Channel论文阅读笔记7-1之multipath interference by Jakes

1. 数学模型

接收机沿 $x$ 轴方向以速度 $v$ 运动，电磁波入射方向与接收机运动方向的夹角为${\alpha }_n$，则多普勒偏移为
$${\omega }_n=\beta v\cos {\alpha }_n\text{\hspace{1em}(1.1-1)}$$
这里 $\beta =2\pi /\lambda$， $\lambda$ 是射频波长。

若传输信号垂直极化，则电场和磁场各分量表示为
.$$E_z=E_0\sum _{n=1}^N{C}_n\cos ({\omega }_{c}t+{\theta }_n)\text{\hspace{1em}(1.1-2)}$$ $$H_x=-\frac{E_0}{\eta }\sum _{n=1}^N{C}_n\sin {\alpha }_n\cos ({\omega }_{c}t+{\theta }_n)\text{\hspace{1em}(1.1-3)}$$ $$H_x=\frac{E_0}{\eta }\sum _{n=1}^N{C}_n\cos {\alpha }_n\cos ({\omega }_{c}t+{\theta }_n)\text{\hspace{1em}(1.1-4)}$$
这里 $${\theta }_n={\omega }_{n}t+{\varphi }_n,\text{\hspace{1em}(1.1-5)}$$ ${\omega }_c$ 是射频信号角频率， ${\varphi }_n\in [0,2\pi )$ 是初始相位， $\eta$ 是自由空间波阻抗，$C_n$ 是归一化系数，满足 $<{\sum }_{n=1}^N{C}_n^2>$。

$E_z$ 可表示为
$$E_z=T_c(t)\cos {\omega }_{c}t-T_s(t)\sin {\omega }_{c}t\text{\hspace{1em}(1.1-7)}$$
这里
$$T_c(t)=E_0\sum _{n=1}^N{C}_n\cos ({\omega }_{n}t+{\varphi }_n)\text{\hspace{1em}(1.1-8)}$$ $$T_s(t)=E_0\sum _{n=1}^N{C}_n\sin ({\omega }_{n}t+{\varphi }_n)\text{\hspace{1em}(1.1-9)}$$ 是高斯随机过程，对应 $E_z$ 的同相 （in-phase）和正交（quadrature）分量。对于给定时间 $t$, $T_c$ 和 $T_s$ 具有相同的均值和方差 $$<T_c^2>=<T_s^2>=\frac{E_0}{2}=<|E_z{|}^2>\text{\hspace{1em}(1.1-10)}$$ $T_c$ 和 $T_s$ 不相关 $$<T_c{T}_s>=0\text{\hspace{1em}(1.1-11)}$$

$T_c$ 和 $T_s$ 的概率密度函数 $$p(x)=\frac{1}{\sqrt{2\pi b}}{e}^{-x^2/2b}\text{\hspace{1em}(1.1-12)}$$ 这里 $b=E_0^2/2$。

$E_z$ 的包络
$$r=(T_c^2+T_s^2{)}^{1/2}\text{\hspace{1em}(1.1-13)}$$ 概率密度函数 $$p(r)=\{\begin{array}{ll}\frac{r}{b}{e}^{-r^2/2b}& r\ge 0\\ 0& r<0\end{array}\text{\hspace{1em}(1.1-14)}$$

2. 功率谱

接收信号的频率 $$f(\alpha )=f_c+f_m\cos \alpha \text{\hspace{1em}(1.2-1)}$$ 这里 $f_m=v/\lambda$ 是最大多普勒频率。从所有方向到达信号的总功率应该与所有频率的总功率相同，另外考虑到 $f(\alpha )=f(-\alpha )$，因此 $$S(f)|df|=b[p(\alpha )G(\alpha )+p(-\alpha )G(-\alpha )]|d\alpha |\text{\hspace{1em}(1.2-2)}$$ 由 $$|df|=f_m|-\sin \alpha d\alpha |=\sqrt{f_m^2-(f-f_c{)}^2}|d\alpha |,$$ 可得到 $$S(f)=\frac{b}{\sqrt{f_m^2-(f-f_c{)}^2}}[p(\alpha )G(\alpha )+p(-\alpha )G(-\alpha )]\text{\hspace{1em}(1.2-3)}$$ 分别代人电场和磁场的接收增益 $G(\alpha )$，即可得到相应的功率谱密度。$$\begin{array}{rl}{S}_{H_x}(f)& ={\sin }^2\alpha \cdot S_{E_z}(f)=\left[1-{\left(\frac{f-f_c}{f_m}\right)}^2\right]S_{E_z}(f),\\ S_{H_y}(f)& ={\cos }^2\alpha \cdot S_{E_z}(f)={\left(\frac{f-f_c}{f_m}\right)}^2{S}_{E_z}(f)\end{array}$$

