A Path-Centric Channel Assignment Framework for Cognitive Radio Wireless Networks

Today’s static spectrum allocation policy results in a situation where the available spectrum is being exhausted while many licensed spectrum bands are under-utilized. To resolve the spectrum exhaustion problem, the cognitive radio wireless network, termed CogNet in this paper, has recently been proposed to enable unlicensed users to dynamically access the licensed spectrum bands that are unused in either temporal or spatial domain, through spectrum-agile cognitive radios. The CogNet plays the role of secondary user in this shared spectrum access framework, and the spectrum bands accessible by CogNets are inherently heterogeneous and dynamic. To establish the communication infrastructure for a CogNet, the cognitive radio of each CogNet node detects the accessible spectrum bands and chooses one as its operating frequency, a process termed channel assignment. In this paper we propose a graph-based path-centric channel assignment framework to model multi-hop ad hoc CogNets and perform channel assignment from a network perspective. Simulation results show that the path-centric channel assignment framework outperforms traditional link-centric approach.

High efficiency GaN-based light-emitting diodes with embedded air voids/SiO2 nanomasks

In this paper, the high performance GaN-based light-emitting diodes (LEDs) with embedded microscale air voids and an SiO(2) nanomask by metal-organic chemical vapor deposition (MOCVD) were demonstrated. Microscale air voids and an SiO(2) nanomask were clearly observed at the interface between GaN nanorods (NRs) and the overgrown GaN layer by scanning electron microscopy (SEM). From the reflectance spectra we show strong reflectance differences due to the different refractive index gradient between the GaN grown on the nanotemplate and sapphire. It can increase the light extraction efficiency due to additional light scattering. The transmission electron microscopy (TEM) images show the threading dislocations were suppressed by nanoscale epitaxial lateral overgrowth (NELOG). The LEDs with embedded microscale air voids and an SiO(2) nanomask exhibit smaller reverse-bias current and large enhancement of the light output (65% at 20 mA) compared with conventional LEDs.     

1.3-$mu$m Amplifier-Free All-Optical 2R Regenerator Using Two-Mode Injection-Locked Distributed Feedback Laser Diode

This study presents a novel and cost-effective 1.3-mum all-optical 2R regenerator based on a two-mode injection-locked distributed feedback laser diode. The proposed 2R regenerator, with 14.13-dB small signal gain, has achieved an amplifier-free 10-Gb/s straight line transmission over 60 km while keeping the power penalty less than 0.84 dB at bit-error rate (BER) =10^-9. In addition, properties like BER degradation, output extinction ratio, gain, and data-rate transparency are also experimentally investigated     

Determination of optical parameters of a twisted-nematic liquid crystal by phase-sensitive optical heterodyne interferometric ellipsometry

What is believed to be a novel phase-sensitive optical heterodyne interferometric ellipsometer is set up to characterize a twisted-nematic liquid crystal (TN-LC) by the elliptical parameters of the output polarization state. This ellipsometer presents the advantages of both polarized optical heterodyne interferometry and optical photometry, which introduce a polarization modulation that is capable of performing with high-sensitivity on phase detection in real time. The twist angle phi and the untwisted phase retardation gamma of TN-LC are measured precisely. The experimental results verify that a TN-LC can be treated as identical to an elliptical retarder.     

Optical phase-locked loop (OPLL) for an amplitude modulatedcommunications link using solid-state lasers

Theoretical analysis is formulated for a solid state laser based optical phase-locked loop (OPLL) disturbed by shot noise, amplitude modulated noise, and frequency noise. The frequency noise spectral density of solid state lasers is modeled to contain a white component, a 1/f component, and a strong 1/f² component at the laser output. This model is verified and the spectral content of each component is measured using an open-loop RF frequency discriminator. The choice of loop filter is made by considering the frequency noise components, transient effects, and the loop damping factor ζ. The total phase error variance as a function of loop bandwidth is displayed for several values of carrier signal-to-noise ratio for the measured frequency noise spectrum. Optimal loop bandwidth is also calculated as a function of carrier signal-to-noise ratio. An OPLL experiment is performed, and measured phase error variance is compared with the theoretical predictions using the measured frequency noise spectrum. The results show that the measured phase error variance closely matches the theoretical predictions     

120-nm bandwidth erbium-doped fiber amplifier in parallel configuration

A new S- to L-band erbium-doped fiber amplifier (EDFA) module, which reaches 120-nm gain bandwidth of 1480 to 1600 nm, has been experimentally investigated and demonstrated by using coupled structure. A 32.8-, 34.7-, and 38.1-dB peak gain is obtained at 1504, 1532, and 1568 nm, respectively, when the input signal power is -30 dBm. In addition, this proposed amplifier also provides a broad-band amplified spontaneous emission (ASE) light source of 1480-1606 nm with the output level above -40 dBm.     

