Ozgur Dural

Localization via TDOA in a UWB Sensor Network using Neural Networks

In an ultra-wide band (UWB) sensor network signal reflections from objects can be used to accurately determine the location. UWB signals are preferred in these types of sensor networks since they provide a very good resolution due to their fine time granularity. We propose an artificial neural network based localization algorithm to detect single object in a sensor network and compare its performance to Cramer-Rao bound and least squares estimator. Then we propose a two phase algorithm for multiple object detection and evaluate the algorithm for the case when there are two objects in a sensor network with three nodes.

Multi-Dimensional Constellations for Coded Transmission in Fading Channels

In high data rate OFDM systems, channel diversity can be a major concern. We revisit multi-dimensional constel- lation design, or short block spreading, and propose several improvements: a fine tuned Chernoff bound, a full optimization using Givens rotations, and a reduced complexity algorithm that exploits symmetry without loss of optimality. In fact, the problems complexity increases exponentially with number of dimensions, and previously proposed algorithms are either sub- optimal or incomplete. We also present a new interleaver scheme that continuously, and possibly adaptively, shuffles the sub- carrier grouping. Some of the ideas can be extended to the design of Space Time Codes. Index Terms—Diversity methods, frequency division multiplex- ing, interleaved coding, multidimensional coding.

Localization via multipath strengths in a CDMA2000 cellular network using neural networks

Localization is becoming more important with increasing number of cellular phone users. Due to safety aspects with increased emergency calls from mobile phones, new applications related to location based services, and the network optimization with increasing load, localization draws interest from both the academia and the industry. In this study, we propose a neural network based algorithm that uses multipath strengths to locate a mobile user without a GPS receiver. We validated our algorithm in a commercial network.

Scalable PLCP header extension within PSDU

Next generation UWB design needs additional header information fields than present in current ECMA-368 specification. However extending the ECMA-368 header beyond the available reserved bits while being compatible with legacy devices is a challenge. In this paper, we propose a scalable Physical Layer Convergence Protocol (PLCP) header extension within the Physical Layer Service Data Unit (PSDU) that allows for header extension beyond the reserved bits. The proposed mechanism also attempts to opportunistically convey the additional header information utilizing the padding bits present in the PSDU.

Compression schemes for the feedback channel in multiband-OFDM UWB systems

In time-varying channels, channel state information (CSI) is often provided via a feedback channel to facilitate dynamic allocation of resources like transmit power and data rates. In OFDM-based ultra wideband (UWB) channels, the signaling overhead for this CSI feedback can be extremely large. In this paper, we propose several compression schemes for the feedback channel that exploit the temporal and spatial correlation properties of the multiband-OFDM UWB channel, viz., fixed and variable spectral binning techniques, hierarchical binary tree-based spectral compression schemes, and Markov state-based differential temporal compression schemes.

MAC performance for second generation UWB

Next generation UWB design needs higher data rates than those proposed by ECMA-368 in order to be competitive. This paper investigates ECMA-368 MAC performance for extended data rates. It is shown that throughput improves at these extended PHY rates, when PHY frame size (L) is increased to 8190 octets under reasonable BER values. This improvement is a function of the transfer size. L=8190 octets exhibits >16% (>100 Mbps) improvement over L=4095 octets, for large transfers (>65 K octets). This is a worthwhile change since doubling PHY frame size imposes minor incremental complexity relative to the overall effort to support the extended PHY rates. It is shown also that scaling L does not improve throughput for bad channels (PER>12% normalized to 4095 octet payload). To maintain higher throughput under bad channel conditions, we need the ability to increase L without increasing PER. This can be achieved by packing multiple MPDUs into a single PSDU, each with its own frame check sum. This reduces PER by making it a function of the MPDU size and not PHY frame size. Complex MAC changes are needed for such a scheme. Fortunately, this scheme is only needed under bad channel conditions which are not suitable for extended data rates to start with.

Low power high speed UWB modem

With many emerging technologies focusing on high speed data delivery at short distance, power consumption of the UWB modem must be optimized for it to be a competitive solution. We show that we can trade extra link margin with savings on power consumption effectively by changing the constraint length of convolutional code and using QPSK when operating at data rates 320Mbps and up. We also present a new link adaptation algorithm using this flexibility, which provides 33% savings on power at LLR and Decoder blocks.

Extended data rates for next generation UWB

For UWB to be a competitive solution, its next generation design must support higher data rates than those proposed by ECMA 368. The purpose of this paper is to investigate methods to double the UWB current maximum data rate. The options considered are 16-QAM and bandwidth expansion. The potential solutions to double the data rates are compared in terms of system performance. Three metrics are used as figure of merit for such comparison.