On Multipath Delay Estimation for OFDM Channel Estimation using Beamforming Techniques
Author
Desiderio, Silvio
Term
10. term
Publication year
2010
Abstract
This thesis investigates the estimation of multipath delay components to improve channel estimation in pilot-aided OFDM systems. It compares the widely used Robust Wiener Filter (RWF) with an alternative method, the Enhanced Noise Reduction Algorithm (ENRA), using a derived discrete-time OFDM model. Simulations show that ENRA achieves lower bit error rates than RWF when the true delay parameters are known. To estimate these delays in practice, beamforming techniques are studied, with a focus on the iterative Sequential Beamforming Algorithm (SBA). The study evaluates SBA for different resolutions in the delay domain and finds that performance improves with finer resolution and by deliberately overestimating the number of channel echoes. Finally, simulations demonstrate that ENRA, when supplied with delays estimated by SBA, outperforms the state-of-the-art approach.
Denne afhandling undersøger estimering af flervejstidsforsinkelser for at forbedre kanalskøn i pilotbaserede OFDM-systemer. Den sammenligner den udbredte Robust Wiener Filter (RWF) med en alternativ metode, Enhanced Noise Reduction Algorithm (ENRA), baseret på en udledt diskret-tids OFDM-model. Simulationer viser, at ENRA opnår lavere bitfejlrater end RWF, når de sande forsinkelsesparametre er kendt. For at kunne estimere disse forsinkelser i praksis undersøges beamforming-teknikker, med særlig fokus på den iterative Sequential Beamforming Algorithm (SBA). Arbejdet vurderer SBA for forskellige opløsninger i forsinkelsesdomænet og viser, at ydeevnen forbedres med finere opløsning samt ved bevidst overestimering af antallet af kanalekkoer. Afslutningsvis demonstrerer simulationer, at ENRA, når den forsynes med forsinkelser estimeret af SBA, overgår den gængse metode.
[This apstract has been generated with the help of AI directly from the project full text]