Learning a discriminative high-fidelity dictionary for single channel source separation

Sparse-representation-based single-channel source separation,which aims to recover each source’s signal using its corresponding sub-dictionary,has attracted many scholars’attention.The basic premise of this model is that each sub-dictionary possesses discriminative information about its corresponding source,and this information can be used to recover almost every sample from that source.However,in a more general sense,the samples from a source are composed not only of discriminative information but also common information shared with other sources.This paper proposes learning a discriminative high-fidelity dictionary to improve the separation performance.The innovations are threefold.Firstly,an extra sub-dictionary was combined into a conventional union dictionary to ensure that the source-specific sub-dictionaries can capture only the purely discriminative information for their corresponding sources because the common information is collected in the additional sub-dictionary.Secondly,a task-driven learning algorithm is designed to optimize the new union dictionary and a set of weights that indicate how much of the common information should be allocated to each source.Thirdly,a source separation scheme based on the learned dictionary is presented.Experimental results on a human speech dataset yield evidence that our algorithm can achieve better separation performance than either state-of-the-art or traditional algorithms.

Study on the Fine Configuration of Reservoir in River Facies Oilfield in Bohai Sea Area

The geological conditions of offshore shallow water delta oil reservoirs were complex, with limited well data and large well spacing. Taking A Oilfield in the Bohai Sea Area, China as an example, the target sand body was formed in a shallow water delta sedimentary environment, with well-developed underwater distributary channels and frequent branching and diversion. The reservoir was strong non-uniformity and uneven plane water cut pressure. To this end, based on the existing work of predecessors, combined with seismic, logging, and production dynamics data, and based on the genesis mechanism of shallow water delta reservoirs, the boundary of composite river channels was identified through seismic facies, and logging facies were used to subdivide them into single river levels within the composite river channels. Then, seismic waveform characteristics were applied to track and characterize the plane distribution of single river channels, guiding the efficient development of offshore shallow water delta oil fields and achieving increased storage and production in Bohai Oilfield, China.

Encryption and Decryption of Color Images through Random Disruption of Rows and Columns

In modern society,information is becoming increasingly interconnected through networks,and the rapid development of information technology has caused people to pay more attention to the encryption and the protection of information.Image encryption technology is a key technology for ensuring the security performance of images.We extracted single channel RGB component images from a color image using MATLAB programs,encrypted and decrypted the color images by randomly disrupting rows,columns and regions of the image.Combined with histograms and the visual judgments of encryption images,it is shown that the information of the original image cannot be obtained from the encryption image easily.The results show that the color-image encryptions with the algorithm we used have good effect and fast operation speed.Thus this algorithm has certain practical value.

Novel Scheme for Multiple Channel Pulse Generation from a Single Laser Diodefor 10 Gbit/s Based WDM Systems Using RZ Modulation Format

A new technique for the generation of multi-channel optical pulse from a single laser diode (LD) is presented in this paper. 35 channel pulse source with 6.5 GHz repetition rate per channel and 32.5 GHz channel spacing was generated from a subharmonically hybrid mode-locked two section monolithic laser with enhanced amplitude modulation. The obtained pulse source exhibits high extinction ratio (>10 dB) and low level of root mean square (RMS) phase noise (<0.11 rad) over all channels from 1556 nm to 1565...

Time Modulated Array Antennas: A Review

Time-modulated array(TMA)antennas,introduce the dimension of time into antenna design to control the radiation patterns and frequency spectral characteristics,thus improve the reconfigurability of array antennas and provide multiple functional-ities.They have great application potential in military and civilian fields,such as precision guidance and mobile communication,and are currently a hot spot of academic research.This article provides a review on the fundamentals and applications of TMAs.First,the basic theory and mathematical formulations of TMAs are introduced.Second,the most important applications of TMAs,namely time-modulated phased arrays(TMPA),are discussed from the perspectives of harmonic suppression and harmonic utiliza-tion,which are used for single-beam and multibeam radiation.Then,we survey the combination of TMA with various types of novel antenna arrays,such as single-channel digital beamforming(DBF)arrays,frequency diverse arrays(FDAs),and retrodirective arrays,to create new hardware implementation methods and enhance their performance.Next,recent advances in dedicated integrated chips for TMA,which have played a significant role in driving the progress of TMAs from academic research to practical applications,are presented.Finally,the challenges and prospects for TMAs are discussed,including new research directions and emerging applica-tion scenarios.