Spyros Kyperountas

Locating the nodes: cooperative localization in wireless sensor networks

Accurate and low-cost sensor localization is a critical requirement for the deployment of wireless sensor networks in a wide variety of applications. In cooperative localization, sensors work together in a peer-to-peer manner to make measurements and then forms a map of the network. Various application requirements influence the design of sensor localization systems. In this article, the authors describe the measurement-based statistical models useful to describe time-of-arrival (TOA), angle-of-arrival (AOA), and received-signal-strength (RSS) measurements in wireless sensor networks. Wideband and ultra-wideband (UWB) measurements, and RF and acoustic media are also discussed. Using the models, the authors have shown the calculation of a Cramer-Rao bound (CRB) on the location estimation precision possible for a given set of measurements. The article briefly surveys a large and growing body of sensor localization algorithms. This article is intended to emphasize the basic statistical signal processing background necessary to understand the state-of-the-art and to make progress in the new and largely open areas of sensor network localization research.

Performance Analysis of Cooperative Spectrum Sensing in Suzuki Fading Channels

This paper examines the performance of cooperative spectrum sensing, using energy detection, in Suzuki fading channels. Sub-optimal centralized detection approaches are examined where decisions are made based on identical tests performed at the individual radios. Decisions are performed at a fusion center using a counting rule that encompasses the OR, AND, and majority rules as special cases. Analytical and simulation results are presented for Rayleigh, Log-normal and Suzuki distributions.

Performance analysis of relative location estimation for multihop wireless sensor networks

In this paper, we present new analytical, simulated, and experimental results on the performance of relative location estimation in multihop wireless sensor networks. With relative location, node locations are estimated based on the collection of peer-to-peer ranges between nodes and their neighbors using a priori knowledge of the location of a small subset of nodes, called reference nodes. This paper establishes that when applying relative location to multihop networks the resulting location accuracy has a fundamental upper bound that is determined by such system parameters as the number of hops and the number of links to the reference nodes. This is in contrast to the case of single-hop or fully connected systems where increasing the node density results in continuously increasing location accuracy. More specifically, in multihop networks for a fixed number of hops, as sensor nodes are added to the network the overall location accuracy improves converging toward a fixed asymptotic value that is determined by the total number of links to the reference nodes, whereas for a fixed number of links to the reference nodes, the location accuracy of a node decreases the greater the number of hops from the reference nodes. Analytical expressions are derived from one-dimensional networks for these fundamental relationships that are also validated in two-dimensional and three-dimensional networks with simulation and UWB measurement results.

Link Maintenance Protocol for Cognitive Radio System with OFDM PHY

For secondary usage of spectrum in a cognitive radio system, in order to maintain reliable continuous communication among secondary users, secondary radios need to make contact, communicate and switch to an open RF channel simultaneously when the current channel is reclaimed by a primary user. In this paper, we describe a link maintenance protocol aimed at solving this issue. The communication channel is monitored and dynamically updated based on spectrum availability. Spectral sensing is employed at the receivers to monitor the spectrum and assist in RF channel synchronization between transmitter and receiver. Simulation results characterizing the performance of this protocol are provided. A prototype system with a pair of cognitive radios featuring this protocol is also introduced.

Implementing an Experimental Cognitive Radio System for DySPAN

Motorola Labs implemented an experimental cognitive radio system for use at the Dynamic Spectrum Access Networks (DySPAN) Symposium 2007. To exercise a practical communication link in a shared spectrum, the system provided a live video feed via a dynamically-allocatable OFDM physical layer. Each demonstration unit included an RF signal conditioning module for DySPAN channel allocations, a custom RF transceiver IC, digital signal processing, and an embedded Linux operating system. Connected to the units via Ethernet, a PC presented a graphical user interface, including the video feed and visualization of the spectral sensing, signal detection, and frequency allocation that was taking place in the units.

Performance comparision between TOA ranging technologies and rssi ranging technologies for multi-hop wireless networks

This paper investigates the relationship between the number of hops from a source node to a destination node and the range estimation accuracy for multi-hop wireless networks with a focus on RSSI-based and TOA-based systems. Analytical results show a dramatic difference between RSSI based ranging and TOA based ranging. With TOA-based systems, it is found that by fixing the source to destination distance, the more number of hops between the source and the destination, the worse the ranging estimation accuracy. In contrast, with RSSI-based location systems, the more number of hops, the better the ranging estimation accuracy. Simulations and experimental measurements employing prototypical UWB and 802.15.4 devices are used to validate the findings. Keywords-TOA; RSSI; networks; location; ranging.

Location estimation in multi-hop wireless networks

We present the analytic and simulation results of the performance of relative location estimation in multi-hop wireless sensor networks. It is found that the location estimation accuracy has a fundamental upper boundary. Location accuracy improves by adding more mobile nodes into the network. However, the resulting accuracy converges towards a fixed asymptotic value that is determined by the total number of links to the reference nodes. It is also established that for multi-hop networks, the location accuracy of a node is highly dependent upon the number of hops it is away from the reference nodes. The further away it is from the reference nodes, the worse the location accuracy.

RF Propagation Simulation in Sensor Networks

Itpsilas rare that real-world sensor network deployments confirm the academic assumptions that are made regarding uniformly, or randomly, distributed nodes and their node-pair signal strength measurements. During simulation time, these assumptions have grave affects on network initialization time, end-to-end delay, number of connections per node, network depth, throughput and so on. This paper introduces a simple RF propagation model that accurately predicts peer-to-peer connectivity based on signal strength. We show that network simulations based on this model agree favorably with real-world deployments.

Range-Free Location Estimation Algorithms for Wireless Networks

In this paper, we describe a suite of new range-free location estimation algorithms. With these algorithms, a blindfolded node is first located within some probable regions by comparing the measured received signal strength between nodes. Then the location of the blindfolded node is narrowed down by overlapping all the probable regions. Finally the geographic center of the overlapped region is calculated as the location estimate of the blindfolded node. Location estimation accuracy can be further improved when a blindfolded node that has already been located serves as additional reference node to determine the positions of other blindfolded nodes within its communication range. When compared with other range-free algorithms, results show improved location estimation accuracy with similar or less computation complexity.

Rigid Body Based Location Technology for Ad Hoc Sensor Networks

As an optimization problem, precision location requires sufficient constraints to warrant unique location estimation. The algorithm to determine the constraint sufficiency is the locatability algorithm. For the classic triangulation in two dimensions, locatability algorithm examines if a sensor node has at least 3 non-collinear reference node (RN) neighbors. This condition is often not met in most ad hoc sensor networks due to the low RN density. Progressive location was developed to turn a located sensor node into an induced RN which in turn is used to locate other sensor nodes. But even after applying progressive location, a lot of sensor nodes are still left un-locatable. A holistic approach, the rigid body (RB) based location technology, is proposed to group together sensors and RNs in a sensor network to form globally rigid bodies (GRBs) and cooperatively estimate sensor locations. The key differentiator of the technology is its locatability algorithm, a bottom-up procedure to identify GRBs in an anchor-free network and to determine the locatabilities of GRBs by grounding the network. The algorithm consists of four processes (node categorization, bilateration extension, trilateration extension, and tri-connectivity test) and locatability rules. It is shown that a bilateratively rigid sub-network is a strongly rigid graph and requires only the tri-connectivity to become globally rigid. Rules are provided for the locatability determination of rigid bodies and their associated sensor nodes. Simulation results show that the RB-based location algorithm locates drastically more sensor nodes than triangulation and progressive location algorithms especially when RNs are sparse