Hou Ligang

Applying Wavelet Support Vector Machine to Analog Circuit Fault Diagnosis

Based on vector wavelet kernel function, a method for analog circuit diagnosis based on genetic algorithm (GA) and least squares wavelet support vector machine (LSWSVM) is proposed. Using wavelet package as a tool for extracting feature, the GA-LSWSVM is then applied to the filet circuit after training by GA; the simulation results have shown that the method can enhances the accuracy and generalization ability.

Anti-Collision Algorithms for Multi-Tag RFID

RFID is one of automatic technology to identify and collect object data quickly through RF digital signals. RFID increases productivity and convenience. RFID is used for hundreds, if not thousands, of applications such as preventing theft of automobiles and merchandise; gaining entrance to buildings; automating parking. But one of the largest disadvantages in RFID system is its low tag (transponder) identification efficiency by tag collision. Collisions are divided into interrogator collisions and tag collisions. Interrogator collisions occur when neighbouring interrogators interrogate a tag simultaneously. Tag collision is the event that the interrogator (reader) cannot identify the data of tag when more than one tag occupies the same communication channel simultaneously. The reason is that whenever two or more users are transmitting on the shared channel simultaneously, a collision occurs and the data cannot be received correctly. This being the case, packets may have to be transmitted and retransmitted until eventually they are correctly received. As the most RFID systems use passive tags, frame sizes are limited in the framed slotted ALOHA algorithm. Especially, since low-functional passive tags can neither detect collisions nor figure out neighboring tags, a tag collision gives rise to the need for a tag anti-collision protocol that enables the recognition of tags with few collisions and also executes in realtime. Active RFID tags contain an on-board battery. They can communicate with interrogator in far distance. Active tags can provide anti-collision by using various combinations of some methods including time scope and frequency scope. When the number of tags is large, for the conventional RFID anti-collision algorithm, the number of slots required to read the tags increases exponentially as the number of tags does. Some methods can solve this problem with complex algorithm consuming long communication time. Based on the analysis above, a good tag collision arbitration protocol for RFID tags should have the following characteristics: First, a interrogator ought to identify all the tags inside its own reading range. Since the interrogator cannot estimate the number of tags precisely, the guarantee of recognizing all tags must be taken into consideration in the design of the tag hard system and anti-collision protocol. Second, a tag should be identified while consuming a small amount of resource, since the tag has low power. Thus, the tag anti-collision protocol must load the tag with the least possible communication time. This paper presents an improved dynamic framed slotted aloha algorithm (IDFSA) that may solve this problem by dividing frequency of tags that is grouping the tags in different Source: Radio Frequency Identification Fundamentals and Applications, Bringing Research to Practice, Book edited by: Cristina Turcu, ISBN 978-953-7619-73-2, pp. 278, February 2010, INTECH, Croatia, downloaded from SCIYO.COM

Research of Fire Detecting System Based on ZigBee Wireless Network

This paper puts forward a strategy of intellectualized fire monitor based on ZigBee transmission. In this system, a low power, multi-parameter composite fire detection node is used, which can detect temperature, smoke concentration and CO gas concentration. The hardware and software implementation are based on 2.4GHz wireless communication chip CC2430. ZigBee protocol is adopted in the system to form reliable wireless communication. The system uses fusion theory to handle the data in order to determine whether the fire occurs.

Implementation of telephone remote control system based on FPGA

This paper presents an intelligent control system with the ability of remote control based on a public telephone communication network. The remote operation with user authority by transmitting password and operation code through telephone network is realized by using a FPGA controller. The main function circuits of the designed system include ring detector, DTMF decoder, pick-phone circuit, audio prompt, and the FPGA controller. The system has highly configurable, stable and reliable feature.

Design and optimization of Serial communication system interface module

Serial communication system has been widely used in data communications and control system because of less hardware resources, anti-jamming ability, and easy to implement features. A FPGA-based high performance Serial communication system interface module which includes full functions of UART16550 is designed and optimized based on the communication protocol and working principles in this paper. Various technologies are adopted during the design and optimization procedure, such as the three always block coding style, EDA optimization, circuit optimization, and so on. The frequency of the optimized design is up to 166MHz, and the power consumptionis reduced to 0.147W by 63.9%. The test data at typical baud-rate of 115200 and the analyzed result by using Matlab are presented. The test results indicate that the optimized design can be communicated correctly and steadily.

Design of 16 bit digital filter used in delta-sigma A/D converter

A digital decimation filter used in Δ-Σ - ADC is designed and introduced in this paper. The designed filter has a multi-stage structure which is comprised of a stage CIC filter, two stage half-band filters, and a stage compensation filter. The CSD coding, “Hognenauer cut-off theory”, frequency response masking approach and some other techniques are used to improve the performance of the chip and reduce the chip area and power consumption. This downsampling filter is achieved by algorithm modeling using MATLAB and hardware implementation using Verilog HDL. The performance indicators raised are achieved.

An universal design of low-power processor based on LED lighting control system by using solar cells

The lighting system posed by the combination of solar photovoltaic power generation and the white LED illumination is an optimal environmental energy-efficient lighting system and an important trend of development in the future. This paper is focusing on the solar energy and the municipal power complementary LED lighting control system, and the research based on the paper is on the processor — the core component of its controller especially in low power state. According to the structural and actual needs of the system, I proposed the synopsis for the processor, presented a reduced instruction set, and completed 5-stage pipeline processor design with peripherals. After finishing RTL design, I optimized the schematic by self clock-gating register in order to reduce the power of clock tree. Finally, I successfully achieved design features through the FPGA This design is the foundation of the controller integrated.

Monte Carlo analysis of a low power domino gate under parameter fluctuation

Using the multiple-parameter Monte Carlo method, the effectiveness of the dual threshold voltage technique (DTV) in low power domino logic design is analyzed. Simulation results indicate that under significant temperature and process fluctuations, DTV is still highly effective in reducing the total leakage and active power consumption for domino gates with speed loss. Also, regarding power and delay characteristics, different structure domino gates with DTV have different robustness against temperature and process fluctuation.

Analog IC Fault Diagnosis based on Wavelet Neural Network Ensemble

A fault diagnosis method for analog IC diagnosis based on wavelet neural network ensemble (WNNE) and Adaboost algorithm, is proposed in this paper. This makes the way of the directory be of use in fault, and enhances the validity of the fault diagnosis. Using wavelet decomposition as a tool for extracting feature, Then, after training the WNNE by faulty feature vectors, the model of the circuit with fault diagnosis system is built. Simulation results have shown that this claim is more effective than wavelet neural network (WNN).