Abstract

In this paper, we address the issue of speech polarity detection using strength of impulse-like excitation around epoch. The correct detection of speech polarity is a crucial step for many speech processing algorithms to extract suitable information. Occurrence of errors in the detection of speech polarity could have an impact on the performance of speech systems. Automatic detection of speech polarity has become an important preliminary step for many speech processing algorithms. We propose a method based on the knowledge of impulse-like excitation of speech production mechanism. The impulse-like excitation is reflected across all frequencies including the zero frequency (0 Hz). Using the slope around zero crossings of the zero frequency filtered signal, an automatic speech polarity detection method is proposed. Performance of the proposed method is demonstrated on 8 different speech corpora. The proposed method is compared with the three existing techniques such as gradient of the spurious glottal waveforms (GSGW), oscillating moments-based polarity detection (OMPD) and residual excitation skewness (RESKEW). From the experimental results, it is observed that the performance of the proposed method is comparable or better than the existing methods for the experiments considered.

Abstract

In this research paper several interesting themes related to Hopfield network are explored and concrete results are derived. For instance, convergence in partial parallel mode of operation, convergence when synaptic weight matrix, W is row diagonally dominant case. Also, structured recurrent Hopfield network with companion matrix as W is studied. The dynamics when the activation function is a step function, the state space is asymmetric hypercube are also studied. Finally several experimental results are derived.

Abstract

The fifth generation (5G) networks employing mil- limeter wave (mmwave) technology will gain the advantages of short distance directive propagation and more spectrum for communication. Device-to-Device communication (D2D) when integrated with mmWave technology can reach the demand for high capacity. In this paper, resource allocation for underlay D2D in outdoor mmWave scenario is presented. We emphasize on fair distribution of resources in a cell while maximizing the spectral efficiency. Also, we discuss the aspects in ad-hoc and cellular communication using the same scheme based on the requirement. The problem is formulated mathematically to maximize sum rate and an approach for resource allocation is proposed. This work aims to achieve higher system capacity with the proposed resource allocation scheme that does not have exclusivity in resource allocation and allows more than one user to share the same resource block without degrading the system spectral efficiency. However, to reduce the complexity, an upper limit on the number of users per resource block is imposed which is adaptive to the traffic environment. Results show that the proposed scheme can significantly enhance the system performance in terms of fairness and spectral efficiency.

Abstract

Fifth generation (5G) cellular networks would include millimeter wave (mmWave) for communication which gains the advantage of extensive frequency reuse and directive propagation. The recent trends in device-to-device (D2D) communication have escalated the importance of proximity based services for next generation cellular systems. The integration of mmWave and D2D technology will improve the performance and spectral efficiency greatly. In areas with high call traffic density, D2D communication can operate in overlay mode giving the advantage of low complexity interference avoidance technique. The complexity is decreased as the resources to overlay mode are dedicated resulting in no interference from cellular user unlike in underlay mode. Such a scenario is considered in this paper where interference management between D2D devices sharing same radio resources plays a key role in system performance. Thus, the inband resources for D2D operation selectively switch to overlay mode in based on the call traffic. Two problems are addressed in this paper. Firstly, assigning the resources for overlay mode which is solved using quadratic programming. Next, the problem is formulated to maximize system capacity in overlay mode which is solved using a heuristic algorithm which is simulated. The proposed scheme has optimal performance for dense network in controlled overlay operation for 5G networks. The scheme also overcomes the problem of spectrum wastage in overlay mode by its selective operation.

Abstract

This paper presents a technique for coronary artery disease (CAD) detection through photoplethysmography (PPG). This work is aimed at developing a non-invasive, inexpensive screening technique suitable for home monitoring. Time domain analysis of PPG signal and its second derivative has been carried out to extract distinguishing features. Support Vector Machine based classifier has been used to classify CAD patients. ICU patient data from MIMIC-II dataset has been used for performance evaluation. Sensitivity of 85% and specificity of 78% has been achieved for the analysed data.

