2015. 3rd Issue

Table of contents 

Full issue  (13MB)



H. Charaf and S. Imre
Special Issue on Advanced Wireless and Mobile Technologies and Services 



S. Baumgartner, Y. El Hajj Shehadeh, and G. Hirtz
Robust CP-based Synchronization for DAB/DAB+ Systems over Dispersive Fading Channels 
In Orthogonal Frequency Division Multiplexing (OFDM) systems, symbol timing errors and frequency offsets may lead to a significant performance degradation. For this reason, it is very important to achieve a high precision time and frequency synchronization. With the absence of pilot symbols, blind synchronization algorithms using different metric functions or statistics of the received signal can be used. This paper investigates blind synchronization over dispersive fading channels by utilizing the redundant information contained in the cyclic prefix (CP). Whereas some proposed methods require channel information, this paper targets the joint estimation of the symbol timing offset, the frequency offset and the channel length. Starting by a previously proposed Maximum Likelihood method, we show that the approximated Log-Likelihood (LL) function exhibits a plateau and not a global maximum, leading to significant fluctuations in the estimated parameters. Yet, the symbol timing offset and the channel length characterize the position and dimensions of this plateau. Based on this finding, a novel method is developed for the joint estimation of the symbol timing offset, the frequency offset and the channel length from the approximated LL function. In addition, a second CP-based synchronization method is proposed, based on a modified timing function that targets the estimation of the parameters from the interference-free region within the cyclic prefix. To demonstrate the performance of the proposed methods and provide a fair comparison with the most recent CP-based synchronization algorithms, Monte Carlo simulations are conducted using the OFDM-based Digital Audio Broadcasting (DAB/DAB+) system under different realistic multipath fading channel conditions. The results show that the proposed algorithms outperform considerably other algorithms in related work and are robust to varying channel conditions.

B. Horváth and P. Horváth
Establishing Lower Bounds on the Peak-to-Average-Power Ratio in Filter Bank Multicarrier Systems 
Filter bank multicarrier (FBMC) modulation is a promising candidate as the primary physical layer waveform in next generation broadband networks. However, FBMC, like other multicarrier schemes, suffer from high peak-to-average power ratio (PAPR). Due to inherent overlapping of the time domain FBMC symbols, the methods designed for PAPR reduction in orthogonal frequency division multiplexing (OFDM) can not be applied in a straightforward manner for FBMC. In this paper, an optimization method is proposed to obtain significant PAPR reduction for short FBMC frames. The formulation of the optimization problem enables the use of both tone reservation and active constellation extension, which are well-known methods for OFDM PAPR reduction. The results establish a practical lower bound on achievable PAPR with these two approaches.

Q. Wang, D. Debbarma, A. Lo, I. Niemegeers, and S. Heemstra de Groot
Sum-rate Performance of Large Centralized and Distributed MU-MIMO Systems in Indoor WLAN 
Large MIMO systems are recognized as an effective technique for increasing the spectral and energy efficiency of wireless networks. The attractiveness of this technique for WLAN is that it can be an alternative approach to cell densification for providing high data rate wireless access. Here we consider large MIMO systems in indoor WLANs for multi-user MIMO (MUMIMO) spatial multiplexing in the 2.4 GHz ISM band. The focus is on analyzing the behaviors of large MIMO systems with both centralized and distributed antenna system (CAS and DAS) architectures. The analysis is based on extensive ray-tracing channel simulations, as well as an i.i.d. Rayleigh model. The numerical results show that the optimum capacity can be closely approached with both CAS and DAS architectures when the number of Access Point (AP) antennas exceeds the users by a few times. DAS is found to be superior to CAS in that the required number of antennas is significantly smaller, and especially performs better in non-rich scattering channels which is the case of practical channels reflected by ray-tracing simulations.

