We consider the problem of dynamically apportioning resources among a set of options in a worst-case online framework. The model we investigate is a generalization of the well studied online learning model. In particular, we allow the learner to see as additional information how high the risk of each option is. This assumption is natural in many applications like horse-race betting, where gamblers know odds for all options before placing bets. We apply Vovk's Aggregating Algorithm to this problem and give a tight performance bound. The results support our intuition that it is safe to bet more on low-risk options. Surprisingly, the loss bound of the algorithm does not depend on the values of relatively small risks.

The electroencephalogram (EEG) has become a widely used tool for investigating brain function. Brain signal source localization is a process of inverse calculation from sensor information (electric potentials for EEG) to the identification of multiple brain sources to obtain the locations and orientation parameters. In this paper, we describe a combination of the backpropagation neural network (BPNN) with the nonlinear least-square (NLS) method to localize two dipoles with reasonable accuracy and speed from EEG data computerized by two dipoles randomly positioned in the brain. The trained BPNN, obtains the initial values for the two dipoles through fast calculation and also avoids the influence of noise. Then the NLS method (Powell algorithm) is used to accurately estimate the two dipole parameters. In this study, we also obtain the minimum distance between the assumed dipole pair, 0.8 cm, in order to localize two sources from a smaller limited distance between the dipole pair. The present simulation results demonstrate that the combined method can allow us to localize two dipoles with high speed and accuracy, that is, in 20 seconds and with the position error of around 6.5%, and to reduce the influence of noise.

To obtain large capacity, high quality mobile satellite communication systems in the future, we must use a multi-beam that can cope with extremely high levels of frequency reuse. This paper describes a novel resource allocation algorithm for multi-beam satellite communication systems that can dynamically adapt to maximum communication capacity without compromising quality. The algorithm combines two resource allocation schemes that enable it to contend with the ever-changing user distribution and inter-beam interference conditions. The first scheme optimizes the resources amongst beams. To minimize interference, the optimal constraint conditions are clarified when all clusters share and occupy the same bandwidth completely. These constraints are used in the optimization algorithm. The second scheme manages the various required resources and adapts them to the beam gain and interference levels at various user locations within a single beam. We propose a fixed power adaptive modulation scheme to obtain stable communications. This two-layered scheme can satisfactorily allocate multi-beam satellite resources to contend with the increasing communication capacity and still improve the quality.

We present a method for classifying image pixels of real images into multiple photometric factors: specular reflection, diffuse reflection, attached shadows and cast shadows. Conventional photometric linearization methods cannot correctly classify pixels under near point light sources, since they assume parallel light. To satisfy this assumption, our method utilizes a photometric linearization method that divides images into small regions. It also propagates linearization coefficients from neighboring regions. Our experimental results show that the proposed method can correctly classify image pixels into photometric factors, even if images are obtained under near point light sources.

This paper deals with a secret key agreement problem from correlated random numbers. It is proved that there is a pair of linear matrices that yields a secret key agreement in the situation wherein a sender, a legitimate receiver, and an eavesdropper have access to correlated random numbers. A relation between the coding problem of correlated sources and a secret key agreement problem from correlated random numbers are also discussed.

This paper investigates some relations among four complexities of sequence over countably infinite alphabet, and shows that two kinds of empirical entropies and the self-entropy rate regarding a Markov source are asymptotically equal and lower bounded by the maximum number of phrases in distinct parsing of the sequence. Some connections with source coding theorems are also investigated.

Recent IDEs have become more extensible tool platforms but do not concern themselves with how other tools running on them collaborate with each other. They compel developers to use proprietary representations or the classical abstract syntax tree (AST) to build source code tools. Although these representations contain sufficient information, they are neither portable nor extensible. This paper proposes a tool platform that manages commonly used, fined-grained, information about Java source code by using an XML representation. Our representation is suitable for developing tools which browse and manipulate actual source code, since the original code is annotated with tags based on its structure and retained within the tags. Additionally, it exposes information resulting from global semantic analysis, which is never provided by the typical AST. Our proposed platform allows the developers to extend the representation for the purpose of sharing or exchanging various kinds of information about the source code, and also enables them to build new tools by using existing XML utilities.

The explosive growth of wireless network users and the existence of various wireless services have demanded high throughput as well as user's quality-of-service (QoS) guarantees. In accordance with, this paper proposes a novel resource allocation scheme improving both the capability of QoS-provisioning for multiple users and the overall data throughput. Towards this, the modified resource allocation technique combined with the modified largest weighted delay first (M-LWDF) scheme will be exploited upon considering statistical channel behavior as well as real time queuing analysis connected to resource allocation. In order to verify the validity of the proposed resource allocation scheme, the time division multiple access (TDMA) system will be considered as a target application. The simulation results confirm that the proposed scheme gives rise to superior performance in a way of showing results of several performance measures under time-varying wireless fading channel.

