High-speed Data Processing Research Articles

Implementation of an SDR system using an MPI-based GPU cluster for WiMAX and LTE

This paper presents an implementation of a 2 × 2 Multi-Input Multi-Output Software Defined Radio (SDR) Base Station system using a Message Passing Interface (MPI)-based Graphic Processing Unit (GPU) cluster as its modem processor for a high-speed data processing. Recently, GPUs have been widely researched especially for SDR systems because of their capability for exploiting parallel processing using a large number of Arithmetic Logic Units. MPI-based GPU clusters have been adopted in order to further increase performance capability. From our experimental results, it has been found that the implemented system consisting of three GPU nodes can enhance the modem speed by more than 2.5 times compared to a single GPU system. A dual-mode Mobile Device (MD) prototype supporting Worldwide Interoperability for Microwave Access and Long Term Evolution communications systems is implemented. In our design, one of the two waveforms can automatically be selected by the MD itself using a dual-mode controller that determines the reconfiguration of the MD modem depending on the received signal quality.

Ultra high capacity 1.28 Tbps DWDM system design and simulation using optimized modulation format

In this paper, an ultra high capacity optical system with transmission capacity of 1.28Tbps using optimized modulation format is presented. The system design is achieved by using a 64-channel 20Gbps dense wavelength division multiplexing (DWDM) technique. The proposed optimized modulation format has narrower bandwidth than conventional intensity modulation (IM) schemes. It has also high dispersion tolerance. Thus, makes it possible to attain an ultra high capacity long distance transmission. The performance characteristics of the system for different modulation formats are compared through simulation at 20Gbps. The system is also analyzed for pre, post and symmetrical dispersion compensation techniques. We have also investigated the filtering performance of the system, using narrow-band optical filters at each input and output channel, to improve the coverage range of the proposed system. These filters have been used for analysing the DWDM system performance, for channel spacings of 100GHz and 50GHz. The system is designed and simulated for a coverage distance of 200km and more. The proposed ultra high capacity system is expected to be more technically viable due to the use of optimized modulation format. Results and analysis of the proposed system are also presented. This system will be highly useful for high speed data processing, in the present communication and computer networks.

Key technologies of dynamic information database for power systems

In the paper,on the basis of analyzing the structure of dynamic information database,and in combination with the feature of the power system,the key technologies of concurrency data processing,memory-mapped file,disk cache management mechanism and associated data storage were discussed,and the data sampling flow and hybrid compression algorithm were also introduced in detail.The application case in the automatic system of power grid dispatching was introduced and the result proves that the dynamic information database can meet the performance requirement of high-speed data processing.

Content security protection for remote sensing images integrating selective content encryption and digital fingerprint

Content security protection methods should be adopted to guarantee the security of highly sensitive remote sensing images during their transmission and usage. The joint fingerprinting and decryption (JFD) framework, which integrates encryption and fingerprinting, can provide comprehensive, effective content security protection for remote sensing information. However, several challenges need to be overcome. In order to solve the contradictory relationship between encryption security and fingerprinted image quality in JFD, a selective content encryption strategy is proposed to enhance encryption security and a new method for choosing fingerprint embedding area is proposed to reduce partial decryption's influence on image quality. Furthermore, a new method for generating and controlling fingerprints is introduced to JFD to eliminate its vulnerability to collusion attacks. The experiment results show that the proposed techniques are highly effective. Because of its high security, good fingerprinted image quality and high data processing and transmission speed, the new JFD scheme is an effective content security protection approach for remote sensing images, especially suitable for massive remote sensing images.

The Implementation Of Cellular Automata With Non-identical Rule On Serial Base

A cellular automaton is a calculation method which strongly needs high-speed processing of data. In cellular automata, a very large volume of data must be processed in a short time, so that the results of cellular automata can be used. Different methods can be used to implement cellular automata, among which the implementation of cellular automata on serial bases is one of the simplest ones. Two types of implementing cellular automata on a serial base were studied in this paper. In a method of implementation, the whole cells of automata the use the same rule; and in another method, the rule related to cells is different depending on the location. Then, the speed of implementation of the two methods is compared: it was shown that the speed of operations in the serial implementation is independent of cellular automata rule.