电场分量磁场分量互相关
$$\begin{array}{rl}{R}_{E_z{H}_x}(\tau )& =<E_z(t),H_x(t+\tau )>\\ & =-\frac{E_0^2}{\eta }\sum _{n,m}{C}_n{C}_m\sin {\alpha }_m<\cos [({\omega }_c+{\omega }_n)t+{\varphi }_n]\cos [({\omega }_c+{\omega }_m)(t+\tau )+{\varphi }_m]\\ & =-\frac{b}{\eta }\sum _{n=1}^N{C}_n^2\sin {\alpha }_n\cos ({\omega }_c+{\omega }_n)\tau \\ & =-\frac{b}{\eta }{\int }_{-\pi }^{\pi }p(\alpha )\sin \alpha \cos ({\omega }_c\tau +{\omega }_m\tau \cos \alpha )d\alpha \end{array}\text{\hspace{1em}(1.2-16)}$$
第三个等号是考虑到在结果的余炫函数中包含 ${\varphi }_n\pm {\varphi }_m$ 项，当 ${\varphi }_n\ne {\varphi }_m$ 时，平均后的结果为 0，第四个等号是由于 $C_n^2=p({\alpha }_n)d\alpha$。当 $p({\alpha }_n)=p(-{\alpha }_n)$，（1.2-16）是奇函数，因此 $R_{E_z{H}_x}(\tau )=0$。
$$\begin{array}{rl}{R}_{E_z{H}_y}(\tau )& =<E_z(t),H_y(t+\tau )>\\ & =-\frac{b}{\eta }{\int }_{-\pi }^{\pi }p(\alpha )\cos \alpha \cos ({\omega }_c\tau +{\omega }_m\tau \cos \alpha )d\alpha \end{array}\text{\hspace{1em}(1.2-20)}$$ 如果 $p(\alpha )=\frac{1}{2\pi }$ $$R_{E_z{H}_y}(\tau )=\frac{b}{\eta }\sin {\omega }_c\tau J_1(\beta v\tau )\text{\hspace{1em}(1.2-21)}$$ 因此在 $\tau =0$ ,场的三个分量均不相关。

(1.2-21) 的简单证明：
首先给出第一类 bessel 函数的积分形式 $$J_n=\frac{1}{\pi }{\int }_0^{\pi }\cos (n\theta -x\sin \theta )d\theta \text{\hspace{1em}(A.1)}$$ $$J_0=\frac{1}{\pi }{\int }_0^{\pi }\cos (x\sin \theta )d\theta =\frac{1}{\pi }{\int }_0^{\pi }\cos (x\cos \theta )d\theta \text{\hspace{1em}(A.2)}$$ $$J_0=\frac{2}{\pi }{\int }_0^{\pi /2}\cos (x\sin \theta )d\theta =\frac{2}{\pi }{\int }_0^{\pi /2}\cos (x\cos \theta )d\theta \text{\hspace{1em}(A.3)}$$ $$J_1=\frac{1}{\pi }{\int }_0^{\pi }\cos (\theta -x\sin \theta )d\theta =\frac{1}{\pi }{\int }_0^{\pi }\cos \theta \sin (x\cos \theta )d\theta \text{\hspace{1em}(A.4)}$$
由 (1.2-20)
$$\begin{array}{rl}{R}_{E_z{H}_y}(\tau )& =-\frac{b}{\eta }{\int }_{-\pi }^{\pi }\frac{1}{2\pi }\cos \alpha \cos ({\omega }_c\tau +{\omega }_m\tau \cos \alpha )d\alpha \\ & =-\frac{b}{\eta \pi }{\int }_0^{\pi }\cos \alpha [\cos ({\omega }_c\tau )\cos ({\omega }_m\tau \cos \alpha )-\sin ({\omega }_c\tau )\sin ({\omega }_m\tau \cos \alpha )]d\alpha \end{array}$$
积分的第一项 $$\begin{array}{rl}& \frac{b}{\eta \pi }{\int }_0^{\pi }\cos \alpha \cos ({\omega }_c\tau )\cos ({\omega }_m\tau \cos \alpha )d\alpha \\ & =\cos ({\omega }_c\tau )\frac{b}{\eta \pi }{\int }_0^{\pi }\sum _{k=0}^{\infty }\frac{({\omega }_m\tau {)}^{2k}{\cos }^{2k+1}\alpha }{(2k)!}d\alpha \\ & =\cos ({\omega }_c\tau )\frac{b}{\eta \pi }\sum _{k=0}^{\infty }\frac{({\omega }_m\tau {)}^{2k}}{(2k)!}\frac{2k(2k-2)\cdots 2}{(2k+1)(2k-1)\cdots 3}{\int }_0^{\pi }\cos \alpha d\alpha \\ & =0\end{array}$$
根据 （A.4），积分的第二项为 $$\frac{b}{\eta \pi }{\int }_0^{\pi }\cos \alpha \sin ({\omega }_c\tau )\sin ({\omega }_m\tau \cos \alpha )d\alpha =\frac{b}{\eta }\sin ({\omega }_c\tau )J_1({\omega }_m\tau )$$
(1.2-21) 得证。

Ref

1. Microwave Mobile Communicatiaons, Ch1.
2. W. B. Davenport, Jr., and W. L. Root, An Introduction to the Theory of Randum Signals and Noise, McGraw-Hill, New York, 1958.
3. https://archive.lib.msu.edu/crcmath/math/math/e/e078.htm
4. https://mathworld.wolfram.com/PochhammerSymbol.html
5. https://mathworld.wolfram.com/GausssHypergeometricTheorem.html
6. https://archive.lib.msu.edu/crcmath/math/math/g/g054.htm