A power and spreading gain allocation strategy is considered for effectively providing data services for mobile users with different levels of priorities in a DS-CDMA system supporting real-time and non-real-time services. Specifically, the uplink in the DS-CDMA system is considered subject to a constraint on total power received at the base station caused by non-real-time data services. The constraint is imposed to meet QoS requirements of real-time services. The priority level of a data user is specified by the weighting factor assigned to the data throughput of the user. Our strategy implements a relative prioritization that affords a trade-off between spectral efficiency and strict prioritization.

The complexity of the optimal decoding for vector quantization (VQ) in code-division multiple access (CDMA) communications prohibits implementation. It was recently shown in [1] that a suboptimal scheme that combines a soft-output multiuser detector and individual VQ decoders provides a flexible tradeoff between decoder's complexity and performance. The work in [1], however, only considers an AWGN channel model. This paper extends the technique in [1] to a frequency-selective Rayleigh fading channel. Simulation results indicate that such a suboptimal decoder also performs very well over this type of channel.

Wireless multi-hop networks have drawn much attention for the future generation mobile communication systems. These networks can establish multiple routes from a source node to a destination node because of flexible construction of network topology. Transmissions by multiple routes have enough capability to achieve reliable communication because we can expect to obtain diversity gain by multiple routes. In this paper, we propose the multi-route coding scheme. At first, we discuss a channel model in multi-hop networks employing regenerative relay, which we named the virtual channel model. By using the virtual channel model, a packet is encoded on multiple routes as follows; a bit sequence of a packet is encoded and divided into subpackets, and each subpacket is transmitted on each route. We evaluate its packet error rate performance, and clarify effectiveness of the proposed scheme. In general, we should face degradation of a route condition such as the case when a subpacket does not reach a destination node. Hence, we have to consider the influence of subpacket loss. We also investigate it, and show tolerance of the proposed scheme over that.

This paper compares selected Optical Packet Switching architectures that use the wavelength conversion technique to solve the packet contention problem. The architectures in question share wavelength converters, which are needed to wavelength translate arriving packets. This paper focuses on two architectures: the Shared Per Output Line (SPOL) and the Shared Per Input Line (SPIL) architectures, in which the wavelength converters are shared per output and input fiber respectively. The performance of the proposed architectures is evaluated for all the balance/unbalance combinations of input/output traffic. Packet loss probability is expressed as a function of the number of wavelength converters used, by means of analytical models validated by simulations. The results obtained show that the SPIL architecture, when compared to the SPOL architecture, allows for greater economies in terms of number of wavelength converters needed. While the performance of the two architectures tends to have similar values in a scenario with unbalanced input traffic and balanced output traffic, in unbalanced output traffic scenarios the SPIL architecture requires about 50% less wavelength converters than the SPOL architecture does, for a given packet loss probability.

IP multicast technology is highly advantageous for various applications and future needs in the Internet. Yet, it is generally recognized that the IP multicast routing protocol is fairly complex and non-scalable and requires additional maintenance and operational cost to network administrators. Although there has been much research related to IP multicast and most router vendors already support basic IP multicast routing protocols, there is still a big gap between what is reported as the state-of-the-art in the literature from what is implemented in practice. In this paper, we clarify the complexities of traditional multicast communication and describe possible solutions using the one-to-many multicast communication model called Source-Specific Multicast (SSM). We explain this communication model and the corresponding routing architecture and examine the statistics obtained for the number of multicast routing entries in our backbone router, which is connected to the international backbone. We also introduce our international collaboration activities that are contributing to the deployment and promotion of IP multicast services in the Internet.

A time domain (TD) speech enhancement technique to improve SNR in noise-contaminated speech is proposed. Additional supplementary scheme is applied to estimate the degree of noise of noisy speech. This is estimated from a function, which is previously prepared as the function of the parameter of the degree of noise. The function is obtained by least square (LS) method using the given degree of noise and the estimated parameter of the degree of noise. This parameter is obtained from the autocorrelation function (ACF) on frame-by-frame basis. This estimator almost accurately estimates the degree of noise and it is useful to reduce noise. The proposed method is based on two-stage processing. In the first stage, subtraction in time domain (STD), which is equivalent to ordinary spectral subtraction (SS), is carried out. In the result, the noise is reduced to a certain level. Further reduction of noise and by-product noise residual is carried out in the second stage, where blind source separation (BSS) technique is applied in time domain. Because the method is a single-channel speech enhancement, the other signal is generated by taking the noise characteristics into consideration in order to apply BSS. The generated signal plays a very important role in BSS. This paper presents an adaptive algorithm for separating sources in convolutive mixtures modeled by finite impulse response (FIR) filters. The coefficients of the FIR filter are estimated from the decorrelation of two mixtures. Here we are recovering only one signal of interest, in particular the voice of primary speaker free from interfering noises. In the experiment, the different levels of noise are added to the clean speech signal and the improvement of SNR at each stage is investigated. The noise types considered initially in this study consist of the synthesized white and color noise with SNR set from 0 to 30 dB. The proposed method is also tested with other real-world noises. The results show that the satisfactory SNR improvement is attained in the two-stage processing.

