#### Table of Content

25 January 2022, Volume 20 Issue S1
Research Paper
An Improved Parasitic Parameter Extraction Method for InP HEMT
DUAN Lanyan, LU Hongliang, QI Junjun, ZHANG Yuming, ZHANG Yimen
2022, 20(S1):  1-6.  doi:10.12142/ZTECOM.2022S1001
Asbtract ( 39 )   HTML ( 180)   PDF (2023KB) ( 82 )
Figures and Tables | References | Related Articles | Metrics

An improved parasitic parameter extraction method for InP high electron mobility transistor (HEMT) is presented. Parasitic parameter extraction is the first step of model parameter extraction and its accuracy has a great impact on the subsequent internal parameter extraction. It is necessary to accurately determine and effectively eliminate the parasitic effect, so as to avoid the error propagation to the internal circuit parameters. In this paper, in order to obtain higher accuracy of parasitic parameters, parasitic parameters are extracted based on traditional analytical method and optimization algorithm to obtain the best parasitic parameters. The validity of the proposed parasitic parameter extraction method is verified with excellent agreement between the measured and modeled S-parameters up to 40 GHz for InP HEMT. In 0.1–40 GHz InP HEMT, the average relative error of the optimization algorithm is about 9% higher than that of the analysis method, which verifies the validity of the parasitic parameter extraction method. The extraction of parasitic parameters not only provides a foundation for the high-precision extraction of small signal intrinsic parameters of HEMT devices, but also lays a foundation for the high-precision extraction of equivalent circuit model parameters of large signal and noise signals of HEMT devices.

Auxiliary Fault Location on Commercial Equipment Based on Supervised Machine Learning
ZHAO Zipiao, ZHAO Yongli, YAN Boyuan, WANG Dajiang
2022, 20(S1):  7-15.  doi:10.12142/ZTECOM.2022S1002
Asbtract ( 24 )   HTML ( 167)   PDF (3209KB) ( 40 )
Figures and Tables | References | Related Articles | Metrics

As the fundamental infrastructure of the Internet, the optical network carries a great amount of Internet traffic. There would be great financial losses if some faults happen. Therefore, fault location is very important for the operation and maintenance in optical networks. Due to complex relationships among each network element in topology level, each board in network element level, and each component in board level, the concrete fault location is hard for traditional method. In recent years, machine learning, especially deep learning, has been applied to many complex problems, because machine learning can find potential non-linear mapping from some inputs to the output. In this paper, we introduce supervised machine learning to propose a complete process for fault location. Firstly, we use data preprocessing, data annotation, and data augmentation in order to process original collected data to build a high-quality dataset. Then, two machine learning algorithms (convolutional neural networks and deep neural networks) are applied on the dataset. The evaluation on commercial optical networks shows that this process helps improve the quality of dataset, and two algorithms perform well on fault location.

Design of Raptor-Like Rate Compatible SC-LDPC Codes
SHI Xiangyi, HAN Tongzhou, TIAN Hai, ZHAO Danfeng
2022, 20(S1):  16-21.  doi:10.12142/ZTECOM.2022S1003
Asbtract ( 48 )   HTML ( 170)   PDF (1213KB) ( 50 )
Figures and Tables | References | Related Articles | Metrics

This paper proposes a family of raptor-like rate-compatible spatially coupled low-density parity-check (RL-RC-SC-LDPC) codes from RL-RC-LDPC block codes. There are two important keys. One is the performance of the base matrix. RL-LDPC codes have been adopted in the technical specification of 5G new radio (5G-NR). We use the 5G NR LDPC code as the base matrix. The other is the edge coupling design. In this regard, we have designed a rate-compatible coupling algorithm, which can improve performance under multiple code rates. The constructed RL-RC-SC-LDPC code property requires a large coupling length $L$ and thus we improved the reciprocal channel approximation (RCA) algorithm and proposed a sliding window RCA algorithm. It can provide lower complexity and latency than RCA algorithm. The code family shows improved thresholds close to the Shannon limit and finite-length performance compared with 5G NR LDPC codes for the additive white Gaussian noise (AWGN) channel.

