
\[f(t)=\frac{1}{2\pi}\int_{-\infty}^{\infty}F(\omega)e^{j\omega t}d\omega
\]
两边对t求导,得:
\[\frac{df(t)}{dt}=\frac{1}{2\pi}\int_{-\infty}^{\infty}[j\omega F(\omega)]e^{j\omega t}dt
\]
所以:
\[\mathscr{F}[\frac{df(t)}{dt}] = j\omega F(\omega)
\]
同理,可以推出:
\[\mathscr{F}[\frac{d^{n}f(t)}{dt^{n}}] = (j\omega)^{n} F(\omega)
\]
同理可以推出,频域的微分特性如下:
\[\begin{aligned}
\mathscr{F}^{-1}[\frac{dF(\omega)}{d\omega}] &= (-jt) f(t) \\
\mathscr{F}^{-1}[\frac{d^{n}F(\omega)}{d\omega^{n}}] &= (-jt)^{n} f(t) \\
\end{aligned} \tag{7a}
\]
8. 积分特性
若\(\mathscr{F}[f(t)]=F(\omega)\),则
\[\mathscr{F}[\int_{-\infty}^{t}f(\tau)d\tau]=\frac{F(\omega)}{j\omega}+\pi F(0)\delta(\omega) \tag{8}
\]
证明:
\[\begin{aligned}
\mathscr{F}[\int_{-\infty}^{t}f(\tau)d\tau]&=\int_{-\infty}^{\infty}[\int_{-\infty}^{t}f(\tau)d\tau]e^{-j\omega t}dt \\
&=\int_{-\infty}^{\infty}[\int_{-\infty}^{\infty}f(\tau)u(t-\tau)d\tau]e^{-j\omega t}dt
\end{aligned} \tag{8a}
\]
上式将被积函数\(f(t)\)乘以\(u(t-\tau)\),同时将积分上限\(t\)改为\(\infty\),结果不变。交换积分次序,并引用延时阶跃信号的傅里叶变换关系式,式\((8a)\)成为:
\[\begin{aligned}
&\quad \int_{-\infty}^{\infty}f(\tau)[\int_{-\infty}^{\infty}u(t-\tau)e^{-j\omega t}dt]d\tau \\
&=\int_{-\infty}^{\infty}f(\tau)\pi \delta(\omega)e^{-j\omega \tau}d\tau + \int_{-\infty}^{\infty}f(\tau)\frac{e^{-j\omega\tau}}{j\omega}d\tau\\
&=\pi F(0)\delta(\omega)+\frac{F(\omega)}{j\omega}
\end{aligned} \tag{8a}
\]
9. 卷积特性
(1)时域卷积定理
给定两个时间函数\(f_{1}(t), f_{2}(t)\),已知\(\mathscr{F}[f_{1}(t)]=F_{1}(\omega),\mathscr{F}[f_{2}(t)]=F_{2}(\omega)\),那么
\[\mathscr{F}[f_{1}(t)*f_{2}(t)]=F_{1}(\omega)F_{2}(\omega) \tag{9a}
\]
证明:
根据卷积的定义
\[f_{1}(t)*f_{2}(t)=\int_{-\infty}^{\infty}f_{1}(\tau)f_{2}(t-\tau)d\tau
\]
因此:
\[\begin{aligned}
\mathscr{F}[f_{1}(t)*f_{2}(t)]&=\int_{-\infty}^{\infty}[\int_{-\infty}^{\infty}f_{1}(\tau)f_{2}(t-\tau)d\tau]e^{-j\omega t}dt \\
&= \int_{-\infty}^{\infty}f_{1}(\tau)[\int_{-\infty}^{\infty}f_{2}(t-\tau)e^{-j\omega t}dt]d\tau \\
&= \int_{-\infty}^{\infty}f_{1}(\tau)F_{2}(\omega)e^{-j\omega\tau}d\tau \\
&= F_{2}(\omega)\int_{-\infty}^{\infty}f_{1}(\tau)e^{-j\omega\tau}d\tau \\
&= F_{1}(\omega)F_{2}(\omega)
\end{aligned}\]
(2)频域卷积定理
若
\[\begin{aligned}
\mathscr{F}[f_{1}(t)]=F_{1}(\omega) \\
\mathscr{F}[f_{2}(t)]=F_{2}(\omega) \\
\end{aligned}\]
那么
\[\mathscr{F}[f_{1}(t)\cdot f_{2}(t)]=\frac{1}{2\pi}F_{1}(\omega)*F_{2}(\omega) \tag{9b}
\]