Three Waveforms for WET-Deterministic Sine, M-QAM, Gaussian

Date: 2024.04.25 15:39
Author: Joffrey LC


Intelligent Reflecting Surface Aided Wireless Power Transfer With a DC-Combining Based Energy Receiver and Practical Waveforms. Qingdong Yue et.al. IEEE Transactions on Vehicular Technology, September 2022 (pdf) (Citations 6)

Quick Overview

  • Multiple antennas energy receiver
  • Three classic waveforms: deterministic waveform, M-QAM waveform and Gaussian waveform are considered for WPT.

  • Each antenna with a EH.

image-20240425160712519

一般文章都是认为总的RF能量为每根天线上的和,然后再进行整理为直流功率。但是由于每根天线上的信号的相位不同,没法很好地直接合并。所以每根天线一个EH,再进行合并。

  • Active beamforming and Passive phase shift 都是closed-from

  • The numerical results demonstrate that the Gaussian waveform has the best energy performance with a low input RF power to the energy harvesters. By contrast, the deterministic waveform becomes superior with a high input RF power to the energy harvesters.

System model

  • single antennas transmitter
  • single antennas receiver
  • single ​ elements IRS
  • Tx and ER are both ULA, while IRS is UPA
  • Tx-ER is Rayleigh fading

考虑三种波形:

  • deterministic sinusoidal wave form
  • M-QAM waveform
  • Gaussian waveform

Three kinds of waveforms

the non-linear energy harvesting function

image-20240506205841898

deterministic sinusoidal wave form

image-20240506162208427

the objective is formulated as

image-20240506162929658

进行一阶泰勒展式:

image-20240506163807634

得到结果,即最大化入射能量即可,因为后面是常数。

通过AO的方案进行:

  • 固定,优化,最优的是对齐级联信道的最大特征向量,

image-20240506164914271

  • 固定,优化,通过 minorize-maximization (MM) technique进行,MM算法的核心是找到一个==替代函数==,对替代函数进行求解

image-20240506203947834

这个替代函数的由来应该是直接打开

并且丢掉和无关的项。

然后通过拉格朗日乘子法,求解对偶问题

image-20240506204417595

然后写出对偶问题:

image-20240506204503576

以及KKT条件:

image-20240506204520454

最后得到闭式解。

M-QAM waveform

image-20240506162233766

乘了==功率==和==概率==

求解过程和deterministic sinusoidal wave form 类似,因为

image-20240506205607935

Gaussian waveform

image-20240506162310295

image-20240506162321040

用PDF(概率)积分算

一系列操作之后也和上面的类似

WITH THE SAME MAXIMUM TRANSMIT POWER CONSTRAINT

deterministic sinusoidal wave form

因为确定正弦波的功率是常数,所以可以直接乘系数搞定

M-QAM waveform

image-20240506211128308

也是乘常数搞定

Gaussian waveform

也差不多

Numerical results

image-20240506211625184

  • 4QAM和sine波相同,因为都有相同的包络
  • 16QAM和Gaussian在低功率区间性能好,因为偶尔的高功率使得非线性EH被激活,并以高RF-DC效率运行
  • 而高功率时,确定性波形一直都在高效率运行,而非确定性波形还可能在低效率区域运行

image-20240506212043729

  • 考虑最大发射功率限制的时候,由于高阶调制会有很多低功率符号,所以WET性能差。

image-20240506212324470

类似于DC combiner and RF combiner

  • 如果只有一个EH,在低功率区,多个RF的和输入EH,可能会使得EH工作在高效率区间。然而可能会使得在高功率情况下工作在饱和区