Joint Pose and Radio Channel Estimation
Forskningsoutput: Avhandling › Licentiatavhandling
A well established approach for pose estimation is using an inertial measurement unit (IMU). Using an inexpensive IMU standalone for dead reckoning pose estimation is tempting but it is not a working solution due to noise and other imperfections in the IMU. There is also a fundamental limitation of inertial sensors, they can not, because of Galileo's principle, obtain any information about absolute velocity of the device. To obtain reliable pose estimates for a longer time, the measurements from the IMU must be fused with some other sensor information. This thesis shows how the pervasive electric magnetic fields from existing radio communication systems such as the cellular mobile systems GSM, 3G, or 4G can be used.
Angle of arrival estimation using antenna arrays is a well studied problem with many different algorithms resolving the individual rays impinging on the array. However, less attention has been given to so called virtual array antennas where only one receiver element is used. By tracking the movement of the element, an array with properties similar to a stationary array with multiple elements is formed. By combining the IMU and the radio channel information, a map of the local radio environment can be obtained. At the same time, the map is used for adjusting for the errors in the IMU that lead to inaccurate pose estimates by using tightly coupled nonlinear state estimation algorithms from the sensor fusion framework.
The goals for this thesis is to develop a dynamic model for kinematics and a ray-trace based radio channel model that can be used together with the particle filter for sensor fusion. It also contains an initial investigation of limitations and achievable performance for the joint pose and radio channel estimation problem, including radio imperfection such as thermal noise, and phase/frequency error. The proposed model is evaluated using both simulations and datasets from experiments. The analysis of the evaluation shows that the proposed model, together with sensor fusion algorithms, provides a breakthrough in pose estimation using a low cost IMU.
|Enheter & grupper|
Ämnesklassifikation (UKÄ) – OBLIGATORISK
|Tilldelningsdatum||2013 jun 17|
|Status||Published - 2013|
Anders Holmqvist, Niklas Andersson, Anton Cervin, Anders Mannesson, Ather Gattami, Andrey Ghulchak, Alessandro Vittorio Papadopoulos, Anders Rantzer, Anders Robertsson, Aivar Sootla, ALFRED THEORIN, Bo Bernhardsson, Björn Olofsson, Björn Wittenmark, Christian Grussler, Charlotta Johnsson, Daria Madjidian, Erik Johannesson, Fredrik Magnusson, Fredrik Ståhl, Giacomo Como, Georgios Chasparis, Gabriel Turesson, Isolde Dressler, Johan Åkesson, Jang Ho Cho, Karl-Erik Årzén, Karl Johan Åström, Kin Cheong Sou, Karl Mårtensson, Karl Berntorp, Kristian Soltesz, Laurent Lessard, Martin Hast, Meike Rönn, Martin Ansbjerg Kjær, Martina Maggio, Maxim Kristalny, Olof Garpinger, Pål Johan From, Per-Ola Larsson, Pontus Giselsson, Rolf Johansson, Tore Hägglund, Vladimeros Vladimerou, Vanessa Romero Segovia, Andreas Aurelius, Gustav Cedersjö, Kaan Bür, Manfred Dellkrantz, Manxing Du, Payam Amani, Robin Larsson, William Tärneberg, Zheng Li, Lianhao Yin, Fredrik Tufvesson, Stefan Höst, Bernt Nilsson, Stig Stenström, Jens A Andersson, Stefan Diehl, Jonas Dürango, Mahdi Ghazaei Ardakani, Per-Ola Forsberg, Fredrik Bengtsson, Henrik Jörntell, Carmen Arévalo, Claus Führer, Christian Andersson, Fatemeh Mohammadi, Per Ödling, Mikael Andersson, Maria Kihl & Per Tunestål
2008/07/01 → 2018/06/30