Research On Cubic Metric Reduction Technology For LTE Downlink

LTE, which has been selected as one of the 4G technology standards by the international Telecommunication Union, uses OFDM technology and high-order Quadrature Amplitude Modulation. Compared with the 3G communication standards, LTE can provide higher transmitted bandwidth and spectrum efficiency.However, one of the major disadvantages of OFDM and high-order QAM is that the signals of LTE have large envelope fluctuations. The main disadvantages of the large envelope fluctuations of LTE signals can be shown as following. Firstly, it could result in poor high power amplifiers efficiency. Secondly, when passing through high power amplifiers, the LTE signals will cause severe nonlinear distortion which affects the in-band distortion and the out-of-band spectrum expansion. Thirdly, it require high dynamic range for digital-to-analog converters, which decrease the conversion efficiency of digital-to-analog.In order to avoid such negative effects, the LTE signals which input high power amplifiers should be operated with an input back-off. Peak-to-average Power Ratio is widely employed to predict the required power de-rating. However, recent studies have shown that PAPR cannot predict the necessary of input back-off, while another metric, named cubic metric, can be better. Therefore, we can decrease the necessary of input back-off by reducing cubic metric in order to improve the high power amplifiers efficiency.This paper proposed two methods on cubic metric reduction in LTE downlink transmission by studying cubic metric.Firstly, a descendent clipping and filtering algorithm has been proposed according to the existing clipping and filtering method. Combine with cubic metric, the detailed theoretical analysis reveals that if we tend to limit these higher the preset threshold level signals with a decreased factor, it can reduce the iterations and improve its convergence. Simulation results have shown that the proposed method has obtained reducing the iterations from 8 times down to 1.Secondly, this study proposed another method to reduce cubic metric using convex optimization. The subcarriers of LTE signals can be divided into three categories: data subcarriers, in-band idle subcarriers and out-of-band idle subcarriers. When designing method to reduce cubic metric of LTE signals, it’s necessary to assure meeting the constraints, such as error vector magnitude and adjacent channel leakage ratio. According to optimization theory, this paper established an optimization model between cubic metric and these constraints. Then, its convex model has been obtained by detailed theoretical analysis. Finally, a custom numerical solution using interior-point method has been proposed. Simulation has shown that the proposed method has good convergence and low computational complexity.The two proposed methods in this paper are both convenient and practical, because it needn’t to change the existing communication protocols.