Derivative-Based Envelope Design Technique for Wideband Envelope Tracking Power Amplifier with Digital Predistortion
YI Xueya, CHEN Jixin, CHEN Peng, NING Dongfang, YU Chao
2022, 20(S1):  22-26.  doi:10.12142/ZTECOM.2022S1004
Asbtract ( 58 )   HTML ( 10)   PDF (878KB) ( 88 )
Figures and Tables | References | Related Articles | Metrics

A novel envelope design for an envelope tracking (ET) power amplifier (PA) based on its derivatives is proposed, which can trade well off between bandwidth reduction and tracking accuracy. This paper theoretically analyzes how to choose an envelope design that can track the original envelope closely and reduce its bandwidth, and then demonstrates an example to validate this idea. The generalized memory polynomial (GMP) model is applied to compensate for the nonlinearity of ET PA with the proposed envelope design. Experiments are carried out on an ET system that is operated with the center frequency of 3.5 GHz and excited by a 20 MHz LTE signal, which show that the proposed envelope design can make a good trade-off between envelope bandwidth and efficiency, and satisfactory linearization performance can be realized.

End-to-End Chinese Entity Recognition Based on BERT-BiLSTM-ATT-CRF
LI Daiyi, TU Yaofeng, ZHOU Xiangsheng, ZHANG Yangming, MA Zongmin
2022, 20(S1):  27-35.  doi:10.12142/ZTECOM.2022S1005
Asbtract ( 100 )   HTML ( 10)   PDF (436KB) ( 90 )
Figures and Tables | References | Related Articles | Metrics

Traditional named entity recognition methods need professional domain knowledge and a large amount of human participation to extract features, as well as the Chinese named entity recognition method based on a neural network model, which brings the problem that vector representation is too singular in the process of character vector representation. To solve the above problem, we propose a Chinese named entity recognition method based on the BERT-BiLSTM-ATT-CRF model. Firstly, we use the bidirectional encoder representations from transformers (BERT) pre-training language model to obtain the semantic vector of the word according to the context information of the word; Secondly, the word vectors trained by BERT are input into the bidirectional long-term and short-term memory network embedded with attention mechanism (BiLSTM-ATT) to capture the most important semantic information in the sentence; Finally, the conditional random field (CRF) is used to learn the dependence between adjacent tags to obtain the global optimal sentence level tag sequence. The experimental results show that the proposed model achieves state-of-the-art performance on both Microsoft Research Asia (MSRA) corpus and people’s daily corpus, with F1 values of 94.77% and 95.97% respectively.

Intelligent Antenna Attitude Parameters Measurement Based on Deep Learning SSD Model
FAN Guotian, WANG Zhibin
2022, 20(S1):  36-43.  doi:10.12142/ZTECOM.2022S1006
Asbtract ( 28 )   HTML ( 2)   PDF (1312KB) ( 9 )
Figures and Tables | References | Related Articles | Metrics

Due to the consideration of safety, non-contact measurement methods are becoming more acceptable. However, massive measurement will bring high labor-cost and low working efficiency. To address these limitations, this paper introduces a deep learning model for the antenna attitude parameter measurement, which can be divided into an antenna location phase and a calculation phase of the attitude parameter. In the first phase, a single shot multibox detector (SSD) is applied to automatically recognize and discover the antenna from pictures taken by drones. In the second phase, the located antennas’ feature lines are extracted and their attitude parameters are then calculated mathematically. Experiments show that the proposed algorithms outperform existing related works in efficiency and accuracy, and therefore can be effectively used in engineering applications.

