## Thesis etd-02132013-161532 |

Thesis type

Tesi di dottorato di ricerca

Author

USAI, PIERPAOLO

URN

etd-02132013-161532

Thesis title

On the use of electromagnetic asymptotic methods for the estimation of communications propagation channel in complex environments

Academic discipline

ING-INF/02

Course of study

INGEGNERIA 'L. DA VINCI'

Supervisors

**tutor**Prof. Monorchio, Agostino

Keywords

- ray tracing
- Physical Optics
- Asymptotic methods
- RCS

Graduation session start date

07/06/2013

Availability

Full

Summary

The ray tracing method and the high frequency theories this method is based on are recapitulated in chapter 1 in the formulation implemented in the software EMviroment-EMv used in this thesis to predict the electromagnetic propagation in complex environment. The estimation of channel parameters by asymptotic propagation model based software could show some drawbacks, such as long simulation times caused by the multipath reconstruction, and a huge memory dimension requirement to store the multipath info that is useful for the frequency analysis of the channel. A lot of energies were spent in optimizing and speeding up the ray tracing algorithms by researchers in the last decades. In the implementation of ray tracing based solver the study of the acceleration of the geometrical algorithm to predict the multipath has a great importance, because the physical coherence of this deterministic approach leads to a boost of the computational time. A method for speeding up a ray tracing based electric field prediction model suitable for urban environments investigation is described in chapter 2.

The physical coherence of this high frequency method allows to accurately estimate the wireless communication channel frequency response and the antennas influence on the transmitted and received signals. A frequency analysis of the channel is required, for example, in ultra-wideband applications. In this case the frequency selective behavior of the channel needs to be estimate to correctly predict the communication link impairment. The frequency response definition by multipath prediction is shown in chapter 3

The Doppler frequency shift is caused by the presence of relative movement by the transmitter, the receiver and the complex objects present in the scene. The power distribution in the Doppler frequency shift domain can be derived by the spreading function estimation. By definition the spreading function could be calculated starting from the knowledge of the impulse response of a channel link by means of the Fourier transform. An alternative direct deterministic approach for its estimation avoiding the Fourier transform has been formulated and is presented in chapter 4 applied to the Munich airport complex scenario.

The radar cross section evaluation of metal and dielectric objects needs to invoke the asymptotic techniques to overcome the limitations the full-wave techniques present at high frequencies. The method of moments, for example, needs a denser and denser mesh definition of the studied object as the wavelength decreases to respect the applicability constraints of the method. Because the complexity of this method is proportional to the currents on the facets, the number of unknowns can get to saturate the computer memory availability and increase the simulation time. The approximations that can be introduced with the assumption of far-field sources and observers at high frequencies allow to unburden the calculation procedures obtaining good estimations of the scattered field. The Physical Optics - PO theory is the most diffuse solution solving these kind of problems, often together with a ray bouncing analysis to predict the illuminated facets of the object. A PO based solver to predict the scattered field by a metallic or dielectric object and the radar cross section spectrum of rotating objects is presented in chapter 5.

The physical coherence of this high frequency method allows to accurately estimate the wireless communication channel frequency response and the antennas influence on the transmitted and received signals. A frequency analysis of the channel is required, for example, in ultra-wideband applications. In this case the frequency selective behavior of the channel needs to be estimate to correctly predict the communication link impairment. The frequency response definition by multipath prediction is shown in chapter 3

The Doppler frequency shift is caused by the presence of relative movement by the transmitter, the receiver and the complex objects present in the scene. The power distribution in the Doppler frequency shift domain can be derived by the spreading function estimation. By definition the spreading function could be calculated starting from the knowledge of the impulse response of a channel link by means of the Fourier transform. An alternative direct deterministic approach for its estimation avoiding the Fourier transform has been formulated and is presented in chapter 4 applied to the Munich airport complex scenario.

The radar cross section evaluation of metal and dielectric objects needs to invoke the asymptotic techniques to overcome the limitations the full-wave techniques present at high frequencies. The method of moments, for example, needs a denser and denser mesh definition of the studied object as the wavelength decreases to respect the applicability constraints of the method. Because the complexity of this method is proportional to the currents on the facets, the number of unknowns can get to saturate the computer memory availability and increase the simulation time. The approximations that can be introduced with the assumption of far-field sources and observers at high frequencies allow to unburden the calculation procedures obtaining good estimations of the scattered field. The Physical Optics - PO theory is the most diffuse solution solving these kind of problems, often together with a ray bouncing analysis to predict the illuminated facets of the object. A PO based solver to predict the scattered field by a metallic or dielectric object and the radar cross section spectrum of rotating objects is presented in chapter 5.

File

Nome file | Dimensione |
---|---|

PhD_Thesis_UsaiP.pdf | 4.61 Mb |

Contatta l’autore |