A new high precision absolute gravimeter

The accurate measurement of local gravitational acceleration (g, normal value 9.81 m/s2) is a key approach to the exploration of the gravitational field of the earth, and it has been applied in metrology, geodesy, geodynamics, seismology, and mineral exploration. With the plan of '2000 national gravity network' and 'China crustal motion observation network', highly precision gravity measurement will be increasingly required in the long term. In order to further investigate the possible systematic error of current absolute gravity measurement, and satisfy various requirements for highly accurate gravity observation, the T-1 absolute gravimeter prototype is designed and built at Tsinghua University. The T-1 instrument adopts the classic free-fall scheme to measure the g value by tracking the trajectory of a free-fall motion in vacuum with a laser interferometer. The T-1 absolute gravimeter consists of several sub-systems: high vacuum free-fall chamber, compact laser interferometer, low-frequency vertical vibration isolation, high-speed signal acquisition system, instrument control and data processing system. The length and time standards used in g measurement are stabilized He-Ne laser and rubidium atomic clock respectively. It's well known that both of the two standards have an uncertainty of less than 1 10-9. The practical gravity measurement results of T-1 absolute gravimeter give a standard deviation of the mean of less than 1 upGal (1 up Gal = 10-8 m/s2) in a typical observation within 12 h. And the reproducibility of less than 3 upGal is verified by repeatedly measuring gravity. The T-1 absolute gravimeter is a promising instrument by which highly accurate gravity measurement at a microgal level can be realized, and it is hopefully to be used in many research and application areas.

Driving magnetic order in a manganite by ultrafast lattice excitation

Optical control of magnetism, of interest for high-speed data processing and storage, has only been demonstrated with near-infrared excitation to date. However, in absorbing materials, such high photon energies can lead to significant dissipation, making switch back times long and miniaturization challenging. In manganites, magnetism is directly coupled to the lattice, as evidenced by the response to external and chemical pressure, or to ferroelectric polarization. Here, femtosecond mid-infrared pulses are used to excite the lattice in La0.5Sr1.5MnO4 and the dynamics of electronic order are measured by femtosecond resonant soft x-ray scattering with an x-ray free electron laser. We observe that magnetic and orbital orders are reduced by excitation of the lattice. This process, which occurs within few picoseconds, is interpreted as relaxation of the complex charge-orbital-spin structure following a displacive exchange quench - a prompt shift in the equilibrium value of the magnetic and orbital order parameters after the lattice has been distorted. A microscopic picture of the underlying unidirectional lattice displacement is proposed, based on nonlinear rectification of the directly-excited vibrational field, as analyzed in the specific lattice symmetry of La0.5Sr1.5MnO4. Control of magnetism through ultrafast lattice excitation has important analogies to the multiferroic effect and may serve as a new paradigm for high-speed optomagnetism.

Overlapped message passing technique with resource sharing for high speed CMMB LDPC decoder

Today, demands for high speed digital communication are getting bigger due to wide deployment of 4G mobile communications and mobile TV services. Requirement for high speed data processing makes forward error correction crucial. LDPC is one of the most popular error correcting codes. In this paper, we propose a novel LDPC decoder for the China Multimedia Mobile Broadcasting (CMMB) standard. The LDPC decoding is carried out iteratively, which leads to relatively long decoding latency. Also, due to long code words, the amount of required memory is huge. To resolve these issues, we propose a novel Overlapped Message Passing (OMP) algorithm with an efficient resource sharing technique. Using the proposed method, we find the best permuted parity check matrix of the CMMB to improve the throughput while minimizing the memory requirement. Using the OMP algorithm only, we could improve the performance by 10%. To avoid potential memory access conflicts, a memory grouping technique to improve pipelining performance is also proposed. By applying all the techniques that we propose in this paper, we could improve the performance up to by 451.

Design and Realization of High-Speed and Real-Time Signal Processing System Based on FPGA

We usually adopt DSP and ARM to perform signal collection and processing. However, functions of DSP and ARM largely depend on the software. Parameter modulation of the software must account for sampling time partly. Moreover, software can not control complex peripheral logic circuit very well. The above problems render DSP and ARM containing big flaws in high-speed data collection and processing. Field-programmable Gate Array (FPGA) possesses the characteristics of timeliness, controllability and rapid processing speed. The paper designed a new circuit based on FPGA to obtain high-speed signal, and utilized FFT IP core in FPGA to perform spectral analysis with the help of digital signals from the circuit. The paper introduced the design based on FPGA and verified that it possesses good performance according to actual experiments.

