Indoor localization consists of locating oneself inside new buildings. GPS does not work indoors due to multipath reflection and signal blockage. WiFi based systems assume ubiquitous availability and infrastructure based systems require expensive installations, hence making indoor localization an open problem. This dissertation consists of solving the problem of indoor localization by thoroughly exploiting the indoor ambient magnetic fields comprising mainly of disturbances termed as anomalies in the Earth’s magnetic field caused by pillars, doors and elevators in hallways which are ferromagnetic in nature. By observing uniqueness in magnetic signatures collected from different campus buildings, the work presents the identification of landmarks and guideposts from these signatures and further develops magnetic maps of buildings - all of which can be used to locate and navigate people indoors. To understand the reason behind these anomalies, first a comparison between the measured and model generated Earth’s magnetic field is made, verifying the presence of a constant field without any disturbances. Then by modeling the magnetic field behavior of different pillars such as steel reinforced concrete, solid steel, and other structures like doors and elevators, the interaction of the Earth’s field with the ferromagnetic fields is described thereby explaining the causes of the uniqueness in the signatures that comprise these disturbances. Next, by employing the dynamic time warping algorithm to account for time differences in signatures obtained from users walking at different speeds, an indoor localization application capable of classifying locations using the magnetic signatures is developed solely on the smart phone. The application required users to walk short distances of 3-6 m anywhere in hallway to be located with accuracies of 80-99%. The classification framework was further validated with over 90% accuracies using model generated magnetic signatures representing hallways with different kinds of pillars, doors and elevators. All in all, this dissertation contributes the following: 1) provides a framework for understanding the presence of ambient magnetic fields indoors and utilizing them to solve the indoor localization problem; 2) develops an application that is independent of the user and the smart phones and 3) requires no other infrastructure since it is deployed on a device that encapsulates the sensing, computing and inferring functionalities, thereby making it a novel contribution to the mobile and pervasive computing domain.

No description available.

No description available.

No description available.

published_or_final_version / Physics / Doctoral / Doctor of Philosophy

We report here the observation of non-ohmic behavior in the dominant oscillatory component (the cigar component) of ρ('H) in ultrapure magnesium when the magnetic field, 'H, is parallel to [0001] and the current density, 'J, lies in the basal plane. In order to study this new phenomenon systematically, we had to overcome two experimental problems. The first was the design of an experimental probe which could reproducibly control, at low temperatures, the contact resistance at one of the points where current is injected into the sample without disturbing the sample's orientation with respect to 'H. Special micromanipulators, controlled by helium gas pressure, were designed into the probe for this purpose. The second problem was the construction of a detector which had the sensitivity to measure small signals from the magnesium samples in an effort to investigate the low current regime where the oscillations appear to satisfy Ohm's Law. A superconducting chopper amplifier was built which had the sensitivity to measure 10⁻¹¹ volt signals. We present evidence which directly relates this non-ohmic behavior to the long-range influence of a relatively large contact resistance at a point where current is injected into the sample. Data are presented which indicate that when this non-ohmic behavior is present, the corresponding oscillation amplitudes are proportional to the contact resistance. Measurements are also presented which show that the effects of this local current injection are so nonlocal that they extend over distances which are comparable to the dimensions of the sample.

Lightning electric field (E) and electric field derivative (dE/dt) signals have been recorded using sensors with 40 ns and 10 ns response times, respectively. Field propagation between the source and the recording antennas was entirely over salt water, so that distortions due to ground wave propagation were minimal below about 20 MHz. The fast-varying, initial portions of return stroke E fields have 10% to 90% risetimes which average 90 ± 40 ns. Peak dE/dt values range from 7 to 71 V/m/μs, with a mean and standard deviation of 33 ± 14 V/m/μs, when normalized to 100 km using an inverse distance dependence. The shapes of first and subsequent stroke fields are similar, but peak subsequent stroke dE/dts are larger than peak first stroke dE/dts in some flashes. The temporal structure of the fast varying fields produced by leader steps near the ground are very similar to return stroke fields. The mean maximum leader dE/dt, at 100 km, is 27 ± 9 V/m/μs. Large amplitude radiation fields produced by cloud discharge processes tend to be bipolar, with either positive or negative initial polarity and usually have several, fast, unipolar pulses superimposed on the initial half cycle. Cloud discharge fields with positive initial polarity usually precede cloud-to-ground flashes and produce a mean maximum dE/dt of 16 ± 8 V/m/μs. The field derivatives for all processes tend to be large when the amplitude of the associated fast field change is large. Estimates of lightning current derivatives, made using range normalized dE/dt measurements, average 155 ± 70 kA/μs, 135 ± 45 kA/μs, and 80 ± 40 kA/μs, for return strokes, leader steps, and cloud discharges, respectively, and a current wavefront velocity of 1 x 10⁸ m/s. These values are about 10 times larger than the maximum dI/dt recorded in strikes to instrumented towers. Lightning field amplitude spectra have been derived by Fourier analyzing dE/dt waveforms, and the spectral amplitudes decrease as 1/f² or faster with increasing frequency in the interval from about 6 to 20 MHz.

