Recent Submissions

Now showing 1 - 5 of 31
  • Item
    A Comparison of the Machine Learning Algorithm for Evaporation Duct Estimation
    (Společnost pro radioelektronické inženýrství, 2013-06) Yang, Chao
    In this research, a comparison of the relevance vector machine (RVM), least square support vector machine (LSSVM) and the radial basis function neural network (RBFNN) for evaporation duct estimation are presented. The parabolic equation model is adopted as the forward propagation model, and which is used to establish the training database between the radar sea clutter power and the evaporation duct height. The comparison of the RVM, LSSVM and RBFNN for evaporation duct estimation are investigated via the experimental and the simulation studies, and the statistical analysis method is employed to analyze the performance of the three machine learning algorithms in the simulation study. The analysis demonstrate that the M profile of RBFNN estimation has a relatively good match to the measured profile for the experimental study; for the simulation study, the LSSVM is the most precise one among the three machine learning algorithms, besides, the performance of RVM is basically identical to the RBFNN.
  • Item
    A BIC Based Initial Training Set Selection Algorithm for Active Learning and Its Application in Audio Detection
    (Společnost pro radioelektronické inženýrství, 2013-06) Leng, Yan; Qi, Guang-hui; Xu, Xin-yan
    To construct a classification system or a detection system, large amounts of labeled samples are needed. However, manual labeling is dull and time consuming, so researchers have proposed the active learning technology. The initial training set selection is the first step of an active learning process, but currently there have been few studies on it. Most active learning algorithms adopt random sampling or algorithms like sampling by clustering (SBC) to select the initial training samples. But these two kinds of method would lose their effectiveness in detecting events of small probability. Because sometimes they could not select or select too few samples of the small probability events. To solve this problem, this paper proposes a BIC based initial training set selection algorithm. The BIC based algorithm performs clustering on the whole training set first. Then uses BIC to judge the status of clusters. Finally, it adopts different selection strategies for clusters of different status. Experimental results on two real data sets show that, compared to random sampling and SBC, the proposed BIC based initial training set selection algorithm can efficiently solve the detection problem of small probability events. In the mean time, it has obvious advantages in detecting events of non-small probability.
  • Item
    Analyses of 100 Gbps Coherent System Performances
    (Společnost pro radioelektronické inženýrství, 2013-06) Skoda, Pavel; Radil, Jan; Vojtech, Josef; Hula, Miloslav
    This paper presents the results of laboratory and field testing of coherent 100 Gbps system with DP-QPSK modulation. Several measurements were performed including power budget, nonlinear threshold, spectrum filtration, constellation diagram, interoperability with 10 Gbps lambdas and dispersion compensation type impact. Field tests addressed transmission of 100 Gbps signal as an Alien Wavelength through multivendor network, influence of photonic service parallel to 100 Gbps signal and performance of 100 Gbps system over single fiber bidirectional transmission lines. 100 Gbps system has been found extremely resilient to most classical impairments thanks to advances error coding and compatible with standard 10 Gbps NRZ lambdas and any type of dispersion compensation. The system was also working over single fiber bidirectional lines and in parallel with Photonic Service of time transfer. The paper also shows recent results of single hop test with 100 Gbps system in laboratory environment.
  • Item
    Diamond Based DDR IMPATTs: Prospects and Potentiality as Millimeter-Wave Source at 94 GHz Atmospheric Window
    (Společnost pro radioelektronické inženýrství, 2013-06) Acharyya, Aritra; Datta, Koyel; Ghosh, Raya; Sarkar, Monalisa; Sanyal, Roshmy; Banerjee, Suranjana; Banerjee, J. P.
    Large-signal simulation is carried out in this paper to investigate the prospects and potentiality of Double-Drift Region (DDR) Impact Avalanche Transit Time (IMPATT) device based on semiconducting type-IIb diamond as millimeter-wave source operating at 94 GHz atmospheric window frequency. Large-signal simulation method developed by the authors and presented in this paper is based on non-sinusoidal voltage excitation. The simulation is carried out to obtain the large-signal characteristics such as RF power output, DC to RF conversion efficiency etc. of DDR diamond IMPATT device designed to operate at 94 GHz. The results show that the device is capable of delivering a peak RF power output of 7.01 W with 10.18% DC to RF conversion efficiency for a bias current density of 6.0×10^8 A m-2 and voltage modulation of 60% at 94 GHz; whereas for the same voltage modulation 94 GHz DDR Si IMPATT can deliver only 693.82 mW RF power with 8.74 efficiency for the bias current density of 3.4×10^8 A m-2.
  • Item
    Contrast Enhancement in Poor Visibility Conditions Using Guided Filtering
    (Společnost pro radioelektronické inženýrství, 2013-06) Chowdhry, Danish Ali; Siddiqui, Adil Masood; Touqir, Imran
    In this paper, extraction of atmospheric veil is proposed to enhance the contrast of the images captured under poor visibility conditions. The method based on guided filtering can accurately recover hidden edges, maintain structural similarity (SSIM) to input image and it is effective for both color and gray level images. The proposed algorithm works without prior information about the scene and its complexity is linear function of the input image size. Experimental comparisons with state of the art algorithms demonstrate that our approach can significantly enhance the contrast and restore the visibility in fine details.