J. Marchang, B. Ghita, and D. Lancaster
Dynamic Queue Utilization Based MAC for Multi-Hop Ad Hoc Networks 
The end-to-end throughput in single flow multi-hop Ad Hoc networks decays rapidly with path length. Along the path, the success rate of delivering packets towards the destination decreases due to higher contention, interference, limited buffer size and limited shared bandwidth constraints. In such environments the queues fill up faster in nodes closer to the source than in the nodes nearer the destination. In order to reduce buffer overflow and improve throughput for a saturated network, this paper introduces a new MAC protocol named Dynamic Queue Utilization Based Medium Access Control (DQUB-MAC). The protocol aims to prioritise access to the channel for queues with higher utilization and helps in achieving higher throughput by rapidly draining packets towards the destination. The proposed MAC enhances the performance of an end-to-end data flow by up to 30% for a six hop transmission in a chain topology and is demonstrated to remain competitive for other network topologies and for a variety of packet sizes.

P. Sarcevic, L. Schaffer, Z. Kincses, and Sz. Pletl
Hierarchical-distributed approach to movement classification using wrist-mounted wireless inertial and magnetic sensor 
Wireless Sensor Networks (WSN) can be used for patient monitoring, analysis of daily activities, and emergency or fall detection. Using a WSN of two wrist mounted 9-degree-offreedom (9DOF) sensor boards, movement classification can be reliably done. The sensor boards or motes contain a tri-axial magnetometer, a tri-axial gyroscope, and a tri-axial accelerometer. If the classification is assigned to only one mote, which is using the data from both sensor boards, high energy consuming wireless data transfer is required. In this paper, a hierarchical-distributed algorithm is presented, where the motes are calculating their own movement classes, which can be combined on one mote, to determine the movement of the entire body and arms. The proposed method requires less and smaller classifiers, which can be easily implemented on low performance motes. Eleven movement classes were constructed, and data were collected with the help of nine subjects. By distributing the process, some movements can be merged and seven classes can be defined for each arm. Their combination determines the class of the entire body. Two classification hierarchies were tested and various Time-Domain Features (TDF) were calculated with different processing window widths. Altogether 48 training and validation data sets were constructed by different configurations of the sensors. The Minimum Distance (MD) with usage of the Linear Discriminant Analysis (LDA) dimension reduction method and the MultiLayer Perceptron (MLP) classifiers with and without LDA were tested.



A. Mandal and R. Mishra
Design of Pipelined Adaptive DFE Architecture For High Speed Channel Equalization 
An adaptive equalizer is a vital and esential component for high-speed reliable data transmission through wired as well as wireless communication channels to minimize Inter Symbol Interference (ISI). The non-linear Adaptive Decision Feedback Equalizer (ADFE) finds enormous use in applications where the channel distortion is severe for a linear equalizer to handle. High hardware cost of the system makes an urge to redesign it in order to minimize overall circuit complexity and cost. In this paper, realization of COordinate Rotation Digital Computer (CORDIC) based pipelined ADFE architecture using reformulated least-mean-square (LMS) algorithm has been presented. The realization results area efficient, better speed and throughput with good convergence. The efficacy of the proposed equalizer is corroborated with MATLAB simulations for mitigation of severe channel distortion arise due to ISI in high speed communication systems.

N. Batfai, R. Besenczi, A. Mamenyak and M. Ispány
Traffic Simulation based on the Robocar World Championship Initiative 
Robocar World Championship or briefly OOCWC is a new initiative to create a community of people who share their interest in investigating the relationship between smart cities and robot cars with particular attention to the spread of robot cars in the near future. At the heart of this initiative is the Robocar City Emulator. It is intended to offer a common research platform for the investigation of the smart city simulations. In this paper, we review the recent advances of the OOCWC.



Special Issue on Applied Cryptography and Security

European Signal Processing Conference – EUSIPCO 2016, Budapest

IEEE International Symposium on Personal, Indoor & Mobile Radio Communications – PIMRC 2016


Guidelines for our Authors

Technical Co-Sponsors





National Cooperation Fund, Hungary