No description available.

No description available.

by Tong Shiu Sing Dominic. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1992. / Includes bibliographical references (leaves 242-245). / Acknowledgements --- p.iv / Abstract --- p.v / Chapter I --- Introduction --- p.1 / Chapter 1.1 --- A general review of the theory --- p.1 / Chapter 1.2 --- An outline of this thesis --- p.6 / Chapter II --- Dispersion surfaces of cold magnetoplasmas --- p.9 / Chapter 2.1 --- Meaning of dispersion surface and wavevector surface --- p.9 / Chapter 2.2 --- Dispersion surfaces of a two component electron-ion magnetoplasma --- p.13 / Chapter 2.3 --- Dispersion surfaces of a three component electron-ion-positron magnetoplasma --- p.35 / Chapter 2.4 --- Dispersion surfaces of a three component electron-ions magnetoplasma --- p.50 / Chapter 2.5 --- Doppler shifted wavevector surfaces (DWS) --- p.61 / Chapter A. --- Examples of DWS in an isotropic cold plasma --- p.61 / Chapter B. --- Examples of DWS in a cold magnetoplasma --- p.62 / Chapter 2.6 --- Dispersive surfaces of a moving magnetoplasma --- p.63 / Chapter III --- Evaluation of far field caused by a moving source --- p.72 / Chapter 3.1 --- Maxwell's equations and constitutive relations --- p.72 / Chapter 3.2 --- Calculation of far field by Lai and Chan's method --- p.75 / Chapter 3.3 --- Radiation energy flow --- p.85 / Chapter IV --- Controversy of Lai and Chan's method --- p.94 / Chapter 4.1 --- Origin of the controversy --- p.94 / Chapter 4.2 --- Evaluating the far field by the method of other authors --- p.97 / Chapter 4.3 --- "Comparsion of the fields found by Lai, Chan and other authors" --- p.100 / Chapter A. --- Comparing the far fields in an uniaxial non-dispersive medium --- p.101 / Chapter B. --- Comparing the far fields in an isotropic cold plasma --- p.104 / Chapter 4.4 --- Some remarks on the method of stationary phase --- p.109 / Chapter V --- Gyro-radiation in a cold magnetoplasma --- p.113 / Chapter 5.1 --- Introduction --- p.113 / Chapter 5.2 --- Radiation energy flux caused by a moving dipole in a magnetoplasma --- p.115 / Chapter 5.3 --- Radiation energy flux caused by a gyrating electron in a magnetoplasma --- p.135 / Chapter VI --- The ratio of emitted to received power in a magnetoplasma --- p.186 / Chapter 6.1 --- Introduction --- p.186 / Chapter 6.2 --- Methol of calculating the ratio of emitted to received power --- p.187 / Chapter 6.3 --- Numerical examples of the power ratio in a magnetoplasma --- p.191 / Chapter VII --- Evaluation of far field in a moving medium --- p.199 / Chapter 7.1 --- Introduction --- p.199 / Chapter 7.2 --- Far field expression in a moving medium --- p.200 / Chapter 7.3 --- Relation between Lai and Chan's far field and the far field in a moving medium --- p.206 / Chapter VIII --- Radiation in some moving media --- p.216 / Chapter 8.1 --- Introduction --- p.216 / Chapter 8.2 --- Radiation in a moving isotropic non- dispersive medium --- p.216 / Chapter 8.3 --- Radiation in a moving isotropic cold plasma --- p.223 / Chapter 8.4 --- Radiation in a moving cold magnetoplasma --- p.226 / Chapter IX --- Conclusions --- p.232 / Appendix A --- p.235 / Appendix B --- p.238 / Appendix C --- p.241 / References --- p.242