Andrew J. Wilkinson

Transmission-line load-network topology for class-E power amplifiers

Class-E amplifiers are a type of switching amplifier offering very high efficiency approaching 100%. In this paper, a topology and design methodology, which could be used for a transmission-line implementation of a class-E power amplifier, is presented. A simple transmission-line class-E load network is proposed that offers combined transformation of the load resistance down to a suitable level, as well as simultaneous suppression of harmonics in the load. The load network was developed and tested with the aid of a time-domain simulator (i.e., SPICE). A microstrip layout was designed and a first prototype was built operating at 1 GHz utilizing a field-effect transistor as the switching device. A drain efficiency of 72% was measured for our prototype after tuning, although better performance can be expected with an improved switching transistor and careful tuning of the load network.

Stepped-frequency processing by reconstruction of target reflectivity spectrum

This paper describes a processing technique for combining stepped-frequency waveforms efficiently to obtain higher range resolution. Essentially this method involves the reconstruction of a wider portion of the targets reflectivity spectrum by combining the individual spectra of the transmitted narrow-bandwidth pulses in the frequency domain. This paper describes the signal processing steps involved, and shows simulation results which validate and illustrate the method.

Electrical resistance tomography using a bi-directional current pulse technique

A novel electrical resistance tomography (ERT) measurement system has been developed, which uses a switched bi-directional constant current source to produce the electric field in the measuring vessel. Potential difference measurements are made on both the positive and negative half cycles and subtracted, cancelling any DC components present at the measuring electrode solution interface. The resulting voltage is a function of conductivity distribution in the medium. This technique is considerably simpler than conventional systems involving the injection of sinusoidal current followed by synchronous or DSP based detection. Further, because the current pulse is DC, the differential measurements do not require demodulating and therefore the technique has the potential for high-speed operation. The current pulse technique is justified in terms of the electrochemistry occurring at the electrodes and electronic circuitry used. The Newton-Raphson (NR) algorithm was used for reconstruction. Experiments were performed on a model system of (a) non-conducting object(s) in a conducting liquid, and on liquid foams of varying bubble size. Visualizations showed that the system could resolve the position and size of the objects, and identify coarse foam regions.

A 1000-measurement frames/second ERT data capture system with real-time visualization

This paper describes the design of a 16-electrode high-speed (1000 frames/s) electrical resistance tomography system with real-time visualization. The instrument utilizes a switched dc current pulse technique in conjunction with parallel data acquisition to achieve the high-data capture rates. The reconstruct algorithm is implemented using a single iteration Newton-Raphson method, which executes in under 1 ms. Data sets are presented that verify its operation. A calibration technique is described which improves the sensitivity of the current pulse measuring system and allows phenomena such as the dynamics of nonuniform slurries and gas distribution in aeration systems to be investigated. Furthermore, the calibration scheme described compensates significantly for the effect of impellers and baffles present in the measuring tank and allows more accurate reconstructions to be performed in the areas of interest.

The design of an ERT system for 3D data acquisition and a quantitative evaluation of its performance

This paper describes a multi-plane implementation of a current-pulse electrical resistance tomography (ERT) data capture system. This is achieved by extending a single plane system, with 16 electrodes and 16 parallel measurement channels, to a one capable of acquiring data in a specified sequence across multiple planes (up to eight) by inserting multiplexer modules in parallel between the instrument and the electrode array. This approach allows high-speed capture systems to be configured for applications such as dual plane cross-correlation velocity measurements or more complex current injection and measurement sequences yielding 3D data sets. The measurement timing and multiplexer measurement sequences are implemented by an embedded processor. Both the executable code and the measurement sequence tables are downloaded to the instrument at start-up. This allows flexibility in specifying the data acquisition sequences and timing required for specific applications without modification of the hardware or embedded code. The effect of measurement noise on the estimated conductivity is quantified and spatial resolution discussed for the case of a 2D online imaging algorithm. Example reconstructions from recorded data sets are presented which verify the operation of the instrument.

