{"title":"Unsaturated MLP neural networks training algorithm using a piecewise error function and adaptive learning rates","authors":"P. Moallem, S. A. Ayoughi","doi":"10.1109/ISTEL.2008.4651271","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651271","url":null,"abstract":"Saturation conditions of the hidden layer neurons are a major cause of learning retardation in multilayer perceptrons (MLP). Under such conditions the traditional backpropagation (BP) algorithm is trapped in local minima. To renew the search for a global minimum, we need to detect the traps and an offset scheme to avoid them. We have discovered that the gradient norm drops to a very low value in local minima. Here, adding a modifying term to the standard error function enables the algorithm to escape the local minima. In this paper, we proposed a piecewise error function; i.e. where the gradient norm remained higher than a parameter we used the standard error function, and added a modifying term to the function below this value. To further enhance this algorithm, we used our proposed adaptive learning rate schema. We performed a selection of benchmark problems to asses the efficiency of our proposed algorithm. Compared to previously proposed algorithms, we recorded higher convergence rates, especially in complex problems with complex input-output mapping.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114259213","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Transmit covariance for spatially-correlated multiple-antenna Ricean fading channels with channel distribution side information at transmitter","authors":"M. Adibi, V. Vakili","doi":"10.1109/ISTEL.2008.4651314","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651314","url":null,"abstract":"In this paper, we cast the problem of finding the capacity-achieving transmit covariance of correlated Ricean MIMO fading channels as an equivalent problem as for uncorrelated Ricean MIMO channels. Perfect channel information is assumed to be available at the receiver, while the transmitter only has channel distribution side information. An iterative algorithm based on Newton method to compute the capacity-achieving transmit covariance for the general case of double-ended correlated MIMO channels is proposed in this paper as well. Additionally, we provide a reduced complexity technique to attain a suboptimal transmit covariance. One of the main contributions in this paper is the characterization of the eigenstructure of the suboptimal transmit covariance. Moreover, Sufficient condition is derived for the optimality of this suboptimal solution.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114710294","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Quantization based audio watermarking in a new transform domain","authors":"M. Akhaee, A. Nikooienejad, F. Marvasti","doi":"10.1109/ISTEL.2008.4651387","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651387","url":null,"abstract":"In this paper, a novel blind watermarking technique based on quantization is proposed. Quantization is performed in a special domain which converts one dimensional signal to a 2-D one named Point to Point Graph (PPG). Basis of the method is on the separation of this domain into two portions; while, only one portion is quantized. Furthermore, in the dewatermarking procedure, by using the unquantized portion and zero norm, the embedded data can be extracted. The performance of the proposed method is analytically investigated and verified by simulation with artificial Gaussian signals. Experimental results over several audio signals shows the great robustness of the technique in comparison with the algorithms presented so far.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126167302","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Hyperspectral classification using Markov random field and “spatial probability density function”","authors":"A. Keshavarz, H. Ghassemian","doi":"10.1109/ISTEL.2008.4651386","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651386","url":null,"abstract":"Hyperspectral images have the unique ability to provide both a spatial sampling and a spectral sampling. Although the hyperspectral data contain a lot of information about the spectral properties of the land cover, but no spatial information is inherent in the spectral data. This problem can be solved using a joint spectral/spatial classifier. In this paper we propose a classification algorithm based on Markov random field and using spatial and spectral information simultaneously. In proposed algorithm spectral features are extracted at first step. At second step an iterative spatial-spectral classification algorithm is applied. ldquospatial probability density functionrdquo of classes is estimated using kernel density estimation in second step. The hyperspectral data set used in our experiments is a scene taken over NW Indianapsilas Indian Pine by the AVIRIS sensor. The obtained results show that proposed classifier improved classification accuracy significantly.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"94 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126862389","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A modified two-ring reference model for MIMO Mobile-to-Mobile communication channels","authors":"G. Bakhshi, Reza Saadat, Kamal Shahtalebi","doi":"10.1109/ISTEL.2008.4651337","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651337","url":null,"abstract":"Analysis and design of multi-element antenna systems in mobile fading channels require a model for the space-time cross-correlation among the links of the underlying multiple-input multiple-output (MIMO) mobile-to-mobile (M-to-M) communication channel. In this paper, we propose a MIMO channel reference model for M-to-M communication systems that we call it as the modified geometrical two-ring model. This model is based on the extension of single-bounce two-ring scattering model for flat fading channel under the assumption that the transmitter and the receiver are moving. Assuming an isotropic and single-bounce scattering model, a closed-form expression for the space-time cross-correlation function (CCF) between any two sub channels is derived. The proposed model provides an important framework in M-to-M system design, where includes many existing correlation models as special cases. Some numerical results are presented as special cases of the derived CCF.