A Novel Modification of PSO Algorithm for SML Estimation of DOA

Area of Science:

• Signal Processing
• Optimization Algorithms
• Array Signal Processing

Background:

• Stochastic Maximum Likelihood (SML) offers high accuracy for Direction-of-Arrival (DOA) estimation in sensor arrays.
• The computational complexity of SML hinders its application in real-time systems.
• Conventional Particle Swarm Optimization (PSO) also presents high computational demands for SML estimation.

Purpose of the Study:

• To significantly reduce the computational complexity of SML estimation for DOA.
• To improve the efficiency of SML algorithms for practical deployment in sensor array systems.
• To propose a novel optimization approach that overcomes limitations of existing methods.

Main Methods:

• A modified Particle Swarm Optimization (PSO) algorithm, termed Joint-PSO, is proposed.
• Joint-PSO utilizes Estimation of Signal Parameters via Rotational Invariance Techniques (ESPRIT) and stochastic Cramer-Rao bound (CRB) for initialization.
• This novel initialization space guides the optimization process, requiring fewer particles and iterations.

Main Results:

• The proposed Joint-PSO algorithm demonstrates a substantial reduction in computational complexity compared to conventional PSO, Altering Minimization (AM), and Genetic Algorithm (GA).
• Simulations confirm the superior efficiency of Joint-PSO in SML-based DOA estimation.
• The algorithm achieves near-optimal performance with significantly reduced computational load.

Conclusions:

• Joint-PSO effectively reduces the computational complexity of SML estimation for DOA.
• The proposed method shows great potential for enabling real-world applications of accurate SML-based DOA estimation.
• This advancement offers a more computationally feasible solution for high-accuracy DOA estimation in sensor arrays.