Técnicas de estimação do ângulo de chegada de sinais: uma análise comparativa

• Ano de defesa: 2008
• Autor(a) principal: Fontes, Rodrigo Tadeu
• Orientador(a): Eisencraft, Marcio
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Presbiteriana Mackenzie
• Palavras-chave em Português: sinais incidentes; técnicas de estimação; relação sinal-ruído; engenharia elétrica
• Palavras-chave em Inglês: incident signals; estimation techniques; signal-to noise ratio; electrical engineering
• Área do conhecimento CNPq: CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: http://dspace.mackenzie.br/handle/10899/24404

This work presents a comparative analysis of direction of arrival estimation techniques that use a Uniform Linear Array (ULA) of sensors. In this analysis, we use large ranges of parameters such as signal-to noise ratio, direction of arrival and number of sensors. Initially, the ULA model and its output are described. The estimation techniques (i) correlation, (ii) Multiple Signal Classification (MUSIC) and (iii) Estimation of Signals Parameters via Rotational Invariant Techniques (ESPRIT) are reviewed and the Cramer-Rao Lower Bound (CRLB) for this estimation problem is derived. The CRLB is the lowest Mean Square Error (MSE) achievable by any unbiased estimator. This limit is compared to the techniques MSE. Furthermore, we numerically obtain estimation resolution as a function of the ULA's number of sensors and the processing time for a pre-determined parameters estimation. We assume that the incident signals are corrupted by additive white gaussian noise. The results show that for only one signal reaching the ULA the correlation and MUSIC MSE are similar and closer the CRLB limit than ESPRIT MSE. Considering two signals reaching the array, we notice that the best choice depends on the signal-to-noise ratio. The same occurs when it comes to resolution. This time, the best method determination is related to the array's number of sensors. ESPRIT is the method that achieves the lowest processing time.