Multi-Task Learning with Dynamic Splitting for Open-Set Wireless Signal Recognition
XU Yujie, ZHAO Qingchen, XU Xiaodong, QIN Xiaowei, CHEN Jianqiang
2022, 20(S1):  44-56.  doi:10.12142/ZTECOM.2022S1007
Asbtract ( 37 )   HTML ( 4)   PDF (1899KB) ( 62 )
Figures and Tables | References | Related Articles | Metrics

Open-set recognition (OSR) is a realistic problem in wireless signal recognition, which means that during the inference phase there may appear unknown classes not seen in the training phase. The method of intra-class splitting (ICS) that splits samples of known classes to imitate unknown classes has achieved great performance. However, this approach relies too much on the predefined splitting ratio and may face huge performance degradation in new environment. In this paper, we train a multi-task learning (MTL) network based on the characteristics of wireless signals to improve the performance in new scenes. Besides, we provide a dynamic method to decide the splitting ratio per class to get more precise outer samples. To be specific, we make perturbations to the sample from the center of one class toward its adversarial direction and the change point of confidence scores during this process is used as the splitting threshold. We conduct several experiments on one wireless signal dataset collected at 2.4 GHz ISM band by LimeSDR and one open modulation recognition dataset, and the analytical results demonstrate the effectiveness of the proposed method.

Multi-Cell Uplink Interference Management: A Distributed Power Control Method
HU Huimin, LIU Yuan, GE Yiyang, WEI Ning, XIONG Ke
2022, 20(S1):  56-63.  doi:10.12142/ZTECOM.2022S1008
Asbtract ( 41 )   HTML ( 7)   PDF (1395KB) ( 184 )
Figures and Tables | References | Related Articles | Metrics

This paper investigates a multi-cell uplink network, where the orthogonal frequency division multiplexing (OFDM) protocol is considered to mitigate the intra-cell interference. An optimization problem is formulated to maximize the user supporting ratio for the uplink multi-cell system by optimizing the transmit power. This paper adopts the user supporting ratio as the main performance metric. Our goal is to improve the user supporting ratio of each cell. Since the formulated optimization problem is non-convex, it cannot be solved by using traditional convex-based optimization methods. Thus, a distributed method with low complexity and a small amount of multi-cell interaction is proposed. Numerical results show that a notable performance gain achieved by our proposed scheme compared with the traditional one is without inter-cell interaction.

SVM for Constellation Shaped 8QAM PON System
LI Zhongya, CHEN Rui, HUANG Xingang, ZHANG Junwen, NIU Wenqing, LU Qiuyi, CHI Nan
2022, 20(S1):  64-71.  doi:10.12142/ZTECOM.2022S1009
Asbtract ( 63 )   HTML ( 1)   PDF (3357KB) ( 35 )
Figures and Tables | References | Related Articles | Metrics

Nonlinearity impairments and distortions have been bothering the bandwidth constrained passive optical network (PON) system for a long time and limiting the development of capacity in the PON system. Unlike other works concentrating on the exploration of the complex equalization algorithm, we investigate the potential of constellation shaping joint support vector machine (SVM) classification scheme. At the transmitter side, the 8 quadrature amplitude modulation (8QAM) constellation is shaped into three designs to mitigate the influence of noise and distortions in the PON channel. On the receiver side, simple multi-class linear SVM classifiers are utilized to replace complex equalization methods. Simulation results show that with the bandwidth of 25 GHz and overall bitrate of 50 Gbit/s, at 10 dBm input optical power of a 20 km standard single mode fiber (SSMF), and under a hard-decision forward error correction (FEC) threshold, transmission can be realized by employing Circular (4, 4) shaped 8QAM joint SVM classifier at the maximal power budget of 37.5 dB.

Review
General Introduction of Non-Terrestrial Networks for New Radio
HAN Jiren, GAO Yin
2022, 20(S1):  72-78.  doi:10.12142/ZTECOM.2022S1010
Asbtract ( 111 )   HTML ( 10)   PDF (1266KB) ( 79 )
Figures and Tables | References | Related Articles | Metrics

In the new radio (NR) access technology, non-terrestrial networks (NTN) are introduced to meet the requirement of anywhere and anytime connections from the world market. With the introduction of NTN, the NR system is able to offer the wide-area coverage and ensure the service availability for users. In this paper, the general aspects of NTN are introduced, including the NTN architecture overview, the impact of NTN on next-generation radio access network (NG-RAN) interface functions, mobility scenarios and other NTN related issues. The current progress in 3GPP Release 17 is also provided.