Abstract

As the number of IoT devices are rapidly increasing and a huge number of access layer devices are underutilized, we propose a self organized software defined edge controller in internet of things (IoT) infrastructure. Our proposed controller gathers information about all the available resources and learns the parameters for serving an IoT node in such a way that the total service time for nodes in the system minimizes. Two methods, round robin and mean time to complete, are utilized for redirecting requests to the resources. MATLAB and mininet simulation results show the self organizing capability of the proposed system.

Abstract

Earthquake induced landslides during rapid flow cause several damages to human life in different ways. In September 2011, Sikkim earthquake with 6.9 Mw is considered to be the biggest earthquake in few decades in Sikkim state that effected with several low to rapid landslides blocked several roadways, destroyed many structures and human life etc. The slope is selected from Northern part of Sikkim state which is 42 Km from Mangan Village at 270 23 1 24.56 11 N - 88 0 38 1 30.35 11 E latitude and longitudes with 5844 feet and 7102 feet elevation and altitude. Due to quite history of several earthquakes from past ten years slope is selected near Mangan village. The Main central thrust is passing from Mangan town having with 0.5H high geohazard, amplification factor of 8.5 and predominant frequency of 10Hz and 0.82g average peak ground acceleration. The slope is divided in too fine mesh using Finite element modeling. Ten nodal points on the slope profile are selected for the analysis. The analysis was done for considering three conditions i.e., saturated, partially saturated and fully saturated depending upon the climatic condition. The acceleration time history data collected from Gangtok station are used for the present analysis. The maximum displacements and velocity flow both in X and Y direction of the slope is calculated under dynamic loading using finite element approach.The results obtained are compared with actual slope profile.

Abstract

Missing data in satellite images is a known phenomenon in remote sensing. There are various reasons for these gaps in satellite images e.g. shadowed area for SAR data sets, cloud coverage for optical imagery and instrument errors such as Scan Line Corrector(SLC-off) failure. The Scan Line Corrector (SLC) of Landsat 7 Enhanced Thematic Mapper Plus (ETM+) sensor failed permanently causing around 23% of pixels not scanned in each scene and this failure has seriously limited the scientific applications of ETM+ data. Reconstructing the gap regions is an important issue in remote sensing. To address this issue, we have proposed an idea to find the homogeneous and heterogeneous regions of ETM+ image using Morphological operations and replace the unscanned pixel value by Neighborhood similar pixel threshold based Local Binary Pattern value. We have applied padding approach with additional rows and columns for unscanned mixels at borders to replace with the corresponding NLBP value. It is observed that our results are promising to fill the data gaps in both homogeneous and heterogeneous regions.

Abstract

Forest fires are the most frequently occurred phenomenon during summer seasons in the state Madhya Pradesh. Monitoring and assessment of forest fires are the crucial steps in effective forest fire management. Forest fire danger estimation helps the disaster management authorities to take necessary mitigation measures for minimizing the losses and to evacuate the local people. Fire danger rating systems predict the fire danger based on the meteorological station parameters and ground datasets. McArthur Forest Fire Danger Index (FFDI) is the most popularly used fire danger rating systems using in the country Australia. This index requires large amount of ground datasets for the computation of drought parameter. In India, it is very difficult to compute the drought parameter due to the unavailability of instruments and man power. In the present research, McArthur Fire Danger Index was modified by inducing Normalized multiband drought index (NMDI) that was generated from Moderate Resolution Imaging Spectroradiometer (MODIS) TERRA surface reflectance product MOD09GA as a substitute for fuel availability parameter. To test the robustness of modified FFDI, the research was carried out on Madhya Pradesh state for the assessment of forest fire danger. The results obtained from

Abstract

Due to the large quantity and diversity of content being eas- ily available to users, recommender systems (RS) have be- come an integral part of nearly every online system. They allow users to resolve the information overload problem by proactively generating high-quality personalized recommen- dations. Trust metrics help leverage preferences of similar users and have led to improved predictive accuracy which is why they have become an important consideration in the design of RSs. We argue that there are additional aspects of trust as a human notion, that can be integrated with col- laborative filtering techniques to suggest to users items that they might like. In this paper, we present an approach for the top-N recommendation task that computes prediction scores for items as a user specific combination of global and local trust models to capture differences in preferences. Our experiments show that the proposed method improves upon the standard trust model and outperforms competing top- N recommendation approaches on real world data by upto 19%.