Li Shufang

Quantitative spectrum occupancy evaluation in China: based on a large scale concurrent spectrum measurement

Abstract Due to the usable frequency becomes more and more crowed, dynamic spectrum access (DSA) is a new hope to solve this problem. However, DSA in China requires a quantitative analysis of the current spectrum utilization in frequency, temporal and spatial domains. In order to free the precious spectrum, spectrum regulation organizations must have a clear, detailed, up-to-date understanding of where, how and by whom spectrum is currently being used—such data is essential to sound policy decisions in the context of cognitive radio (CR). In this paper, a concurrent spectrum occupancy measurement in south China was conducted to evaluate the practical spectrum occupancy with a digital wideband receiver covering from 20 MHz to 3 GHz. We also propose systemic spectrum measurement methodology, matrix format data storage, duty cycle (DC) evaluation metric and data mining process which can be a guideline for other researchers when they conduct the similar experiments. Quantitative analysis and characterization of the 4 different measurement locations are evaluated to promote the popularization of CR application in China. And a uniform Beta distribution channel occupancy model is also validated using real-scene measurement data. The experimental results demonstrate that there is a significant scope for license-exemption use of the released spectrum using CR technology.

Design of wideband dual-band dual-polarized dipole for base station antenna

Abstract This article put forward a novel dual-band dual-polarized magneto-electric dipole antenna excited by F-shaped strips. The proposed antenna achieved a common impedance bandwidth of 25.5% and 39.5% in the lower and the upper bands at both input ports, ranged from 0.75 GHz to 0.97 GHz and from 1.73 GHz to 2.59 GHz respectively. The antenna has good performance in isolation, which is more than −32 dB between the two input ports, and the gain of the antenna is average 4.3 dB and 7.8 dB in the lower and upper bands. The antenna has a stable broadside radiation pattern with low cross polarization and low back lobe radiation over the operating band. Metallic side walls on the ground are added for better performance in gain and radiation pattern.

Novel error resilient tag estimation for RFID system in interference scenarios

Abstract In a radio frequency identification (RFID) system, the backscattered signal is small and prone to interference. The performance of RFID tag identi?cation in interference scenarios is degraded compared to that in error-free scenarios. In this paper, a novel Mahalanobis distance estimate (MDE) method is proposed to jointly estimate the number of tags and packet error rate (PER). The MDE method is error resilient owing to its ability to achieve a stable estimation when interference is impairing the observed information. The proposed method achieves significantly enhanced accuracy over existing methods by taking all the information and correlations among the observed results into account. The MDE method improves the estimate performance based on efficient decorrelation and classification of the observed information. Moreover, the performance of the PER estimate is analyzed both in theory and through simulations. It can be concluded from the analysis that the estimated PER is unbiased and variance-bounded. Simulations show that the proposed estimate outperforms the previous proposals in terms of accuracy and stability, which makes it suitable for application in interference scenarios.

Anomaly detection in smart grid based on encoder-decoder framework with recurrent neural network

Abstract Anomaly detection in smart grid is critical to enhance the reliability of power systems. Excessive manpower has to be involved in analyzing the measurement data collected from intelligent motoring devices while performance of anomaly detection is still not satisfactory. This is mainly because the inherent spatio-temporality and multi-dimensionality of the measurement data cannot be easily captured. In this paper, we propose an anomaly detection model based on encoder-decoder framework with recurrent neural network (RNN). In the model, an input time series is reconstructed and an anomaly can be detected by an unexpected high reconstruction error. Both Manhattan distance and the edit distance are used to evaluate the difference between an input time series and its reconstructed one. Finally, we validate the proposed model by using power demand data from University of California, Riverside (UCR) time series classification archive and IEEE 39 bus system simulation data. Results from the analysis demonstrate that the proposed encoder-decoder framework is able to successfully capture anomalies with a precision higher than 95%.

Conversion and Comparison of the Test Results in Fully and Semi-Anechoic Chamber

This paper illustrates the test method of the radiated emission in Semi-Anechoic Chamber (SAC), and the radiated power in Fully Anechoic Chamber (FAC) between 30 MHz and 1 GHz. The test results are conversed and compared by analysis, while the correction factors that influence the performance are given.