当前位置: X-MOL 学术Optim. Control Appl. Methods › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
A new method for the solution of hybrid analog digital beamforming problems
Optimal Control Applications and Methods ( IF 1.8 ) Pub Date : 2020-07-06 , DOI: 10.1002/oca.2636
Ebrahim Amini 1 , Hamid Reza Marzban 2 , Mahdi Rastegari 3
Affiliation  

By going to millimeter wave (mmWave) we can use large scale MIMO due to short mmWave wavelength to overcome path loss by using beamforming to focus power of signal to the receiver. System structure of mmWave band is different with conventional MIMO because of large scale MIMO which is leading to use many RF‐chains. For this reason Hybrid structure have been proposed for large Scale MIMO. By going to Hybrid structure a new issue has been created with phase shifter constraint. In this paper we propose a novel solution to make a hybrid precoding and combing to reach high spectral efficiency. Our problem includes a maximizing Frobenius norm of multiplying five complex matrices. As constraints, all elements of some matrices must have unit norm, and also the multiplication of some matrices must have a specific Frobenius norm. A novel solution is proposed for the problem, which is gained in two steps. At the first step, we propose a new decomposition usable for kind of complex matrices which could lead to reduce the dimension of the objective function, and eliminating constraints completely and in the second step, we propose an iterated algorithm for the resulting problem. By proposing another novel technique, we show that in each iteration, the optimal problem is equivalent to a quadratic optimal problem with limitations on the vector norm, and as a result, an optimal problem is obtained quickly. Moreover, some other real examples are included to demonstrate the usefulness and effectiveness of the proposed algorithm.

中文翻译:

解决混合模拟数字波束成形问题的新方法

通过使用毫米波(mmWave),由于毫米波波长短,我们可以使用大规模MIMO,以通过使用波束成形将信号功率集中到接收器来克服路径损耗。mmWave频带的系统结构与常规MIMO不同,因为大规模MIMO导致使用许多RF链。因此,已经提出了用于大规模MIMO的混合结构。通过进入混合结构,已经产生了具有移相器约束的新问题。在本文中,我们提出了一种新颖的解决方案,可以使混合预编码和合并达到高频谱效率。我们的问题包括将五个复杂矩阵相乘的最大化Frobenius范数。作为约束,某些矩阵的所有元素都必须具有单位范数,并且某些矩阵的乘法也必须具有特定的Frobenius范数。针对该问题提出了一种新颖的解决方案,该解决方案分两步获得。第一步,我们提出了一种适用于复杂矩阵类型的新分解方法,该方法可以减少目标函数的维数,并完全消除约束,第二步,我们针对结果问题提出一种迭代算法。通过提出另一种新颖的技术,我们表明在每次迭代中,最优问题都等同于对向量范数有限制的二次最优问题,因此,可以很快获得最优问题。此外,还包括一些其他实际示例来证明所提出算法的有用性和有效性。我们提出了一种适用于复杂矩阵类型的新分解方法,该分解方法可以减少目标函数的维数,并完全消除约束,第二步,针对结果问题提出一种迭代算法。通过提出另一种新颖的技术,我们表明,在每次迭代中,最优问题都等同于对向量范数有限制的二次最优问题,因此,可以很快获得最优问题。此外,还包括一些其他实际示例来证明所提出算法的有用性和有效性。我们提出了一种适用于复杂矩阵类型的新分解方法,该分解方法可以减少目标函数的维数,并完全消除约束,第二步,针对结果问题提出一种迭代算法。通过提出另一种新颖的技术,我们表明,在每次迭代中,最优问题都等同于对向量范数有限制的二次最优问题,因此,可以很快获得最优问题。此外,还包括一些其他实际示例,以证明所提出算法的有用性和有效性。最优问题等效于对向量范数有限制的二次最优问题,结果,可以很快获得最优问题。此外,还包括一些其他实际示例来证明所提出算法的有用性和有效性。最优问题等效于对向量范数有限制的二次最优问题,结果,可以很快获得最优问题。此外,还包括一些其他实际示例来证明所提出算法的有用性和有效性。
更新日期:2020-07-06
down
wechat
bug