Research On Concurrent Dual-band Digital Predistortion Linearization Method

For the current situation of multi-frequency multi-mode in wireless mobile communication systems,a dual-band communication system capable of supporting two different frequency bands simultaneously working has attracted much attention.At the same time,spectrum resources are becoming increasingly tight,and modulation schemes and transmission systems with high spectral efficiency are widely used to ensure linearity while improving efficiency.In the dual-band communication system,the dual-band power amplifier is one of the most important nonlinear devices in the RF communication system,and its linearity directly affects the transmission performance of the system.In order to ensure the efficiency and linearity of the communication system,the corresponding linearization technique is needed to compensate the nonlinear distortion generated by the power amplifier.The dual-time digital pre-distortion technology can compensate for such nonlinear distortion well,and has become a hot research topic in domestic and oversea.This thesis focus on the research of the low-complexity concurrent dual-band pre-distortion model and the inter-band intermodulation distortion suppression method for concurrent near-band-interval dual-band signals are studied.The main works and innovations points are as follows.1.The 2D-RSBMP was proposed.Since the model reduces the three in-line summations to the quadratic inline summation,the model coefficients and implementation complexity are greatly reduced.At the same time,it is proved by mathematical theory that 2D-RSBMP can contain all distortion terms of 2D-DPD model,which guarantees the nonlinear compensation performance of the model.The test results show that compared with the 2D-DPD model,the proposed 2D-RSBMP model reduces the model coefficient by more than 66%,and the ACPR performance is improved by about 2-4dBc.The normalized mean square The NMSE performance is improved by about 11dB;compared with the 2D-SOC model,the 2D-RSBMP model coefficient is reduced by 50%,the ACPR performance is improved by about 2-4dBc,and the NMSE performance is improved by about 14dB;and 2D-MMP Compared to other models,the 2D-RSBMP model requires less runtime than other performance indicators.2.For the phenomenon that the distance between two frequency bands is close,the interband distortion,in-band distortion and intermodulation distortion are aliased together.A 2D-CSSMP model is proposed.The model considers both the in-band distortion term and the out-of-band distortion term in the expression,so it can effectively suppress the distortion aliasing caused by the close band spacing of the dual-frequency signals.The test results show that the proposed 2D-CSSMP model can reduce ACPR by 15dBc and NMSE by 15dB compared with the 2D-DPD model.Compared with the single-frequency Memory Polynomial model,the proposed 2D-CSSMP model can achieve similar nonlinear compensation performance.And the sampling rate is reduced to 1/4 of the single-frequency MP model,but the complexity will increase.