Recently, classification and gene selection of DNA microarray data are important in biomedical research. DNA microarray data provide useful information that can be used to discover the complex mechanism of cancer development. DNA computing techniques are alternative approaches to analyze the DNA microarray data. In this paper, we propose the VLSI implementation of a prostate cancer classification processor. The proposed architecture uses parallel and pipelined processing to improve speed and uses cyclic random masks to reduce memory size. We evaluated the prostate cancer classification processor by testing its performance on prostate cancer microarray data. From the experimental results, the proposed architecture reduced the memory size and classification time with little loss of classification accuracy.
A novel gender classification method is presented which fuses information acquired from multiple facial regions for improving overall performance. It is able to compensate for facial expression even when training samples contain only neutral expression. We perform experimental investigation to evaluate the significance of different facial regions in the task of gender classification. Three most significant regions are used in our fusion-based method. The classification is performed by using support vector machines based on the features extracted using two-dimension principal component analysis. Experiments show that our fusion-based method is able to compensate for facial expressions and obtained the highest correct classification rate of 95.33%.
This paper focuses on reproducible low frequency ranges of electrodynamic loudspeaker drivers. The research findings lead to a guide that allows the optimal driver to be selected at the early stages of the design when the consumer electronic products such as home theater systems, digital TVs or PC speakers are being considered. The suggested frequency ranges are based on the acoustics and estimated from statistical data from 1,473 loudspeaker drivers of various sizes. In particular, the lower frequency limit is discussed together with the peak volume displacement which is closely related to the maximum reproducible sound pressure in a loudspeaker.
We developed a broadband, highly sensitive heterodyne THz-wave receiver with a superconductor-insulator-superconductor (SIS) mixer. To fully utilize wide-bandwidth characteristics of the SIS mixer, we pump the SIS mixer with a photonics-based local oscillator (PLO) with fast frequency-sweeping speed. The received THz-wave signal down-converted to an intermediated frequency band is rapidly acquired by a high-speed analog-to-digital converter. The noise level of the developed receiver is equal to or smaller than twenty times the quantum-limited noise level between 225 and 425GHz. With this receiver, the frequency sweep from 225 to 475GHz with frequency resolution of 0.5GHz takes only about 2.5sec. Emission spectra of N2O gases measured with this receiver show its excellent performance.
We propose a new backoff scheme, called exclusive backoff scheme (EBS), for ad hoc mode in IEEE 802.11 wireless LANs. Each station generates a unique backoff number before starting backoff procedure by using timing synchronization function (TSF) timer and a unique user's ID (UID), which makes a collision free access. The EBS shows better performance in the view of channel utilization and packet delay than original IEEE 802.11 CSMA/CA backoff scheme.
Air-coupled ultrasonic waves in the megahertz range are applied for noncontact detection in a pipe wall reduction model. Various widths and depths of pipe wall reductions were used, and the pipes were irradiated with 1.2MHz air-coupled ultrasonic waves. F-mode guide waves were used to detect the defects in the pipe walls. The experimental results confirmed that ultrasonic waves reflecting from areas of wall reduction in the pipes could be detected. A wall reduction of about 10mm could be deduced from the reflected waveforms. A clear correlation between the depth of the wall reduction and the amplitude of the reflected waves was found.
This paper presents a novel microstrip dual-band filter in which a shunt-stub bandpass filter and an open-stub bandstop filter are integrated into a single structure of short length. The bandstop filter is centered within the passband of the bandpass filter, resulting in a filter having dual passbands. Good return loss is obtained using an optimization routine. This method is suitable for designing dual-band filters having wide bandwidths. Good agreement between theory and measurements in respect of low insertion loss and high isolation is demonstrated.
Writer identification is a popular research field in many languages such as English, Persian, Chinese, etc. The approaches of writer identification methods are dependent on the language because different languages letters have different pattern. In this paper, we have presented XGabor filter and proposed a language independent writer identification system. In the feature extraction phase of proposed method, Gabor and XGabor filters are used while in the classification phase, a new classification method is defined that is not based on any kind of distances among feature vectors. The proposed classifier uses the sequence similarity of the Sorted Order of Features (SOF). To measure this similarity, the Longest Common Subsequence (LCS) algorithm is employed. In simulation phase, two databases in different languages have been used. First one consisted of 100 people's Persian handwritings and second one had 30 people's English handwritings. The accuracy of the system was satisfactory in both of them.
This paper describes using notch filters for improving the gain and gain flatness of an UWB CMOS LNA designed with a 0.13µ m CMOS technology. Also, the performance of LNA is examined by replacing the passive inductances in notch filters with Q-enhanced cascade-grounded active inductors to decrease the silicon area. Simulation results illustrate that both passive and active notch filters at 2.4GHz approximately improves the gain flatness by 0.75dB, while the gain performance with passive notch is about 2dB better. In addition, passive notch filters at 2.9GHz, in parallel with a passive and active notch at 5.2GHz for interference reduction, increases the gain by 3dB and 4dB, respectively.