Time filter

Source Type

Tamulevicius G.,VU Institute of Mathematics and Informatics | Kaukenas J.,VU Institute of Mathematics and Informatics
2016 IEEE 4th Workshop on Advances in Information, Electronic and Electrical Engineering, AIEEE 2016 - Proceedings | Year: 2016

Autoregressive model order and parameter estimation technique is proposed and applied for modeling of Lithuanian semivowels. According to experimental results adequate modeling of semivowels requires for model order 72 in average. The appropriate order value differs for female and male voices. Besides, there are remarkable differencies between word starting and middle phones - the last ones are influenced by surrounding vowels inevitably. This results in need for higher order models and higher prediction error attained. The adequacy of autoregressive model was confirmed objectively and subjectively. © 2016 IEEE.


Kurmis M.,VU Institute of Mathematics and Informatics | Andziulis A.,Klaipeda University | Dzemydiene D.,Mykolas Romeris University | Jakovlev S.,Klaipeda University | And 2 more authors.
Wireless Personal Communications | Year: 2015

One of the most interesting recent topics in pervasive computing is the smart adaptive service support systems. These systems are impossible without knowing the context with the entity at the moment, what context was in the past and what is possible in the near future. The context alone is not so important in human oriented services provided in connected vehicles. In this environment we need to know not only the exact context but also the higher level information—situation. The situation awareness can be increased by exchanging it to other participating nodes—vehicles. Because the vehicular communication network is a very dynamic environment, it is essential to adequately respond to the user’s needs and to provide all the needed services in the right place at the right moment and in the right way. In this work we present our developed cooperative context data acquisition and dissemination model for situation identification in vehicular communication networks. Our solution is different from others as it uses additional virtual context information source—information from other vehicles is weighted and exchanged using the utility function. The proposed decision support system decides if the message should be transmitted to other vehicles, sent to the cloud, saved locally or dismissed. The simulation results show the promising context exchange rate between vehicles and huge saving on channel utilization. © 2015, Springer Science+Business Media New York.


Stankus A.,Klaipeda University | Lukosius Z.,Klaipeda University | Aponkus D.,Klaipeda University | Andziulis A.,Klaipeda University | And 4 more authors.
Elektronika ir Elektrotechnika | Year: 2012

The purpose of the study is to compare the accuracy of point-to-point measurement method of pulse wave propagation time from the multi-point. Electroimpedance method recorded pulse waves in the knees and ankles. Both the ECG signal and the digitized 16-bit analog-digital converter with a frequency of 1 kHz per channel. The study was conducted with 18 healthy volunteers, 20-22 years in the supine position for 3-4 minutes. With the help of LabVIEW tools created by the nine algorithms each subject were analyzed 250-300 pulse waves, highlighting the figure of the coefficient of variation. Their comparison showed that the analysis of pulse waves using different algorithms give different results. The most stable were made when times were measured between the highest peaks of the first derivatives of the forward pulse wave fronts. Similar results were found cross-correlation and cross-spectral analysis of second derivatives of the forward pulse wave fronts. It can be argued that the results of these algorithms allow the parameters obtained by averaging closer to the true results.


Minkevicius S.,VU Institute of Mathematics and Informatics
AIP Conference Proceedings | Year: 2012

The performance in terms of reliability of computer multiserver networks motivates this paper. The probability limit theorem is derived on the extreme queue length in open multiserver queueing networks in heavy traffic and applied to a reliability model for multiserver computer networks where we relate the time of failure of a multiserver computer network to the system parameters. © 2012 American Institute of Physics.


Pyz G.,VU Institute of Mathematics and Informatics | Simonyte V.,Lithuanian University of Educational Sciences | Slivinskas V.,Lithuanian University of Educational Sciences
Elektronika ir Elektrotechnika | Year: 2012

We present a new Lithuanian speech phoneme synthesis method based on the principle of additive synthesis in this paper. An assumption is made that phoneme models consist of the sum of harmonics which could be generated by properly chosen formant synthesizer parameters. In order to estimate the synthesizer parameters, we use the real sound signals that are expanded into harmonics by the inverse fast Fourier transform method. The harmonic synthesizer parameters (amplitudes, damping factor, and phases) are estimated by Levenberg-Marquardt method. We present an example of the synthesized female vowel /a/ and compare it with the true sound signal. © Kauno technologijos universitetas, 2012.


Astrauskas A.,VU Institute of Mathematics and Informatics
Journal of Statistical Physics | Year: 2013

In this paper, we study the asymptotic localization properties with high probability of the Kth eigenfunction (associated with the Kth largest eigenvalue, K≥1 fixed) of the multidimensional Anderson Hamiltonian in torus V increasing to the whole of lattice. Denote by zK,V∈V the site at which the Kth largest value of potential is attained. It is well-known that if the tails of potential distribution are heavier than the double exponential function and satisfies additional regularity and continuity conditions at infinity, then the Kth eigenfunction is asymptotically delta-function at the site zτ(K),V (localization centre) for some random τ(K)=τV(K)≥1. We study the asymptotic behavior of the index τV(K) by distinguishing between three cases of the tails of potential distribution: (i) for the "heavy tails" (including Gaussian), τV(K) is asymptotically bounded; (ii) for the light tails, but heavier than the double exponential, the index τV(K) unboundedly increases like {pipe}V{pipe}o(1); (iii) finally, for the double exponential tails with high disorder, the index τV(K) behaves like a power of {pipe}V{pipe}. For Weibull's and fractional-double exponential types distributions associated with the case (ii), we obtain the first order expansion formulas for logτV(K). © 2012 Springer Science+Business Media New York.


Astrauskas A.,VU Institute of Mathematics and Informatics
Journal of Statistical Physics | Year: 2012

We study the asymptotic structure of the first K largest eigenvalues λκ, V and the corresponding eigenfunctions ψ(·;λκ, V) of a finite-volume Anderson model (discrete Schrödinger operator) HV=κΔv+ξ(·)on the multidimensional lattice torus V increasing to the whole of lattice ℤν, provided the distribution function F(·) of i. i. d. potential ξ(·) satisfies condition -log(1-F(t))=o(t3) and some additional regularity conditions as t → ∞. For z∈V, denote by λ0(z) the principal eigenvalue of the "single-peak" Hamiltonian κΔV+ξ(z)δz in l2(V), and let λ0 κ,V be the kth largest value of the sample λ0(·) in V. We first show that the eigenvalues λκ,V are asymptotically close to λ0 κ, V. We then prove extremal type limit theorems (i. e., Poisson statistics) for the normalized eigenvalues (λκ,V-BV)aV, where the normalizing constants aV>0 and BV are chosen the same as in the corresponding limit theorems for λ0 κ,v The eigenfunction ψ(·;λκ,V) is shown to be asymptotically completely localized (as V ↑ ℤ) at the sites zk,V∈V defined by λ0(Zκ,V)= λ0 κ, V. Proofs are based on the finite-rank (in particular, rank one) perturbation arguments for discrete Schrödinger operator when potential peaks are sparse. © 2011 Springer Science+Business Media, LLC.

Loading VU Institute of Mathematics and Informatics collaborators
Loading VU Institute of Mathematics and Informatics collaborators