Incremental expectation maximization principal component analysis for missing value imputation for coevolving EEG data

Missing values occur in bio-signal processing for various reasons, including technical problems or biological characteristics. These missing values are then either simply excluded or substituted with estimated values for further processing. When the missing signal values are estimated for electroencephalography (EEG) signals, an example where electrical signals arrive quickly and successively, rapid processing of high-speed data is required for immediate decision making. In this study, we propose an incremental expectation maximization principal component analysis (iEMPCA) method that automatically estimates missing values from multivariable EEG time series data without requiring a whole and complete data set. The proposed method solves the problem of a biased model, which inevitably results from simply removing incomplete data rather than estimating them, and thus reduces the loss of information by incorporating missing values in real time. By using an incremental approach, the proposed method also minimizes memory usage and processing time of continuously arriving data. Experimental results show that the proposed method assigns more accurate missing values than previous methods.

Design and FPGA Implementation of Systolic Array Architecture for Matrix Multiplication

ABSTRACT The evolution of computer and Internet has brought demand for powerful and high speed data processing, but in such complex environment fewer methods can provide perfect solution. To handle above addressed issue, parallel computing is proposed as a solution to the contradiction. This paper provides solution for the addressed issues of demand for high speed data processing. This paper demonstrates an effective design for the Matrix Multiplication using Systolic Architecture on Reconfigurable Systems (RS) like Field Programmable Gate Arrays (FPGAs). Here, the systolic architecture increases the computing speed by combining the concept of parallel processing and pipelining into a single concept. Here, the RTL code is written for matrix multiplication with systolic architecture and matrix multiplication without systolic architecture in Verilog HDL, compiled and simulated by using Modelsim XE III 6.4b, Synthesized by using Xilinx ISE 9.2i and targeted to the device xc3s500e-5-ft256 and then finally the designs are compared to each other to evaluate the performance of proposed architecture. The proposed Matrix Multiplication with systolic architecture is enhances the speed of matrix multiplication by twice of conventional method.

3D optical Yagi–Uda nanoantenna array

Future photonic circuits with the capability of high-speed data processing at optical frequencies will rely on the implementation of efficient emitters and detectors on the nanoscale. Towards this goal, bridging the size mismatch between optical radiation and subwavelength emitters or detectors by optical nanoantennas is a subject of current research in the field of plasmonics. Here we introduce an array of three-dimensional optical Yagi–Uda antennas, fabricated using top-down fabrication techniques combined with layer-by-layer processing. We show that the concepts of radiofrequency antenna arrays can be applied to the optical regime proving superior directional properties compared with a single planar optical antenna, particularly for emission and reception into the third dimension. Measuring the optical properties of the structure reveals that impinging light on the array is efficiently absorbed on the subwavelength scale because of the high directivity. Moreover, we show in simulations that combining the array with suitable feeding circuits gives rise to the prospect of beam steering at optical wavelengths.

Design of a High-speed Data Processing Channel based on DSD for Digital Audio System

Abstract With the development of digital audio technology, recently, a so-called DSD(direct stream digital) coding technique becomes a hot subject of research. Now, engineers all over the world are researching on digital power amplifier and digital speaker, hoping that 1-bit stream in the form of DSD can be amplified by a digital power amplifier and then the amplified signal can be passed to a digital speaker directly so that the high-fidelity of digital audio can be achieved. In the procedure of development, the demand of high-speed is placed on the process of filtering and noise-shaping. In order to fulfill this demand, a high-speed data processing channel based on DSD is presented in this paper, the speed of data processing is enhanced effectively by using RNS FIR filter in DSP.

Design of the Control Circuit Based on FPGA for the Underwater Trash Detection System

For the detection of underwater trashes, the ultrasonic detection system is a necessary deceive. In the traditional ultrasonic detection system, echo signal is generally processed by analog circuit, and only the distance information is obtained. In order to determine the distance and material of target, the amplitude of echo wave also need to be extracted, and FPGA is employed to control high-speed AD and process data. The design of control circuit based on FPGA is presented here. The control circuit includes modules of clock, high-speed data processing, communication and transmitting control. On the basis of the system design, simulation, implementation, the system experiments are completed. The results show: the functions of the control circuit based on FPGA are implemented correctly, but the system power consumption needs to be reduced more.

基于PSD的新型触摸板技术

In order to meet the big screen display's requirements on touchpad and overcome the disadvantages of interactive touchpad on property and cost,a new type of touchpad technology was invented which was an application of optical system imaging principle as well as optical sensing device of PSD.The infrared spotlight source of contact point of touch pen and touch board imaging on the optical sensor through simple optical system consisted of optical lens and the location of the touch point will be determined by data processing such as coordinates correction.The low cost,easy integration,high sensitivity,high speed of data processing and simple system structure of the new touchpad can be applied to the big size and proportion of the touchpad.

