Quasi-Orthogonal Design and Performance Analysis of Amplify-And-Forward Relay Networks with Multiple-Antennas

This paper is on the design and performance analysis of practical distributed space-time codes for wireless relay networks with multiple antennas terminals. The amplify-andforward scheme is used in a way that each relay transmits a scaled version of the linear combination of the received symbols. We propose distributed generalized quasi-orthogonal space-time codes which are distributed among the source antennas and relays, and valid for any number of relays. Assuming M-PSK and M-QAM signals, we derive a formula for the symbol error probability of the investigated scheme over Rayleigh fading channels. For sufficiently large SNR, this paper derives closed-form average SER expression. The simplicity of the asymptotic results provides valuable insights into the performance of cooperative networks and suggests means of optimizing them. Our analytical results have been confirmed by simulation results, using full-rate full-diversity distributed codes.

A fast quasi-Newton method for semi-supervised SVM

Due to its wide applicability, semi-supervised learning is an attractive method for using unlabeled data in classification. In this work, we present a semi-supervised support vector classifier that is designed using quasi-Newton method for nonsmooth convex functions. The proposed algorithm is suitable in dealing with very large number of examples and features. Numerical experiments on various benchmark datasets showed that the proposed algorithm is fast and gives improved generalization performance over the existing methods. Further, a non-linear semi-supervised SVM has been proposed based on a multiple label switching scheme. This non-linear semi-supervised SVM is found to converge faster and it is found to improve generalization performance on several benchmark datasets. (C) 2010 Elsevier Ltd. All rights reserved.

Impact of satellite observed microwave SST on the simulation of tropical cyclones

The impact of realistic representation of sea surface temperature (SST) on the numerical simulation of track and intensity of tropical cyclones formed over the north Indian Ocean is studied using the Weather Research and Forecast (WRF) model. We have selected two intense tropical cyclones formed over the Bay of Bengal for studying the SST impact. Two different sets of SSTs were used in this study: one from TRMM Microwave Imager (TMI) satellite and other is the weekly averaged Reynold's SST analysis from National Center for Environmental Prediction (NCEP). WRF simulations were conducted using the Reynold's and TMI SST as model boundary condition for the two cyclone cases selected. The TMI SST which has a better temporal and spatial resolution showed sharper gradient when compared to the Reynold's SST. The use of TMI SST improved the WRF cyclone intensity prediction when compared to that using Reynold's SST for both the cases studied. The improvements in intensity were mainly due to the improved prediction of surface latent and sensible heat fluxes. The use of TMI SST in place of Reynold's SST improved cyclone track prediction for Orissa super cyclone but slightly degraded track prediction for cyclone Mala. The present modeling study supports the well established notion that the horizontal SST gradient is one of the major driving forces for the intensification and movement of tropical cyclones over the Indian Ocean.

Multi-spectral satellite image classification using Glowworm Swarm Optimization

This paper investigates a new Glowworm Swarm Optimization (GSO) clustering algorithm for hierarchical splitting and merging of automatic multi-spectral satellite image classification (land cover mapping problem). Amongst the multiple benefits and uses of remote sensing, one of the most important has been its use in solving the problem of land cover mapping. Image classification forms the core of the solution to the land cover mapping problem. No single classifier can prove to classify all the basic land cover classes of an urban region in a satisfactory manner. In unsupervised classification methods, the automatic generation of clusters to classify a huge database is not exploited to their full potential. The proposed methodology searches for the best possible number of clusters and its center using Glowworm Swarm Optimization (GSO). Using these clusters, we classify by merging based on parametric method (k-means technique). The performance of the proposed unsupervised classification technique is evaluated for Landsat 7 thematic mapper image. Results are evaluated in terms of the classification efficiency - individual, average and overall.