T. H. Ismail

Array Pattern Synthesis Using Digital Phase Control by Quantized Particle Swarm Optimization

A modified PSO algorithm called quantized PSO (QPSO) is described for the synthesis of antenna array patterns by digital phase shifters. The solution space of the QPSO algorithm is restricted to finite quantized integer values of the array phase coefficients. The QPSO search for an optimal solution within the available quantized values of the digital phase shifters to minimize the fitness function that includes the SLL value and interference suppression while maintains the main beam unchanged. The array factor is expressed using a linear transform based on discrete cosine transform (DCT) with pre-computed DCT matrix which allows a high speed computation of the fitness function. The results show better synthesis performance using QPSO, as compared to BPSO and efficient computation time for large arrays.

Genetically evolved phase-aggregation technique for linear arrays control

This paper presents a modified Genetic Algorithm (GA) technique in which the large phase perturbations are calculated by aggregating small phase increments. The proposed aggregation GA technique overcomes the major drawback of the large solution space required by the classical GA techniques. The proposed method adopts small ranges for increments of the parameters and the optimality is reached via aggregation of the best increments of phases. Consequently, the GA searches in a smaller solution space and finds the solution with reduced number of iterations. Simulation results show the achieved improvement of the proposed technique over the classical GA. The suppressed sectors using phase-only control are accomplished with and without element failures. Problems like imposing symmetrical nulls around the mainbeam and compensation for the failure of center element have been achieved.

Design of Linear Phased Array for Interference Suppression Using Array Polynomial Method and Particle Swarm Optimization

In this work, a linear phased array pattern design with null steering is achieved using the array polynomial technique and the Particle Swarm Optimization (PSO) algorithm. The null steering for interference suppression is obtained by controlling some of the roots on the Schelkunoff’s unit circle while keeping the roots responsible for the main beam unchanged. The rest of the roots are controlled to minimize the Side Lobe Level (SLL) of the array pattern using the PSO algorithm. It has been demonstrated that this technique achieved more than 50% reduction in the parameters needed to be optimized compared with the conventional complex coefficients optimization techniques. Consequently, the fitness function is only responsible for the SLL as the prescribed controlled nulls and the mainbeam characteristics are solved analytically. The simulated results show the effectiveness of the proposed technique.

Genetically Evolved Phase-Aggregation Technique for Linear Arrays Control

This paper presents a modified Genetic Algorithm (GA) technique in which the large phase perturbations are calculated by aggregating small phase increments. The proposed aggregation GA technique overcomes the major drawback of the large solution space required by the classical GA techniques. The proposed method adopts small ranges for increments of the parameters and the optimality is reached via aggregation of the best increments of phases. Consequently, the GA searches in a smaller solution space and finds the solution with reduced number of iterations. Simulation results show the achieved improvement of the proposed technique over the classical GA. The suppressed sectors using phase-only control are accomplished with and without element failures. Problems like imposing symmetrical nulls around the mainbeam and compensation for the failure of center element have been achieved.

Partial control for wide-band interference suppression using eigen approach

This paper presents a simple and efficient array pattern synthesis technique for multiple wideband interference suppression using only the edge elements of linear arrays. By minimizing the mean-square value of the pattern in the directions of interference, the eigenvector that corresponds to the minimum eigenvalue is computed to obtain the required current excitation of the edge elements, the number of computations is greatly reduced since partial amplitude control of the edge elements is used.

Digital Amplitude Control for Interference Suppression Using Immunity Genetic Algorithm

In this paper, we propose a novel genetic algorithm (GA) called immunity GA (IGA) for array pattern synthesis with interference suppression using digital amplitude only control. The IGA is based on crossover evolution where the crossover operator is a variant of the known GA operator. A new formulation of the array factor transform for a speciflc number of elements N is expressed by a discrete cosine transform (DCT) with pre-computed DCT matrix. Evaluating thousands of candidate solutions generated by the IGA using the precomputed DCT matrix will result in a high speed computation. This high performance allows us to flnd a good approximation of the absolute minimum SLL of synthesized arrays with digital amplitude control. Simulation results show the efiectiveness of this new algorithm for pattern synthesis with low SLL and null steering.

Pattern Synthesis with Phase-Only Control using Array Polynomial Technique

A new synthesis technique using the phase shifters of the linear array elements is developed using the array polynomial technique. The array factor is expressed as the product of sub polynomials such that their roots are located on the unit circle. All array elements are active for any arbitrary non-prime number of the array elements. The results show that the developed method in the form of an analytical solution can synthesize the prescribed patterns using the phase shifters of the linear antenna arrays.

Optimization of thinned arrays using stochastic Immunity Genetic Algorithm

In this paper we propose a novel genetic algorithm called Immunity Genetic Algorithm (IGA) based on stochastic crossover evolution to solve the synthesis problem of thinned arrays. Our crossover operator is a variant of the known GA operator. A new expression of the array factor for a specific number of elements N is expressed as a linear Discrete Cosine Transform (DCT). Using IGA to generate thousands of array bit patterns and the DCT to compute the fitness function will result in a very high speed computation compared to traditional computation techniques. This high performance allows us to find a good approximation of the absolute minimum SLL of synthesized thinned arrays. Simulation results of this novel array signal processing technique show the effectiveness for pattern synthesis with low SLL.

Null Steering with Element Failures Using Partial Controlled Linear Arrays

This paper deals with re-optimizing linear array patterns with multiple wideband null steering when arbitrary element failures occur. Using partial controlled arrays, the re-optimization method is formulated as a minimax problem solved by a linear programming technique. A weighting function is provided to control the relative error along the re-optimized array pattern. The results indicate that only a few elements are needed for the compensation. Furthermore, the suppressed sectors and the SLL can be improved by increasing the number of controlled elements. The results validate the capability of the proposed method to compensate for arbitrary element failures to re-create the multiple wideband nulls.

Maximum Likelihood Estimation of the Double-Directional Parameters in the Multiple-Input-Multiple-Output Communication System Using the Particle Swarm Optimization

Abstract In this article, we describe and evaluate the performance of the particle swarm optimization (PSO)-based maximum likelihood estimation (MLE), which will be used to estimate the angular parameters of the double-directional (DD) channel model in the multiple-input–multiple-output (MIMO) wireless system. The presented technique is shown to have a performance that is better than that achieved by the genetic algorithm (GA), and, hence, it yields an asymptotically efficient directional parameters estimate. The obtained estimates have approximately 2.5-dB saving in the threshold region over the GA-based MLE or the data-supported grid search. Another advantage of the PSO-based MLE technique is its simple implementation with few operations.