Digital signal processor-based three-dimensional wavelet error-resilient lossless compression of high-resolution spectrometer data

Ultraspectral sounders revolutionize the way remote sensing data are collected, retrieved, and assimilated to provide better weather, climate, and environmental prediction and monitoring. This unprecedented amount of ultraspectral data increases the burden of data bandwidth and the chance of transmission noise and error contamination. It is desired that source coding of ultraspectral data also have some error-resilient capability, in addition to the error-correcting channel coding. Earlier, we developed three-dimensional wavelet reversible variable-length coding (3DWT-RVLC) for lossless compression of ultraspectral sounder data, which has significantly better error resilience than JPEG2000 Part 2 at only a small reduction in compression gain. The reversible variable-length codes allow instantaneous decoding in both directions, which affords better detection of bit errors due to synchronization losses over a noisy channel. To explore the feasibility of 3DWT-RVLC for real-time ultraspectral data processing, we implement a memory-limited digital signal processor (DSP) version of 3DWT-RVLC. Compression experiments on 10 ultraspectral test granules obtained from the NASA Atmospheric Infrared Sounder show that the memory-limited DSP-based 3DWT-RVLC is able to perform high-speed data processing at only a small reduction in compression ratio as compared to the original 3DWT-RVLC.

Status report on a MicroTCA card for HCAL trigger and readout at SLHC

We present recent measurements performed using a prototype MicroTCA card for CMS-HCAL Trigger and Readout at SLHC. Our second generation prototype uses a Xilinx XC5VFX70T FPGA to perform the high-speed communication and data processing for up to eight Readout Module fibers that are streaming data at 4.8 Gbps each. The FPGA also uses two SFP+ optical interfaces at 6.4 Gbps each for data transfer to the Trigger System. A local DAQ interface in the FPGA communicates via Gigabit Ethernet with the MicroTCA MCH. Bit Error Rate Test (BERT) results and data integrity analyses are presented in challenging clocking environments including a legacy TTC system. In addition, the status of the IPbus concept for control of deeply embedded devices is presented.

Interconnect Technologies and Advanced Packaging for High Performance Computing

The quest for high speed data processing continues to drive the industry to increase L1 and L2 on -chip caches and near-processor memories in large-scale computer systems. However, the increased latency and process yields limit chip area for L1 and L2 caches. Therefore alternative solutions beyond on-chip memory are being aggressively pursued. In this paper a high-level computer memory hierarchy and the corresponding interconnect technologies on each level of memory structures are summarized. We introduce a new MCM package technology based on proximity communication (PxC-MCM package) and discuss its merits compared to conventional MCM. Challenges and recent research results on PxC-MCM solution are presented as well.

Dynamic tensile testing of fabric–cement composites

Dynamic tensile tests were conducted using a high speed servo-hydraulic testing machine on three types of fabric reinforced cement composites. The high speed testing procedure and data processing method are presented. Quasi-static tests were also conducted on the composites. Effects of strain rate on the mechanical properties of fabric–cement composites are noted. A good correlation was found between the properties of the fabrics and the composites, with the carbon fabric exhibiting the highest strength and ductility performance in high speed tensile tests. The differences in tensile behavior of the various composites were correlated with the differences in the role of the fabric materials. Composites tested under high speed loading exhibited different responses as compared to similar composites tested under quasi-static condition.

Dynamic application of digital image and colour processing in characterizing flame radiation features

In this work, the experimental investigation of the dynamic flame properties of flame flickering and equivalence ratio sensing of a combustion process was done. In particular, the time-varied flame properties were examined using a novel digital image and colour processing methodology. This technique makes use of the observed correlation between a digital image colour signal and physical flame radiation characteristics in the visible wavelength domain. Aspects of RGB and HSV colour modelling principles were applied to show that the addition of colour identification in the image processing of high-speed flame image data could yield three useful parameters which are related to the dynamic behaviour of different flame emanating components. First, the validity of the colour identities for tracking the yellowish-red diffusion and greenish-blue premixed flame colourations were examined by comparing their respective flickering frequency profiles. Then, the usefulness of the extracted Rdiffusion, Gpremixed and Bpremixed colour signals to abstractly represent the behaviour of soot, C2* and CH* emission characteristics in a dynamic flame transition from diffusion to stoichiometric premixed condition was demonstrated. In particular, the colour signal ratio Bpremixed/Gpremixed was correlated to exemplify the approximate time-varied state of the equivalence ratio from the imaged combustion phenomenon.