未编码下QAM,PAM,PSK系统的误码率理论公式

误符号率

最佳基带传输系统（匹配滤波），信道误符号率(symbol error rate) 的最终表达式。
符号能量 $E_s$ 与 比特能量 $E_b$ 之间的关系是 $E_s=E_b{\log }_{2}M$

PAM,ASK

对于PAM 或ASK星座映射，星座大小为$M$，两个星座的最小距离为
$$d_{\min }=\sqrt{\frac{12{\log }_{2}M}{M^2-1}{E}_b}$$
$$P_e=2\left(1-\frac{1}{M}\right)Q\left(\frac{d_{\min }}{\sqrt{2N_0}}\right)=\left(1-\frac{1}{M}\right)\mathrm{erfc}\left(\frac{d}{\sqrt{2}\sigma }\right)=\left(1-\frac{1}{M}\right)\mathrm{erfc}\left[\sqrt{\frac{3}{M^2-1}\cdot \frac{E_b{\log }_{2}M}{N_0}}\right]$$
特别的，当 $M=2$, bpsk

$$P_b=\frac{1}{2}\mathrm{erfc}(\sqrt{E/N_0})$$

QAM

对于QAM，可以相当于两路PAM信号，因此有
$$\begin{array}{c}\begin{array}{rl}{P}_{e,M-\text{QAM}}& =1-(1-P_{e,\sqrt{M}-\text{PAM}}{)}^2\\ & =2P_{e,\sqrt{M}-\text{PAM}}\left(1-\frac{1}{2}{P}_{e,\sqrt{M}-\text{PAM}}\right)\\ & \approx 2P_{e,\sqrt{M}-\text{PAM}}\\ & =2\left(1-\frac{1}{\sqrt{M}}\right)\mathrm{erfc}\left[\sqrt{\frac{3}{M-1}\cdot \frac{0.5E_b{\log }_{2}M}{N_0}}\right]\end{array}\end{array}$$

For QPSK system,BER
$$P_{e,\text{QPSK}}=\mathrm{erfc}(\sqrt{E_b/N_0})$$

PSK

$$P_e=2Q\left(\sqrt{(2{\log }_{2}M){\sin }^2\left(\frac{\pi }{M}\right)\frac{E_b}{N_0}}\right)=\mathrm{erfc}\left(\sin \left(\frac{\pi }{M}\right)\sqrt{\frac{E_b{\log }_{2}M}{N_0}}\right)$$

误比特率 Pb

典型的通信系统使用格雷编码调制映射，即相邻符号中对应的信息比特只相差一个比特。当一个符号被错误判决时，它通常会落入相邻的符号内，因此每个符号错误都会导致 ${\log }_{2}M$ 比特种的一个比特出错，因此符号错误和比特错误（bit errior rate)之间的关系是
$$P_b\approx P_s/{\log }_{2}M$$

PAM

$$P_{b,\text{PAM}}=\frac{1}{{\log }_{2}M}\left(1-\frac{1}{M}\right)\mathrm{erfc}\left[\sqrt{\frac{3}{M^2-1}\cdot \frac{E_b{\log }_{2}M}{N_0}}\right]$$

QAM

$$P_{b,\text{QAM}}=\frac{2}{{\log }_{2}M}\left(1-\frac{1}{\sqrt{M}}\right)\mathrm{erfc}\left[\sqrt{\frac{3}{M-1}\cdot \frac{0.5E_b{\log }_{2}M}{N_0}}\right]$$

PSK

$$P_{b,\text{PSK}}=\frac{1}{{\log }_{2}M}\mathrm{erfc}\left(\sin \left(\frac{\pi }{M}\right)\sqrt{\frac{E_b{\log }_{2}M}{N_0}}\right)$$

注意 $P_{b,\text{bpsk}}=P_{b,\text{QPSK}}=0.5\mathrm{erfc}\sqrt{E_b/N_0}$

Q(x) and erfc

$$Q(x)=\frac{1}{\sqrt{2\pi }}{\int }_x^{\infty }\exp (-t^2/2)dt$$
The Q function is related to the complementary error function, erfc, according to
$$Q(x)=\frac{1}{2}\mathrm{erfc}\left(\frac{x}{\sqrt{2}}\right),\mathrm{erfc}(x)=2Q(\sqrt{2}x)$$

[1] Modulation roundup: error rates, noise, and capacity
[2] J. Proakis and M. Salehi, Digital Communications, 5th Edition, 5th edition. Boston: McGraw-Hill Education, 2007.