Taijun Liu

Taijun Liu

Auto-generated profile, based on publicly available information

Taijun Liu

Ningbo University

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
26 November 2014 in Photonics
Verified
Journal Article
Open access
PDF file
Xiupu ZhangXiupu Zhang
Xiupu Zhang
RZRan Zhu
Ran Zhu
DSDongya Shen
Dongya Shen
TLTaijun Liu
Taijun Liu
Linearization technologies that can be used for linearizing RoF transmission are reviewed. Three main linearization methods, i.e. electrical analog linearization, optical linearization, and electrical digital linearization are presented and compared. Analog linearization can be achieved using analog predistortion circuits, and can be used for suppression of odd order nonlinear distortion components, such as third and fifth order. Optical linearization includes mixed-polarization, dual-wavelength, optical channelization and the others, implemented in optical domain, to suppress both even and odd order nonlinear distortion components, such as second and third order. Digital predistortion has been a widely used linearization method for RF power amplifiers. However, digital linearization that requires analog to digital converter is severely limited to hundreds of MHz bandwidth. Instead, analog and optical linearization provide broadband linearization with up to tens of GHz. Therefore, for broadband radio over fiber transmission that can be used for future broadband cloud radio access networks, analog and optical linearization are more appropriate than digital linearization. Generally speaking, both analog and optical linearization are able to improve spur-free dynamic range greater than 10 dB over tens of GHz. In order for current digital linearization to be used for broadband radio over fiber transmission, the reduced linearization complexity and increased linearization bandwidth are required. Moreover, some digital linearization methods in which the complexity can be reduced, such as Hammerstein type, may be more promising and require further investigation.

6 December 2007 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
SBSlim Boumaiza
Slim Boumaiza
MHMohamed Helaoui
Mohamed Helaoui
OHOualid Hammi
Oualid Hammi
TLTaijun Liu
Taijun Liu
FGFadhel Ghannouchi
Fadhel Ghannouchi
This paper proposes a comprehensive and systematic characterization methodology that is suitable for the forward and reverse behavior modeling of wireless transmitters (Txs) driven by wideband-modulated signals. This characterization approach can be implemented in adaptive radio systems since it does not require particular signal or training sequences. The importance of the nature of the driving signal and its average power on the behavior of radio-frequency Txs are experimentally investigated. Critical issues related to the proposed characterization approach are analytically studied. This includes a new delay-estimation method that achieves good accuracy with low computational complexity. In addition, the receiver linear calibration and its noise budget are investigated. To demonstrate the accuracy and robustness of the proposed method, a full characterization (including the memoryless nonlinearity and the memory effects) of a 100-W Tx driven by a multicarrier wideband code-division multiple-access signal is carried out, and its forward and reverse models are identified. Cascading the identified reverse model derived using the proposed methodology and the Tx prototype leads to excellent compensation of the static nonlinearities and the memory effects exhibited by the latter. Critical issues in implementing this approach are also discussed.

2 June 2021 in IEEE Antennas and Wireless Propagation Letters
Verified
Journal Article
Closed access
MHMinyong Huang
Minyong Huang
YLYunlong Lu
Yunlong Lu
QZQi-Ang Zhu
Qi-Ang Zhu
MSMilan Salek
Milan Salek
YWYi Wang
Yi Wang
TLTaijun Liu
Taijun Liu
This letter presents a highly integrated power amplifier (PA) antenna with the capability of frequency tuning. A planar inverted-F antenna (PIFA) is used to directly match the PA transistor. This mitigates the lossy output matching network in the PA-integrated antenna (PAIA). In order to achieve a wide frequency tuning range, a combination of capacitors is loaded at the PIFA to control the antenna input impedances so that matching can be maintained with frequency tuning. The antenna analysis and design procedure of the tunable PAIA are detailed. For verification, a prototype is designed, fabricated, and measured. The experimental results show that the frequency tuning range of the proposed PAIA covers from 2 to 4 GHz. Over this frequency tuning range, the effective isotropic radiation power and power-added efficiency are higher than 40.5 dBm and 60% at the input power of 26.5 dBm, respectively.

10 December 2010 in IEEE Photonics Technology Letters
Verified
Journal Article
Closed access
BHBouchaib Hraimel
Bouchaib Hraimel
Xiupu ZhangXiupu Zhang
Xiupu Zhang
WJWei Jiang
Wei Jiang
KWKe Wu
Ke Wu
TLTaijun Liu
Taijun Liu
KXKun Xu
Kun Xu
We experimentally investigate optical mixed polarization for the linearization of an electro-absorption modulator (EAM) in radio-over-fiber links. A linear polarizer is used before and after a polarization-dependent EAM. Thus, both transverse-electric and transverse-magnetic polarized lights are simultaneously modulated by different amounts. By carefully optimizing two polarization angles, the third-order intermodulation distortion (IMD3) is suppressed. The linearization leads to suppression of the IMD3 by more than 16 dB and improvement of spurious-free dynamic range by 8.1 and 9.5 dB for back-to-back and after 20 km of fiber transmission, respectively, compared to the nonlinearization case.

1 June 2014 in Electronics Letters
Verified
Journal Article
Open access
PDF file
MHMing Hui
Ming Hui
TLTaijun Liu
Taijun Liu
MZMeng Zhang
Meng Zhang
YYYan Ye
Yan Ye
DSDongya Shen
Dongya Shen
XYXiangyue Ying
Xiangyue Ying
An augmented radial basis function neural network (ARBFNN) is proposed for modelling and linearising a wideband Doherty power amplifier (DPA) with strong memory effects and static nonlinearity. To evaluate the performance of the ARBFNN, a 51 dBm DPA and a 25 MHz mixed test signal were used in modelling and linearisation measurement. Compared with the memory polynomial (MP) model and the real‐valued time‐delay neural network (RVTDNN), the ARBFNN is highly effective, leading to 3 and 5 dB improvements in the normalised mean square error. More importantly, the ARBFNN predistorter represents a significant improvement over the RVTDNN and MP in the suppression of the out‐of‐band spectral regrowth. In addition, the ARBFNN has a similar linearisation capability as the generalised MP model, but has much better numerical stability.

26 March 2025 in IEEE Signal Processing Letters
Verified
Journal Article
Closed access
WLWenhan Li
Wenhan Li
JWJiangong Wang
Jiangong Wang
TLTaijun Liu
Taijun Liu
Gaoming XuGaoming Xu
Gaoming Xu
Radio frequency (RF) fingerprinting is a technique used to identify different transmitters by analyzing the unique hardware impairments of RF transmitters, known as RF fingerprints. However, many existing studies have primarily focused on transmitter impairments, while the impact of receiver hardware impairments on RF signals has often been overlooked. To alleviate this issue, this letter proposes a receiver-agnostic RF fingerprinting method using unsupervised domain adaptation. The method employs prototypical contrastive learning to align the features of source domain and target domain data, while simultaneously learning the features of both domains. Experimental results on the real-world dataset (WiSig) demonstrate that the proposed method outperforms other receiver-agnostic RF fingerprinting methods.