Applied Mechanics and Materials Vols. 333-335

Abstract: Combining complex FastICA and L-shaped array DOA (Direction of Arrival) estimation, a new algorithm is proposed. Firstly, the received signal is whitened to reduce the computational load, Secondly, the original signal is restored by the ICA algorithm, Thirdly, the generalized separation matrix is inversed to obtain the estimation of the array manifold. Finally, the estimated steering vector is projected to the ideal L-shaped array to get the final results. This algorithm doesnt need a spectral peak search, What is more, it is able to recover the source signal during the estimation. The formed new algorithm can get faster speed without losing the intrinsic advantage of the original DOA estimation algorithm. Simulation results confirm the validity of the new algorithm.

Abstract: Mobile OFDM refers to OFDM systems with fast moving transceivers, contrastive to traditional OFDM systems whose transceivers are stationary or with a low velocity. An efficient implementation of the channel equalization for mobile OFDM is presented in this paper. Based on the particular OFDM subcarrier allocations, the channel equalizer is split into separated sub-equalizers, enabling a concurrent implementation. This parallel equalizer is implemented on an FPGA platform. The experimental results show that without an efficient design, mobile OFDM leads to an unacceptable hardware cost. The proposed parallel equalizer for mobile OFDM can compensate for time-varying channels, in which a traditional OFDM receiver fails to operate, although the paid price is that the hardware resource is reasonably increased.

Abstract: The paper proposes a smart, reliable and robust algorithm for motion detection, tracking and activity analysis. Background subtraction is considered intelligent algorithms for the same. Mount the web camera focused to the patient. PC should have a unique external Internet IP Address. Android mobile phone should be GPRS enabled. GSM technology is used for sending SMS. It is a client-server technology wherein client captures the images, checks for motion if any, discards the packets until motion is detected. Use background subtraction algorithm to check the motion. The surveillance camera does not move and has a capture of the static background it is facing. It uses image subtraction to determine object motion. It provides more reliable information about moving object, but it is so sensitivity to the dynamic changes such as lighting. Once motion is detected, camera stops monitoring further motion. Instead, it starts capturing the video. Simultaneously, SMS alert is sent to the responsible doctors and also alerting the medical staff with audio speaker in the hospital. Java mail API is used to mail the captured video to the entered e-mail IDs. Once the doctor demands for video, socket is established between the PC and the mobile phone and video (series of images) are streamed to the doctors mobile phone. Save live video of first few seconds at the server end for future use. Activate alert at the remote end.

Abstract: A new signal processing method for phase difference estimation was proposed based on time-varying signal model, whose frequency, amplitude and phase are time-varying. And then be applied Coriolis mass flowmeter signal. First, a bandpass filtering FIR filter was applied to filter the sensor output signal in order to improve SNR. Then, the signal frequency could be calculated based on short-time frequency estimation. Finally, by short window intercepting, the DTFT algorithm with negative frequency contribution was introduced to calculate the real-time phase difference between two enhanced signals. With the frequency and the phase difference obtained, the time interval of two signals was calculated. Simulation results show that the algorithms studied are efficient. Furthermore, the computation of algorithms studied is simple so that it can be applied to real-time signal processing for Coriolis mass flowmeter.

Abstract: Polarization filtering is an effective method to process the ocean bottom multi-component seismic data but distort signal amplitude and phase. To solve this problem, a method of polarization filter based on oblique projection operators (OOPF) in process of ocean bottom multi-component seismic data is proposed in this paper. First, we compute polarization parameters based on singular value decomposition (SVD); secondly, an oblique projecting polarization is construct by polarization parameters, finally, the simulation and experiment data results and indicates that the proposed method effectively processing the ocean bottom multi-component seismic.

Abstract: The basic approach of generating arbitrary waveform is the digital synthesis method. Appling this method, one first samples and quantifies a period of the signal waveform to be generized, memorizes the obtained binary waveform data in a memory, and then reads it via hardware circuit in certain sequence, converts the data via digital-to-analog (DA) circuit, finally output the analog waveform through a filter. In this work, an arbitrary waveform generator (AWG) is designed based on the theory of direct digital synthesis (DDS) ^[1-7] and on the analysis of the performance of the output signal. The design uses a field programmable-gate-array (FPGA) ^[3] chip to utilize the AWG. The preset and display of the output frequency and phase are controlled by a micro computer unit (MCU). The artribary waveform data can be downloaded and updated from a communication interface. The AWG can produce a high-resolution arbitrary waveform. The dissertation focuses on hardware circuit design, which has been accomplished, including power supply module, MCU system, high-speed DA converter, and filter, etc. The experimental prototype of the AWG has been made and tested systemly.At the end of the dissertation, the measurement result of the system is given and its error is analyzed. It is shown the AWG can output a sine wave, a triangle wave, a sawteeth wave, or a square wave within the frequency range from 0.01Hz to 15MHz with the step of 10mHz, or output an arbitrary waveform within the range from 0.01Hz to 20kHz.

Abstract: Multiple input multiple output (MIMO) systems could increase wireless communication system capacity enormously. The best optimal detection algorithm for MIMO systems was maximum likelihood (ML) detection algorithm, which could provide the best bit error rate (BER) performance for MIMO systems. However, the computational complexity of ML detection algorithm grew exponentially with the number of transmit antennas and the order of modulation, which resulted in difficult using for practice. A modified MIMO signal detection algorithm which combined ML detection with stack algorithm was presented in this paper. After performing QR decomposition of the channel matrix, the ML detection with length L was done firstly. The partial accumulated metrics were calculated and sorted, which produced an ordered set secondly. Based on the ordered set, stack algorithm was performed to search for the symbol with the minimum accumulated metrics. The proposed algorithm reduced the probability of look back in stack algorithm.

Abstract: OFDM system is sensitive to the frequency offset, by which the orthogonality will be destroyed between the sub-carriers and the Inter-Carrier Interference (ICI) will be generated. Considered that the colored noise is often studied in the dynamic data processing, an adaptive Unscented Kalman Filter (UKF) is presented to estimate and eliminate the frequency offset in this paper. Adaptive factors are introduced into the iteration of the state variance and the noise variance to inhibit the misalignment in the filtering caused by the correlation of the noise. The results of simulation show favorable estimation of the frequency offset, as well as the smoothing effect of the adaptive UKF, which is the evidence of the strong tracking effect of the proposed algorithm.

Abstract: This paper presents an efficient and novel framework for human action recognition based on representing the motion of human body-joints and the theory of nonlinear dynamical systems. Our work is motivated by the pictorial structures model and advances in human pose estimation. Intuitively, a collective understanding of human joints movements can lead to a better representation and understanding of any human action through quantization in the polar space. We use time-delay embedding on the time series resulting of the evolution of human body-joints variables along time to reconstruct phase portraits. Moreover, we train SVM models for action recognition by comparing the distances between trajectories of human body-joints variables within the reconstructed phase portraits. The proposed framework is evaluated on MSR-Action3D dataset and results compared against several state-of-the-art methods.

Abstract: A software defined radio architecture algorithm is applied to Digital BPM processor for its signal processing. The algorithm is evaluated in MATLAB with data acquired from the instrument with a commercial signal source as input.The data flow diagram of the algorithm is shown and explained, critical parameters of the filters are displayed and the results are shown which testify the feasibility of the method.