Adaptive Beamforming and Power Allocation for mmWave Communication in High-Speed Railway
高速铁路毫米波通信的自适应波束成形和功率分配

1. Introduction 1.导言

1. Combination of MR and separation is applied to HSR mmWave beam switching architecture
   将 MR 和 分离结合应用于 HSR 毫米波波束切换架构
2. By introducing a factor related to train speed, the optimized beam is jointly generated by multiple APs, and each beam has optimal power setting.
   通过引入与列车速度相关的因子，优化后的波束由多个接入点共同产生，每个波束都有最佳功率设置。

2. System Model 2.系统模型

3. Optimization and Simulation Results
3.优化和模拟结果

3.1 Optimization Scheme 3.1 优化方案

Algorithm 1: Service range Optimization
Input \(a, \beta, d, d_{\text {min }}\)
if \(a>=0\)
    service range of \(\mathrm{AP}_{1}=a \tan \left(\frac{O P T_{1}}{d_{\min }}\right)\);
    service range of \(\mathrm{AP}_{2}=a \tan \left(\frac{O P T_{2}}{d_{\min }}\right)\);
else
    service range of \(\mathrm{AP}_{1}=a \tan \left(\frac{O P T_{2}}{d_{\min }}\right)\);
    service range of \(\mathrm{AP}_{2}=a \tan \left(\frac{O P T_{1}}{d_{\min }}\right)\);
    end
    Output service range of \(\mathrm{AP}_{1}\) and \(\mathrm{AP}_{2}\)

Algorithm 2: Optimal beamwidth and BSP generation
Input \(P t 、 d_{\min } 、 R 、 h\)
Initialize \(N=0 ; \theta_{\text {sum }}=0\);
for \(k=1\) to \(\operatorname{round}\left(180 / \theta_{\min }\right)\)
    \(\theta(k)=\theta_{\max }\)
While \(\theta_{\text {sum }}<a \tan \left(R / d_{\text {min }}\right) \& \& \theta(k)>\theta_{\text {min }}\)
    While \(P t-P L(D(k))+10 \lg (\operatorname{gain}(\theta(\mathrm{k}), \theta(\mathrm{k}) / 2))<R S\)
                \(\theta(k)=\theta(\mathrm{k})-h\)
                    if \(\theta(k)<\theta_{\min }\)
                    \(\theta(k)=\theta_{\min }\)
                    break
                    end if
                    end while
                    \(\theta_{\text {sum }}=\theta_{\text {sum }}+\theta(k)\)
                    \(\mathbf{B S P}(\mathbf{k})=d_{\text {min }} * \tan \left(\theta_{\text {sum }}\right)\)
                        \(N=N+1\)
end while
Output \(\theta\), BSP, \(N\)

Algorithm 3: Power Optimization
    Input \(a, \beta, \mathbf{B S P}, \max P_{t}, \min P_{t}\)
    Initialize \(N=\) length(BSP);
            \(\boldsymbol{P t}=\min P_{t} *\) ones \((1, N)\)
                opt \(1=1-\beta-(1-\beta)^{*} a^{\wedge} 2 ;\)
                opt \(2=1-\beta+(1-\beta)^{*} a^{\wedge} 2\);
    if \(\mathrm{a}>0\)
        \(\boldsymbol{P}_{t}\left(1: \operatorname{ceil}\left(N^{*}(\right.\right.\) opt1 \(\left.\left.)\right)\right)=\operatorname{linspace}\left(\max P_{t}, \min P_{t}\right.\), ceil \(\left(N^{*}(\right.\) opt1 \(\left.\left.)\right)\right)\)
    else
        \(\boldsymbol{P}_{t}\left(1: \operatorname{ceil}\left(N^{*}(\operatorname{opt} 2)\right)\right)=\operatorname{linspace}\left(\max P_{t} \max P_{t}, \min P_{t}, \operatorname{ceil}\left(N^{*}(\operatorname{opt} 2)\right)\right)\);
    end if
    Output \(P_{t}\)

3.2 Simulation Results 3.2 模拟结果

Fig. 3. Optimized beam width when
图 3.当 时的优化波束宽度

4. Conclusion 4.结论

References 参考资料