Generalized eigenspace-based beamformers

The generalized eigenspace-based beamformer (GEIB) is presented here, which utilizes the eigenstructure of the correlation matrix to enhance the performance of the linearly constrained minimum variance beamformer (LCMVB). The weight vector of the GEIB is found by projecting the LCMVB weight vector onto a vector subspace constructed from the eigenstructure of the correlation matrix. The GEIB and the LCMVB have the same responses to the desired signal and the interferers. However, the weight vector of the GEIB has a smaller norm and generates a lower output noise power. An additional advantage of the GEIB is that the linear constraints can be treated flexibly, i.e. each linear constraint can be chosen to be preserved or not preserved. The cost of preserving a linear constraint is to get more output noise power. In addition to developing the GEIB, we discuss the effects of imposing linear constraints on the output noise powers of the GEIB and the LCMVB. Computer simulations are also presented that demonstrate the merits of the GEIB     

The effect of non-reactive alloying elements on the growth kinetics of the intermetallic compound between liquid Sn-based eutectic solders and Ni substrates

The effect of non-reactive alloying elements on the growth kinetics of the intermetallic compound formed between liquid Sn-based eutectic solders and Ni substrates at 250–280 °C was studied, with Ag, Pb and Bi as the alloying elements. Experimental results showed that only the intermetallic compound (IMC) Ni₃Sn₄ was present as the reaction product. Based on the diffusion controlled reaction mechanism, a parabolic kinetic equation in which both Sn concentration and alloying element effects were quantified, was derived and verified by comparing the kinetic data obtained by using four different solders with different Sn concentrations and alloying elements. The apparent activation energies for pure Sn and eutectic Sn–Ag, Sn–Pb and Bi–Sn solders reacting with Ni, the substrate, are in the range of 28.7–32.7 kJ mol⁻¹, which indicates slight variations in activation energy. The similarity between the activation energies of these four solders suggests that the diffusion of Sn atoms through the IMC is the rate controlling step. The effect of non-reactive alloying elements exerted on the pre-exponential factor of the equilibrium constant which multiplied by the bulk concentration of Sn in solder gives the surface concentration of Sn in Ni₃Sn₄.     

Human action recognition based on graph-embedded spatio-temporal subspace

Human action recognition is an important issue in the pattern recognition field, with applications ranging from remote surveillance to the indexing of commercial video content. However, human actions are characterized by non-linear dynamics and are therefore not easily learned and recognized. Accordingly, this study proposes a silhouette-based human action recognition system in which a three-step procedure is used to construct an efficient discriminant spatio-temporal subspace for k-NN classification purposes. In the first step, an Adaptive Locality Preserving Projection (ALPP) method is proposed to obtain a low-dimensional spatial subspace in which the linearity in the local data structure is preserved. To resolve the problem of overlaps in the spatial subspace resulting from the ambiguity of the human body shape among different action classes, temporal data are extracted using a Non-base Central-Difference Action Vector (NCDAV) method. Finally, the Large Margin Nearest Neighbor (LMNN) metric learning method is applied to construct an efficient spatio-temporal subspace for classification purposes. The experimental results show that the proposed system accurately recognizes a variety of human actions in real time and outperforms most existing methods. In addition, a robustness test with noisy data indicates that our system is remarkably robust toward noise in the input images.     

RedTweet: recommendation engine for reddit

Twitter and Reddit are two of the most popular social media sites used today. In this paper, we study the use of machine learning and WordNet-based classifiers to generate an interest profile from a user’s tweets and use this to recommend loosely related Reddit threads which the reader is most likely to be interested in. We introduce a genre classification algorithm using a similarity measure derived from WordNet lexical database for English to label genres for nouns in tweets. The proposed algorithm generates a user’s interest profile from their tweets based on a referencing taxonomy of genres derived from the genre-tagged Brown Corpus augmented with a technology genre. The top K genres of a user’s interest profile can be used for recommending subreddit articles in those genres. Experiments using real life test cases collected from Twitter have been done to compare the performance on genre classification by using the WordNet classifier and machine learning classifiers such as SVM, Random Forests, and an ensemble of Bayesian classifiers. Empirically, we have obtained similar results from the two different approaches with a sufficient number of tweets. It seems that machine learning algorithms as well as the WordNet ontology are viable tools for developing recommendation engine based on genre classification. One advantage of the WordNet approach is simplicity and no learning is required. However, the WordNet classifier tends to have poor precision on users with very few tweets.