The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Massive MIMO is a promising technology for 5G. While its advantages of increasing capacity and reducing output power are clear, understanding its SNR requirements in practical cases lacks sufficient theoretical support. This paper computes the relative power levels of the different signal, interference and noise terms in a Massive MIMO system using Maximum-Ratio Transmission (MRT). This brings light...
This report proposes a detection method based on using kurtosis and maximum amplitude information, both extracted from the spectral domain. At the end we face a classification problem which is solved by introducing a linear decision boundary. This new method is proved to result into overall good and stable detection results.
Accurate channel state information at the transmitter (CSIT) is essential to achieve multiplexing gains in multi-user (MU) MIMO systems through precoding at the base station (BS). In conventional frequency division duplexing (FDD) systems, CSIT is obtained through downlink pilot-aided channel estimation and limited uplink feedback. However, the resources consumed by training and feedback become unacceptable...
Reducing the size, weight, and power (SWAP) of airborne radar systems improves efficiency and decreases operational cost. Platform SWAP requirements are directly related to the complexity of algorithms used to process the received signal. Several reduced dimension space-time adaptive processing (RD-STAP) algorithms, which significantly decrease computations over traditional STAP, have been shown to...
Accurate classification or recognition of phase modulated radar waveforms, typically accomplished via the combination of pulse parameter estimates and matched filtering, poses a simple problem in ideal conditions. In less than ideal conditions, carrier frequency, time offset, pulse amplitude, initial phase, and bandwidth are unknown to the electronic warfare (EW) receiver rendering the application...
The Human Auditory System for speech recognition is highly robust against background noise compared to state-of-the-art Automatic Speech Recognition (ASR) systems. One of the best ways to add robustness to a speech recognition system is to have a compressed and highly robust feature set. In this paper, we present a novel approach for feature compression which makes the proposed noise-robust ASR system...
Based on Gibbs sampling technology, this paper puts forward a new semi-blind joint channel estimation and equalization algorithm that used in signal carrier coherent underwater acoustic communication system. First of all, we estimate the channel by using the train sequence with a loop structure before the data, and according to the sparse characteristics of underwater acoustic channel, this paper...
In this paper, we propose a practical hybrid precoding system to implement the MIMO transmission with data stream adaptation for high throughput millimeter-wave (mmWave) communications. Hybrid precoding by using both analog radio frequency (RF) precoding (beamforming) and digital baseband precoding is a practical way to achieve the required array gain and multiplexing gain, and to reduce system complexity...
Cognitive radio is the best solution for spectrum scarcity and spectrum underutilization over wireless communication challenges. It empowers secondary users to use primary user's spectrum without any interference. In this paper, a new approach of artificial neural network based spectrum sensing is proposed which senses the availability of a vacant channel in the primary user's spectrum and allocates...
The millimeter wave frequencies (roughly above 10 GHz) offer the availability of massive bandwidth to greatly increase the capacity of fifth generation (5G) cellular wireless systems. However, to overcome the high isotropic pathloss at these frequencies, highly directional transmissions will be required at both the base station (BS) and the mobile user equipment (UE) to establish sufficient link budget...
Machine learning based approaches for spectrum sensing and spectrum occupancy prediction in cognitive radio applications appear to have attracted sufficient interest in the current literature. In this paper, K-mean clustering based unsupervised learning method has been adopted for the performance enhancement of cooperative spectrum sensing in generalized к-μ fading channels. Extensive simulation has...
Underwater acoustic channels are characterized by a time-varying multipath structure with a long delay spread. The multipath is usually very sparse. Direct-adaptation based turbo equalizer is a low-complexity and effective method to combat the inter-symbol interference caused by the time-varying multipath. This paper studies the determinations of the tap positions of the linear equalizers in the direct-adaptation...
In this paper, a methodology for a texture-based classification between sand and rock images will be proposed, and an object recognition algorithm is added for the sand images. Eighteen types of sand and rock textures have been used to extract different type of features which are hypothetically should help in the classification process. The features discrimination ability were tested and ranked using...
This paper presents an efficient DNN design with stochastic computing. Observing that directly adopting stochastic computing to DNN has some challenges including random error fluctuation, range limitation, and overhead in accumulation, we address these problems by removing near-zero weights, applying weight-scaling, and integrating the activation function with the accumulator. The approach allows...
In this paper, we propose a multi-stage speech enhancement technique for speech recognition. At first, a multi-channel speech enhancement method takes advantage of the spatial information of speech source. Then, in the second stage, single-channel speech enhancement based on data-driven approach is adopted to improve performance of speech recognition at server side. This method can improve the quality...
Several distributed coordinated precoding methods relying on over-the-air (OTA) iterations in time-division duplex (TDD) networks have recently been proposed. Each OTA iteration incurs overhead, which reduces the time available for data transmission. In this work, we therefore propose an algorithm which reaches good sum rate performance within just a few number of OTA iterations, partially due to...
In this paper, we explore the potential of using deep learning for extracting speaker-dependent features for noise robust speaker identification. More specifically, an SNR-adaptive denoising classifier is constructed by stacking two layers of restricted Boltzmann machines (RBMs) on top of a denoising deep autoencoder, where the top-RBM layer is connected to a soft-max output layer that outputs the...
A method to detect an abnormal situation inside a public transport bus using audio signals is presented. Mel Frequency Cepstral Coefficients (MFCC) were used as a feature vector and a multilayer backpropagation neural network as a classifier. Audio samples were taken inside the bus running along Epifanio Delos Santos Avenue (EDSA), Metro Manila, Philippines. The audio samples depict sounds under normal...
Speech separation based on deep neural networks (DNNs) has been widely studied recently, and has achieved considerable success. However, previous studies are mostly based on fully-connected neural networks. In order to capture the local information of speech signals, we propose to use convolutional maxout neural networks (CMNNs) to separate speech and noise by estimating the ideal ratio mask of the...
Fine feature extraction is the basement of specific emitter identification, but existing researches rarely take the channel environment into consideration. To improve the practicability of the technology under the realistic channel, this paper researches on the method extracting of the fine feature under a simple multipath channel, which estimates the impulse of channel response by small-amplitude...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.