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124933225","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Multi-service path switching in all-optical GMPLS core network","authors":"H. Beyranvand, J. Salehi","doi":"10.1109/ISTEL.2008.4651352","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651352","url":null,"abstract":"In this paper, we present a novel labeling scheme for multi-service path switching in the all-optical generalized multiprotocol label switched (GMPLS) core networks. In this scheme, we employ the multi-length variable-weight optical orthogonal code (MLVW-OOC) as a multiplexing signature sequence and identifying label. In order to analyze the performance of the network, the probability of error is evaluated for arbitrary number of classes and cross-correlation values. Furthermore, the probability of performance degradation is investigated by considering the activity coefficient of connected paths. To assure the demanded performance, label distributed protocol (LDP) is accommodated by a controlling mechanism which limits the number of connected paths. Finally, a four-class network is designed by the proposed labeling scheme and the probabilities of error and the performance degradation are investigated; results indicate that decreasing the upper bound of the connected paths will reduce the probability of the performance degradation.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123618816","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Characterization of ultra wideband temporal dispersion parameters in indoor laboratory environments","authors":"N. Noori, R. Baee, A. Abolghasemi","doi":"10.1109/ISTEL.2008.4651276","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651276","url":null,"abstract":"In this paper, the results of some ultra wideband (UWB) time-domain propagation measurements performed in modern laboratory (Lab) environments are presented. These measurements have been down for both line of site (LOS) and non-LOS (NLOS) scenarios. The measurement results are used to investigate temporal dispersion parameters of the channel. The appropriate model parameters are presented for UWB signal transmission in the Lab environment based on the obtained results.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"134 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123207017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Enhanced object tracking with received signal strength using Kalman filter in sensor networks","authors":"M. Nabaee, Ali Pooyafard, A. Olfat","doi":"10.1109/ISTEL.2008.4651321","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651321","url":null,"abstract":"The importance of localization task has drawn much attention to location estimation and object tracking systems in wireless sensor networks. Many methods have been proposed to improve the location accuracy in which received signal strength (RSS) values of sensor nodes are used as an indicator of the distance between sensor node and the source node. Some of the previously proposed tracking algorithms are based on Kalman Filtering (KF) which makes us capable of tracking the location of a mobile node (MN). In this paper, the implementation and enhancement of a tracking system based on RSS indicator with the aid of an Extended Kalman Filter (EKF) is described and an adaptive filter is derived. The dynamic characteristic of channel requires considering the variations of path loss exponent of the space. Fast variations in the movement path of the source node can explicitly interrupt the performance of the localization because of inappropriate initial conditions. This imperfect behavior of the initially modelled EKF is improved and the simulation results are provided to assess the achieved enhancement. It is shown that MSE of the proposed algorithm is considerably lower than other modelled EKFs and that in presence of high measurement noise or with fewer sensor nodes this method clearly outperforms the conventional approach.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131803058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"A new eigenspace subband beamformer for cancellation of broadband interferences using Subband Adaptive Filtering","authors":"M.R. Edalatzadeh, M. Samiei","doi":"10.1109/ISTEL.2008.4651315","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651315","url":null,"abstract":"At this paper a new broadband beamformer is presented that is realized in eigenspace. That is designed using generalized sidelobe canceller (GSC) structure, linearly constrained minimum variance (LCMV) algorithm, subband adaptive filtering (SAF) method, and uses effective eigenvalues of received data to cancel narrowband and broadband interferers that arrive to array systems through their antenna sidelobes. In nonstationary environments and in environments that number of input samples is low, the new eigenspace subband beamformer is better than sample matrix inversion (SMI) counterpart and gives higher array gain. Also, it requires fewer samples to reach to a given SNR at output. Thus, computational complexity of system is reduced. This design can be used for multipath fading effect mitigation in wireless communications, jamming cancellation in radar systems, interference signals elimination in noisy places such as stadiums, many reflection locations like conference rooms, and etc.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"97 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128227694","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Characterization of the impact of power normalization on the MIMO mutual information","authors":"O. Darvishi, B. Abolhassani, H. S. Rad","doi":"10.1109/ISTEL.2008.4651336","DOIUrl":"https://doi.org/10.1109/ISTEL.2008.4651336","url":null,"abstract":"In this paper, we investigate the impact of power normalization on the mutual information of Multiple-Input Multiple-Output (MIMO) systems. We quantify the difference in the mutual information when using two different power normalization methods - corresponding to slow and fast power control - and characterize this difference in the high SNR regime in terms of its distribution, mean and variance. We assume Rayleigh fading channels with perfect channel state information at the receiver. Also, we assume equal power allocation strategy is applied at the transmitter. It turns out that the difference in mutual information is governed by the temporal variations of the SNR. In MIMO systems, the inherent spatial diversity of the channel makes the difference negligible in Rayleigh fading environments.","PeriodicalId":133602,"journal":{"name":"2008 International Symposium on Telecommunications","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2008-10-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130681514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}