The main advantage of tight frame-based time-frequency representations is to warrant a robust and stable signals reconstruction. In this paper we propose to study the speech reconstruction stability, in distortion sense, of three tight frame-based representations: the tight framelet transform, the higher density tight framelet transform and the tight framelet packet transform. The reconstruction stability is assessed using objective criteria. The results are compared with the critically wavelet packet transform. We perceive that the representations satisfying the frame theory conditions are more stable and resistant to distortions.
The study of the human handwriting process shows that bell-shaped velocity profiles are generally observed in the handwriting motion. In this paper, a mathematical model, using input and output data for forearm muscles activities, named ElectroMyoGraphic signals (EMG), and velocities of a pen tip moving on (x, y) plan, is proposed. Using the outputs of this mathematical velocity model, we also propose a novel bi-axis control algorithm to generate handwritten graphic traces. The effectiveness of this approach should be observed on different cursive Arabic letters and geometric shapes constituting the recorded experimental basis.
Cognitive radio (CR) is a kind of “access technology” used to co-exist among more than one radio technology without (or with minimal) interference to each other. Typically, there is one primary user of the spectrum, whereas other users are “allowed” to communicate as long as they don’t interfere the primary user. One of the well-known techniques is to sense the spectrum before initiating the communication in order to analyze the occupancy of radio frequency spectrum. In this paper, we present a comparative analysis between Energy Detection (ED) and Maximum Eigenvalue Detection (MED) spectrum sensing techniques.The performance of these two methods is evaluated in terms of their Receiver Operating Characteristics (ROC) and their detection probability for different values of Signal to Noise Ratio (SNR).
Nowadays, underwater vehicles (UVs) are greatly developed in almost domains. However, using this kind of vehicles for in-service oil tank inspection is not yet advanced enough. As part of an effort to design a new class of underwater vehicles suitable for inspection, the present work lays down requirements devoted to each component of those vehicles. A detailed design process is introduced and the main parts of the vehicle are described with emphasis on its related constraints. Finally prototype implementation is discussed and contributions of ROS middleware are highlighted.
Breast cancer is now one of the main sources of death among women throughout the world. The detection of cancerous tumors in mammographic images is a great challenge for radiologists and physicians. It requires precise, experience and time. Fortunately, the evolution of science enabled the development of medical imaging techniques which are very efficacious for the benefit of health in order to detect any abnormalities in the breast parenchyma. It is in this context that our work fits. In fact, in this paperwe have implemented a computer-assisted diagnostic system (CAD) aiming at facilitating the early detection of this cancer. The developed method proceeds in three steps: pretreatment mammograms to enhance the segmentation step, segmentation based on morphological operators to detect breast masses. Finally, the segmentation of the breast contour based on the method of Marching Square was adopted. Our approach has been tested on the basis of pictures “MIAS” showing its efficiency. In addition, a graphical interface was performed to facilitate the task for radiologists.
In this work a proposition has been studied in order to check the Terrestrial Digital Video Broadcasting (DVB-T) enhancement. It deals with overcoming the effect of Peak-to-Average Power Ratio (PAPR) drawback on wireless system performances. For this purpose, a special averaging technique has been introduced in this work and consists of three different stages; signal noise removal, adaptively peak detection, and averaging filtration, consequently. In order to check the work efficiency, a MATLAB simulation has been conducted at the same bandwidth and channel conditions to deals with both of BER curves and CCDF curves. As a result, the proposed work shows promising results over the proposed work in the literature and over our previously published work. The proposed work gives extra immunity against the channel noise reaches up to 25% over the literature. Furthermore, it gives extra 15% reduction in combatting the PAPR effect.
In this work a comparison has been made among different proposed algorithms in order to improve the performance of a power consumption wireless network. They areused to combat one of the Orthogonal Frequency Division Multiplexing (OFDM) technique drawbacks, which are considered as a key technique in enhancing the new era of wireless systems’ quality of service (QoS).Therefore, three different propositions have been investigated and covered by this work and classified as; linear coding based, wavelet transformation based, PWM based. Furthermore, a MATLAB program has been run to check their performance and covers two different criteria; the curves of CCDF and the SER curves. This is in order to reduce the Peak-to-Average Power Ratio (PAPR) effect.
In this contribution, we present a multiple-input multiple-output (MIMO) transceiver scheme for very high-speed off-chip communication over low-cost electrical interconnects. Linear MIMO pre-equalization at the transmitter is combined with decision feedback equalization (DFE) at the receiver to counteract the adverse effect of inter symbol interference (ISI) and crosstalk (XT). Neat closed-form expressions for the minimum mean square error equalization filters are derived under a total transmit energy constraint. Numerical analysis shows that the combination of linear MIMO pre-equalization and MIMO DFE allows to significantly improve the reliability of future high-speed off-chip communication.