Design of a Sensor Coil and Measurement Electronics for Magnetic Induction Tomography

Magnetic induction tomography (MIT) is a tomographic imaging technique that is able to map the electromagnetic properties within an object or vessel from magnetic field measurements. Excitation coils are used to induce eddy currents in the medium, and the magnetic field produced by the induced eddy current is then sensed by receiver coils. Because of its noncontact nature, MIT is particularly attractive for biomedical and some industrial applications, such as pipe-flow monitoring, when compared with traditional contact electrode-based electrical impedance tomography. This paper describes the design and performance of an MIT transceiver circuit that can operate from 400 kHz to 12 MHz. The in-phase and quadrature (I/Q) demodulation technique is used to measure the signal perturbation due to the induced conduction eddy currents. The transceiver circuit design employs a single integrated circuit, containing a variable-gain amplifier and an I/Q demodulator. This paper contains characterizations of the transceivers measurement noise, system stability, and sensitivity for detecting saline solutions and metal plates. A novel balanced coaxial screened coil structure with integrated current sensing was also developed to minimize capacitive coupling between coils and to allow measurement of the current in the driving coils. Experiments were carried out at 3 and 10 MHz using bottles of saline solutions (1%-5% concentration) and metal sheets (aluminum and steel) to verify the sensitivity for conductivity imaging.

Interferometric SAR topographic mapping, a Bayesian model-based approach

A novel treatment of interferometric SAR as a Bayesian estimation problem is presented. Application of the Bayesian approach involves devising: a model relating system geometry, surface topography and reflectivity to the measured radar signals, a stochastic model of the topographic surface embodying the dominant terrain characteristics of slope and curvature and a stochastic model of the surface reflective properties. This information is combined into a conditional probability density function of the model parameters given the measured data. System models are proposed, and possible algorithms for estimating model parameters are considered. The issue of formulating an error estimate is discussed.

Synthetic aperture radar interferometry: a statistical model for layover areas

SAR interferometry (InSAR) for topographic mapping is in essence an inverse problem. The author has previously proposed a formal treatment of the problem within the framework of Bayesian inference. To implement this approach it is necessary to construct probabilistic models which relate the measured SAR data to the scene properties, i.e. the likelihood function P (data|/spl theta/). The authors now proposes a simple model for the joint statistics in layover areas (i.e. where two or more parts of the scene map to the same area in slant range).

The adaptive social hierarchy: a self organizing network based on naturally occurring structures

Wireless networks can be used for relaying information acquired by mobile animal borne tags. To date, no research has considered the large amount of diversity presented by the animal kingdom which impacts the design of the network. We consider here how the weight of the animal affects the size of the tag, and hence the energy that it can carry. We use a common structure in nature, the social dominance hierarchy, and form a similar hierarchy based on energy. Nodes adjust their perceived rankings through continual tournaments using simple, locally applied rules to result in a stable and adaptive structure. We show that the number of levels in the hierarchy controls traffic density and consequently energy usage. To further conserve energy of low ranked nodes, we propose a simple cross-layer protocol. We show through simulation that our power-aware protocol outperforms those with no knowledge of energy

EcoLocate: A Heterogeneous Wireless Network System for Wildlife Tracking

All research to date using wireless networks for wildlife tracking has concentrated on monitoring a single species, using large GPS enabled collars. These collars are too heavy to attach on smaller animals. Rather than omit small animals from the tracking spectrum, we show how a fusion of GPS tracking (where applicable) and an improved version of VHF tracking can result in a system which is able to track a wide range of animal species using the same underlying wireless network for information transfer. Tags are equipped with radio transceivers, which are used to both transmit and receive beacons. Received beacons are used to construct radio proximity maps which characterize co-location of various animals at different points in time. Furthermore, as the locations of some nodes are known, coarse estimates of animal locations can be determined, especially around focal points such as waterholes. We present the components of our system and discuss our prototype implementation.