【摘要】：針對(duì)射頻功放的非線性特性進(jìn)行了研究,提出一種新的稀疏化的Volterra級(jí)數(shù)模型。該模型基于壓縮感知算法,將稀疏系統(tǒng)的辨識(shí)等效為信號(hào)的重構(gòu)問題,利用正則正交匹配(ROMP)算法對(duì)核系數(shù)進(jìn)行稀疏化并選擇出活躍的核系數(shù)。將提出的模型與記憶多項(xiàng)式(MP)模型、通用記憶多項(xiàng)式(GMP)模型進(jìn)行比較,較MP模型的建模精度提升10.7 dB,模型系數(shù)減少25%;較GMP模型的建模精度提升3.9 dB,模型系數(shù)減少57.65%。仿真結(jié)果表明,提出的方法實(shí)現(xiàn)了良好的預(yù)失真線性化性能,極大地降低了模型系數(shù),優(yōu)于傳統(tǒng)的功放行為模型,由此驗(yàn)證對(duì)功放的線性化技術(shù)發(fā)展具有參考價(jià)值。
[Abstract]:The nonlinear characteristics of RF power amplifier are studied and a new sparse Volterra series model is proposed. Based on the compression sensing algorithm, the model equates the identification of sparse systems to the problem of signal reconstruction. The regularized orthogonal matching (ROMP) algorithm is used to sparse the kernel coefficients and select the active kernel coefficients. The proposed model is compared with the memory polynomial (MP) model and the general memory polynomial (GMP) model. Compared with the MP model, the modeling accuracy is increased by 10.7 dB and the model coefficient is reduced by 25 dB, and the modeling precision is increased by 3.9 dB and the model coefficient is reduced by 57.65 dB compared with that of the GMP model. The simulation results show that the proposed method achieves good predistortion linearization performance, greatly reduces the model coefficient, and is superior to the traditional power amplifier behavior model, which has a reference value for the development of power amplifier linearization technology.
【作者單位】： 遼寧工程技術(shù)大學(xué)電子與信息工程學(xué)院;
【基金】：國家自然科學(xué)基金面上項(xiàng)目(61372058) 遼寧省教育廳科學(xué)研究一般項(xiàng)目(L2015209) 遼寧省高等學(xué)校重點(diǎn)實(shí)驗(yàn)室資助項(xiàng)目(LJZS007)
【分類號(hào)】：TN722.75
，

本文編號(hào)：2128469