Daily Archives: November 17, 2015


This paper presents a speaker identification algorithm using the deep neural network (DNN) as the classifier to learn the features of the voiceprints represented by spectrogram. The collected speech signals are pre-emphasized, windowed, divided into some chunks, then calculated to obtain the magnitude of the frequency spectrum, which creates the spectrograms. The […]

A spectrogram-based voiceprint recognition using deep neural network


The early detection of wildfires with monitoring sensors can improve the effectiveness of fire responders, decreasing the destruction of property and saving lives. However, distributing these sensors across a wide area is an expensive and laborsome task. Furthermore, the cost of the sensors themselves may be high, limiting their practicality […]

Samara: Biologically Inspired Self-Deploying Sensor Networks


Many Wireless Sensor Network (WSN) applications operate autonomously in unreliable or inaccessible environments, precluding maintenance or human intervention. Redundant deployment schemes are usually considered in this scenario, making the network resilient to failure and environmental changes. Furthermore, sleep-scheduling techniques can also be applied, enabling redundant nodes to turn off their radios, while active nodes […]

A sleep-scheduling scheme for enhancing QoS and network coverage in IEEE 802.15.4 WSN



Secure delivery networks can help prevent or mitigate the most common attacks against mission-critical websites. A case study from a leading provider of content delivery services illustrates one suchnetwork’s operation and effectiveness. The Web extra at https://youtu.be/4FRRI0aJLQM is an overview of the evolving threat landscape with Akamai Director of Web Security Solutions […]

Protecting Websites from Attack with Secure Delivery Networks


We study the problem of survivable Virtual Optical Network (VON) mapping in a flexible grid opticalnetwork with programmable regenerators that are capable of converting spectrum and modulation format. We propose heuristic algorithms to address the problem with the objective of minimizing blocking probability for dynamic VON requests.

Survivable Virtual Optical Network mapping in spectrum and modulation format convertible flexible grid optical networks


A SiGe BiCMOS power amplifier (PA) with a switchable output matching network (OMN) is designed and measured. A MOS switch is incorporated in the OMN of the PA. According to the state of the switch, the proposed PA can be operated in either the high-power (HP) or low-power (LP) modes. In the […]

SiGe BiCMOS power amplifier with a switchable output matching network for efficiency enhancement



Translating behavioral information learned from analysis of a botnet’s software in controlled environments, to a model of how the botnet behaves in the wild is complicated by the fact that controlled environments do not account for a wide variety of user behavior, and that machines are not associated one-to-one with […]

Visualizing and Modeling the Scanning Behavior of the Conficker Botnet in the Presence of User and Network Activity


We address the problem of identifying the topology of an unknown weighted, directed network of LTI systems stimulated by wide-sense stationary noises of unknown power spectral densities. We propose several reconstruction algorithms by measuring the cross-power spectral densities of the networkresponse to the input noises. The measurements are based on a series of […]

Topology Identification of Directed Dynamical Networks via Power Spectral Analysis


This paper introduces an event-driven feedforward categorization system, which takes data from a temporal contrast address event representation (AER) sensor. The proposed system extracts bio-inspired cortex-like features and discriminates different patterns using an AER based tempotron classifier (a network of leaky integrate-and-fire spiking neurons). One of the system’s most appealing characteristics is […]

Feedforward Categorization on AER Motion Events Using Cortex-Like Features in a Spiking Neural Network



This paper presents an energy management approach for a hybrid energy system comprised of a photovoltaic (PV) array and a polymer electrolyte membrane fuel cell (PEMFC). A single storage device, i.e., a Li-ion battery module, is used in the proposed structure. Linear proportional-integral (PI) and nonlinear flatness-based controllers for dc […]

DC Bus Stabilization of Li-Ion Battery Based Energy Storage for a Hydrogen/Solar Power Plant for Autonomous Network Applications