A refined genetic algorithm for accurate and reliable DOA estimation with a sensor array

M. Li, Y. Lu

Research output: Contribution to journalArticle

20 Citations (Scopus)

Abstract

Maximum likelihood (ML) direction-of-arrival (DOA) estimation algorithm is a nearly optimal technique. In this paper, we present a modified and refined genetic algorithm (GA) to find the exact solutions to the complex, multi-modal, multivariate and highly nonlinear likelihood function. With the newly introduced features such as intelligent initialization and the emperor-selective mating scheme, carefully selected crossover and mutation operators, and fine-tuned parameters such as the population size, the probability of crossover and mutation, the GA-ML estimator achieves fast global convergence. The GA-ML estimator has been compared with various DOA estimation methods in a variety of scenarios, and the simulation results demonstrate that in most scenarios the proposed GA-ML estimator is the fastest and its performance is the best among popular ML-based DOA estimation methods.
LanguageEnglish
Pages533-547
Number of pages15
JournalWireless Personal Communications
Volume43
Issue number2
DOIs
Publication statusPublished - Oct 2007

Fingerprint

Direction of arrival
Sensor arrays
Maximum likelihood
Genetic algorithms

Keywords

  • array signal processing
  • direction finding
  • genetic algorithms
  • maximum likelihood

Cite this

@article{e159c3165f5f4190b8762b3ab978b16c,
title = "A refined genetic algorithm for accurate and reliable DOA estimation with a sensor array",
abstract = "Maximum likelihood (ML) direction-of-arrival (DOA) estimation algorithm is a nearly optimal technique. In this paper, we present a modified and refined genetic algorithm (GA) to find the exact solutions to the complex, multi-modal, multivariate and highly nonlinear likelihood function. With the newly introduced features such as intelligent initialization and the emperor-selective mating scheme, carefully selected crossover and mutation operators, and fine-tuned parameters such as the population size, the probability of crossover and mutation, the GA-ML estimator achieves fast global convergence. The GA-ML estimator has been compared with various DOA estimation methods in a variety of scenarios, and the simulation results demonstrate that in most scenarios the proposed GA-ML estimator is the fastest and its performance is the best among popular ML-based DOA estimation methods.",
keywords = "array signal processing, direction finding, genetic algorithms, maximum likelihood",
author = "M. Li and Y. Lu",
year = "2007",
month = "10",
doi = "10.1007/s11277-007-9248-5",
language = "English",
volume = "43",
pages = "533--547",
journal = "Wireless Personal Communications",
issn = "0929-6212",
number = "2",

}

A refined genetic algorithm for accurate and reliable DOA estimation with a sensor array. / Li, M.; Lu, Y.

In: Wireless Personal Communications, Vol. 43, No. 2, 10.2007, p. 533-547.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A refined genetic algorithm for accurate and reliable DOA estimation with a sensor array

AU - Li, M.

AU - Lu, Y.

PY - 2007/10

Y1 - 2007/10

N2 - Maximum likelihood (ML) direction-of-arrival (DOA) estimation algorithm is a nearly optimal technique. In this paper, we present a modified and refined genetic algorithm (GA) to find the exact solutions to the complex, multi-modal, multivariate and highly nonlinear likelihood function. With the newly introduced features such as intelligent initialization and the emperor-selective mating scheme, carefully selected crossover and mutation operators, and fine-tuned parameters such as the population size, the probability of crossover and mutation, the GA-ML estimator achieves fast global convergence. The GA-ML estimator has been compared with various DOA estimation methods in a variety of scenarios, and the simulation results demonstrate that in most scenarios the proposed GA-ML estimator is the fastest and its performance is the best among popular ML-based DOA estimation methods.

AB - Maximum likelihood (ML) direction-of-arrival (DOA) estimation algorithm is a nearly optimal technique. In this paper, we present a modified and refined genetic algorithm (GA) to find the exact solutions to the complex, multi-modal, multivariate and highly nonlinear likelihood function. With the newly introduced features such as intelligent initialization and the emperor-selective mating scheme, carefully selected crossover and mutation operators, and fine-tuned parameters such as the population size, the probability of crossover and mutation, the GA-ML estimator achieves fast global convergence. The GA-ML estimator has been compared with various DOA estimation methods in a variety of scenarios, and the simulation results demonstrate that in most scenarios the proposed GA-ML estimator is the fastest and its performance is the best among popular ML-based DOA estimation methods.

KW - array signal processing

KW - direction finding

KW - genetic algorithms

KW - maximum likelihood

U2 - 10.1007/s11277-007-9248-5

DO - 10.1007/s11277-007-9248-5

M3 - Article

VL - 43

SP - 533

EP - 547

JO - Wireless Personal Communications

T2 - Wireless Personal Communications

JF - Wireless Personal Communications

SN - 0929-6212

IS - 2

ER -