在各种信道类型下的慢速链路自适应方法

摘要

本发明公开了一种在各种信道类型下的慢速链路自适应方法，所述方法包括：判断通信链路的信道类型；根据信道类型以及链路自适应门限确定不同的QAM调制级数。本发明的特点在于：(1)本发明是基于信道实际状况的，而不是采用假设信道，更符合通信场景的实际情况，提高了通信系统的性能；(2)本发明对信道类型进行了判断，获得了通信链路准确的信道类型，不仅可用于链路自适应技术领域，而且可以用于中继选择、接收机等众多通信领域，应用广泛；(3)本发明仍然是一种慢速链路自适应技术，因此具备慢速链路自适应技术的特点，能有效降低系统实现复杂度。

主权项

一种在各种信道类型下的慢速链路自适应方法，所述方法包括：判断通信链路的信道类型；根据信道类型以及链路自适应门限确定不同的QAM调制级数；所述通信链路的信道类型的判断步骤如下：(1)通信链路的n个即时信噪比样本为其中：n表示即时信噪比的样本数，n的取值大于1000；p＝1时γ_pi表示源节点S到第i个中继节点链路的即时信噪比，p＝2时γ_pi表示第i个中继节点到目的节点D链路的即时信噪比，i＝1,2,…,L，L表示中继节点的数量；(2)分别计算通信链路在第j种信道类型下的即时信噪比样本均值${\bar{\gamma }}_{\mathrm{pi}}^{\left(j\right)}=\frac{1}{n}\underset{l=1}{\overset{n}{\Sigma }}{\gamma }_{\mathrm{pi}}^{\left(l\right)}$和标准差$S_{\mathrm{pi}}^{\left(j\right)}=\sqrt{\frac{1}{n-1}\underset{l=1}{\overset{n}{\Sigma }}{({\gamma }_{\mathrm{pi}}^{\left(l\right)}-{\bar{\gamma }}_{\mathrm{pi}}^{\left(j\right)})}^2},$估计该通信链路在第j种信道下的信噪比概率密度函数的参数：利用得到该通信链路在第j种信道下的即时信噪比近似概率密度函数$f_0^{\left(j\right)}\left({\gamma }_{\mathrm{pi}}\right)=\underset{h=1}{\overset{N_j}{\Sigma }}\widehat{{\alpha }_h^{\left(j\right)}}·{\left({\gamma }_{\mathrm{pi}}\right)}^{\widehat{{\beta }_h^{\left(j\right)}-1}}·e^{\widehat{{-\xi }_h^{\left(j\right)}{\gamma }_{\mathrm{pi}}}},$其中：N_j、为不同信道类型下的函数参数；j表示信道类型的种类数，j＝1,2,3,…；(3)令求的概率密度函数并计算的k阶矩：${\eta }_{\mathrm{pi}}^{(k,j)}={\int }_a^{+\infty }{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^k·f^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)d{\eta }_{\mathrm{pi}}^{\left(j\right)},$$a=-\frac{{\mu }_{\mathrm{pi}}^{\left(j\right)}}{{\sigma }_{\mathrm{pi}}^{\left(j\right)}},$其中：和分别为即时信噪比近似概率密度函数f₀^(j)(γ_pi)的期望和方差；(4)构造关于概率密度函数的正规直交系：$B_t^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)=\frac{1}{G[1,{\eta }_{\mathrm{pi}}^{\left(j\right)},···,{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1}]}\left|\begin{array}{ccccc}{\mu }_0& {\mu }_1& ...& {\mu }_{t-1}& 1\\ {\mu }_1& {\mu }_2& ...& {\mu }_t& {\eta }_{\mathrm{pi}}^{\left(j\right)}\\ .& .& & .& .\\ .& .& ...& .& .\\ .& .& & .& .\\ {\mu }_t& {\mu }_{t+1}& ...& {\mu }_{2t-1}& {\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^t\end{array}\right|,$且计算$\stackrel{‾}{{\left[{\eta }_{\mathrm{pi}}^{\left(j\right)}\right]}^t\left[B_t^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)\right]}={\int }_a^{+\infty }{f}^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right){\left[B_t^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)\right]}^2{\mathrm{d\eta }}_{\mathrm{pi}}^{\left(j\right)};$式中，为的格拉姆行列式：$G[1,{\eta }_{\mathrm{pi}}^{\left(j\right)},···,{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1}]=\left|\begin{array}{cccc}<\mathrm{1,1}>& <1,{\eta }_{\mathrm{pi}}^{\left(j\right)}>& ...& <1,{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1}>\\ <{\eta }_{\mathrm{pi}}^{\left(j\right)},>1& <{\eta }_{\mathrm{pi}}^{\left(j\right)},{\eta }_{\mathrm{pi}}^{\left(j\right)}>& ...& <{\eta }_{\mathrm{pi}}^{\left(j\right)},{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1}>\\ .& .& & .\\ .& .& ...& .\\ .& .& & .\\ <{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1},1>& <{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1},{\eta }_{\mathrm{pi}}>& ...& <{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1},{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^{t-1}>\end{array}\right|,$其中：内积$<{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^p,{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^q>\stackrel{\Delta }{=}{\int }_a^{+\infty }{f}^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)·{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^p·{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^q{\mathrm{d\eta }}_{\mathrm{pi}}^{\left(j\right)},$高阶矩${\mu }_t\stackrel{\Delta }{=}{\int }_a^{+\infty }{f}^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)·{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}^t{\mathrm{d\eta }}_{\mathrm{pi}}^{\left(j\right)};$(5)分别用通信链路在第j种信道类型下的即时信噪比样本均值和标准差代替即时信噪比的期望和方差，即：令标准化的即时信噪比变量求的概率密度函数并计算的k阶矩：$\overline{{\eta }_{\mathrm{pi}}^{(k,j)}}={\int }_b^{\infty }{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^k·f^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}d\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},$$b=-\frac{{\bar{\gamma }}_{\mathrm{pi}}^{\left(j\right)}}{S_{\mathrm{pi}}^{\left(j\right)}};$(6)构造并计算关于概率密度函数的正规直交系：$B_t^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}=\frac{1}{G[1,\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},···,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}]}\left|\begin{array}{ccccc}\overline{{\mu }_0}& \overline{{\mu }_1}& ...& \overline{{\mu }_{t-1}}& 1\\ \overline{{\mu }_1}& \overline{{\mu }_2}& ...& \overline{{\mu }_t}& \overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}}\\ .& .& & .& .\\ .& .& ...& .& .\\ .& .& & .& .\\ \overline{{\mu }_t}& \overline{{\mu }_{t+1}}& ...& \overline{{\mu }_{2t-1}}& {\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^t\end{array}\right|;$式中，$G[1,\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},···,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}]$分别为$1,\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},···,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}$的格拉姆行列式：$G[1,\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},···,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}.]=\left|\begin{array}{cccc}<\mathrm{1,1}>& <1,\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}}>& ...& <1,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}>\\ <\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},1>& <\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}}>& ...& <\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}>\\ .& .& & .\\ .& .& ...& .\\ .& .& & .\\ <{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1},1>& <{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1},\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}}>& ...& <{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1},{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t-1}>\end{array}\right|$其中：内积$<{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^p,{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^q>\stackrel{\Delta }{=}{\int }_b^{+\infty }{f}^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}·{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^p·{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{q}d\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}},$高阶矩$\overline{{\mu }_t}\stackrel{\Delta }{=}{\int }_b^{+\infty }{f}^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}·{\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}^{t}d\overline{{\eta }_{\mathrm{pi}}^{\left(j\right)}};$(7)计算通信链路在第j种信道类型下概率密度函数的级数展开式：$f_{\eta }^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}=f^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)\underset{t=0}{\overset{\infty }{\Sigma }}\frac{B_t^{\left(j\right)}\overline{\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)}}{{\left[{\eta }_{\mathrm{pi}}^{\left(j\right)}\right]}^t\left[B_t^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)\right]}\left[B_t^{\left(j\right)}\left({\eta }_{\mathrm{pi}}^{\left(j\right)}\right)\right];$(8)计算通信链路的即时信噪比概率密度函数的级数展开式：$f_{{\gamma }_{\mathrm{pi}}}^{\left(j\right)}\left({\gamma }_{\mathrm{pi}}\right)=\frac{1}{S_{\mathrm{pi}}^{\left(j\right)}}{f}_{\eta }^{\left(j\right)}\left(\frac{{\gamma }_{\mathrm{pi}}-{\bar{\gamma }}_{\mathrm{pi}}^{\left(j\right)}}{S_{\mathrm{pi}}^{\left(j\right)}}\right);$(9)确定通信链路的信道类型：将f₀^(j)(γ_pi)和代入到公式中，得到各通信链路的值，最小值所对应的信道类型便是最符合实际的信道类型，即为相应通信链路的信道类型。