In this paper, the magnetic field properties and the dosimetry at intermediate frequency (21 kHz) for an induction heater are investigated with the coil model, which is prescribed as substitute source model in the European standard EN50366 (CENELEC). The accuracy of the magnetic field vectors and the values of the induced current density, which are achieved with the coil model, are compared with the results of a realistic model of the induction heater obtained from the equivalent source model. It is shown that the coil model coincides well for the magnitude of the magnetic field strength around the induction heater. On the other hand, the dominant field vector of the coil model differs significantly from the real induction heater, which leads to induced current densities in the body model which are three time larger. Owing to these results, the applicability of the coil model prescribed in the EN50366 is confirmed for the induction heater.

In this paper, a new implementation of the Viterbi decoder is proposed. The Modified State-Mapping VD algorithm combines the TB algorithm with the RE algorithm. By updating the starting point of the state for each memory bank, and by using Trace Back and Trace Forward information, LIFO (Last Input First Output) operation can be eliminated, which reduces the latency of the TB algorithm and decreases the resource usage of the RE algorithm. When the memory unit is 3, the resource usage is 13184 bits and the latency is 54 clocks. The latency of the proposed algorithm is 25% smaller than the MRE algorithm and 50% smaller than the k-pointer even TB algorithm. In addition, resource usage is 50% smaller than the RE algorithm. The resource usage is a little larger than that of the MRE algorithm for the small value of k, but it becomes smaller after k is larger than 16.

In this study, we propose a robust approach for blind source separation (BSS) by using radial basis function networks (RBFNs) and higher-order statistics (HOS). The RBFN is employed to estimate the inverse of a hypothetical complicated mixing procedure. It transforms the observed signals into high-dimensional space, in which one can simply separate the transformed signals by using a cost function. Recently, Tan et al. proposed a nonlinear BSS method, in which higher-order moments between source signals and observations are matched in the cost function. However, it has a strict restriction that it requires the higher-order statistics of sources to be known. We propose a cost function that consists of higher-order cumulants and the second-order moment of signals to remove the constraint. The proposed approach has the capacity of not only recovering the complicated mixed signals, but also reducing noise from observed signals. Simulation results demonstrate the validity of the proposed approach. Moreover, a result of application to X-ray image separation also shows its practical applicability.

The Hilbert transformation together with empirical mode decomposition (EMD) produces Hilbert spectrum (HS) which is a fine-resolution time-frequency representation of any nonlinear and non-stationary signal. The EMD decomposes the mixture signal into some oscillatory components each one is called intrinsic mode function (IMF). Some modification of the conventional EMD is proposed here. The instantaneous frequency of every real valued IMF component is computed with Hilbert transformation. The HS is constructed by arranging the instantaneous frequency spectra of IMF components. The HS of the mixture signal is decomposed into subspaces corresponding to the component sources. The decomposition is performed by applying independent component analysis (ICA) and Kulback-Leibler divergence based K-means clustering on the selected number of bases derived from HS of the mixture. The time domain source signals are assembled by applying some post processing on the subspaces. We have produced experimental results using the proposed separation technique.

A new system identification based method has been proposed for accurate estimation of vocal tract parameters. An often encountered problem in using the conventional linear prediction analysis is due to the harmonic structure of the excitation source of voiced speech. This harmonic characteristic is coupled with the estimation of autoregressive (AR) coefficients that results in difficulties in estimating the vocal tract filter. This paper models the effective voice source from the residual obtained through the covariance analysis in the first-pass which is then used as input to the second-pass least-square analysis. A better source-filter separation is thus achieved. The formant frequencies and corresponding bandwidths obtained using the proposed method for synthetic vowels are found to be accurate up to a factor of more than three (in percent) compared to the conventional method. Since the source characteristic is taken into account, local variations due to the positioning of analysis window are reduced significantly. The validity of the proposed method is also examined by inspecting the spectra obtained from natural vowel sounds uttered by high-pitched female speaker.

Supporting diversified rates for real-time communications will become possible and essential with the rapidly increasing transmission rates provided by the 4th generation (4G) mobile communication systems. In this paper, a novel wireless Quality of Service (QoS) scheme suitable for broadband CDMA packet cellular systems with adaptive modulation coding is proposed and its characteristics are described. The proposed QoS scheme comprises several control factors laid on the MAC and RRC layers, and can be harmonized with IP-QoS. Two important control factors are proposed: radio-condition-aware admission control and resource allocation reflected multistage scheduling. Computer simulations and testbed experiments indicate that by using the radio-condition-aware admission control, stable and guaranteed service can be provided to real-time users regardless of the interference and the variation in the location of the mobile station. Moreover, resource allocation reflected multistage scheduling maintains guaranteed rates for real-time users and provides high resource utilization efficiency for best-effort users. Consequently, by using the proposed wireless QoS scheme, it is possible to provide users with high quality and diversified real-time services, on a packet based radio network for enhanced 3G and beyond.