TY  - THES
U1  - Dissertation oder Habilitation
A1  - Oshiga, Omotayo Olabowale
T1  - Efficient Advanced Indoor Localization: Analysis and Algorithms
N2  - Wireless localization is a very mature area of research, with plenty of work done in recent years both in academia and industry. Despite the amount of effort put into this problem, wireless positioning systems are still far off their potential as a real time locating technology (which requires automatic identification and tracking). It is commonly known that wireless localization systems are still inaccurate and unreliable in indoor environment, as a result indoor positioning systems are still quite frail and under-deployed.
One possible reason for this is that numerous constituents are available in the literature to solve parts of this problem, but still do not collectively combine to provide a complete solution. To qualify the above, two very important problems within the area of wireless localization have been treated as separate challenges. These problems are ranging (as defined by the process of estimating distances from physical quantities) and trilateration (as defined by the process of estimating the absolute location of sources given their distances to a set of references).
This is rationalized by the fact that the fundamental tools required to design accurate distance estimators and positioning algorithms are clearly distinct. From an error analysis point of view, these problems are intrinsically interdependent, by the reason of the fundamental limits on the root mean square error on the corresponding estimates (both distances and locations) being governed by the same likelihood function given as the product of the ranging error distributions. Therefore, an attempt at the unification of these problems, with the aim at improving the accuracy, precision, complexity and robustness of wireless localization for indoor positioning systems is reasonable.
During the course of this thesis, we provided estimation and reconstruction analyses on the statistics of the ranging error distributions, we then refrained from pursuing further positioning algorithms as both ranging and trilateration are governed by the same likelihood function, but rather presented efficient ranging and multipoint ranging techniques through the efficient collection of ranging information using Sparse and Golomb rulers obtained utilizing evolutionary genetic techniques, with the adjustment of the ranging techniques to provide highly accurate solutions which aim at improving the quality of distance estimation. We then pursued effective trilateration techniques which allow results and information typically restricted to the ranging problem, to inform positioning algorithms, thereby conditioning results in light of knowledge extracted from ranging information in order to provide accurate wireless localization. Therefore, we bridged and inter-connected both ranging and trilateration methods, which resulted in efficient, robust, accurate, precise and low-complex ranging and trilateration techniques for advanced indoor localization.
AB  - Wireless localization is a very mature area of research, with plenty of work done in recent years both in academia and industry. Despite the amount of effort put into this problem, wireless positioning systems are still far off their potential as a real time locating technology (which requires automatic identification and tracking). It is commonly known that wireless localization systems are still inaccurate and unreliable in indoor environment, as a result indoor positioning systems are still quite frail and under-deployed.
One possible reason for this is that numerous constituents are available in the literature to solve parts of this problem, but still do not collectively combine to provide a complete solution. To qualify the above, two very important problems within the area of wireless localization have been treated as separate challenges. These problems are ranging (as defined by the process of estimating distances from physical quantities) and trilateration (as defined by the process of estimating the absolute location of sources given their distances to a set of references).
This is rationalized by the fact that the fundamental tools required to design accurate distance estimators and positioning algorithms are clearly distinct. From an error analysis point of view, these problems are intrinsically interdependent, by the reason of the fundamental limits on the root mean square error on the corresponding estimates (both distances and locations) being governed by the same likelihood function given as the product of the ranging error distributions. Therefore, an attempt at the unification of these problems, with the aim at improving the accuracy, precision, complexity and robustness of wireless localization for indoor positioning systems is reasonable.
During the course of this thesis, we provided estimation and reconstruction analyses on the statistics of the ranging error distributions, we then refrained from pursuing further positioning algorithms as both ranging and trilateration are governed by the same likelihood function, but rather presented efficient ranging and multipoint ranging techniques through the efficient collection of ranging information using Sparse and Golomb rulers obtained utilizing evolutionary genetic techniques, with the adjustment of the ranging techniques to provide highly accurate solutions which aim at improving the quality of distance estimation. We then pursued effective trilateration techniques which allow results and information typically restricted to the ranging problem, to inform positioning algorithms, thereby conditioning results in light of knowledge extracted from ranging information in order to provide accurate wireless localization. Therefore, we bridged and inter-connected both ranging and trilateration methods, which resulted in efficient, robust, accurate, precise and low-complex ranging and trilateration techniques for advanced indoor localization.
KW  - Wireless Communication
KW  - Wireless Localization
KW  - Ranging and Positioning
KW  - Signal Processing
KW  - Genetic Algorithm and Optimization
Y1  - 2015
U6  - https://nbn-resolving.org/urn:nbn:de:gbv:579-opus-1004864
UN  - https://nbn-resolving.org/urn:nbn:de:gbv:579-opus-1004864
ER  -