one, as compared with the conventional case. This is due to the increased surface area of the rounded edge portions without a corresponding increase in the actual diameter. CONCLUSIONS In this work we have explored the design of a new discone-type antenna [3]. When compared with the performance of the conven- tional discone’s configuration, the design examples have shown that it is possible to obtain, simultaneously, wider operational frequency band, lower return loss at the coaxial feeder, and reduc- tion of the antenna’s volume. This improvement in the perfor- mance was obtained by exploring the additional freedom provided by the shaping of the rotationally symmetric metallic structure. ACKNOWLEDGMENT This work was supported by the Brazilian National Research Council (CNPq) under Covenant PRONEX 664041/1996-S and Projects 462669/00-9 and 470495/2001-4. REFERENCES 1. J.J. Nail, Designing discone antennas, Electron (1953), 167–169. 2. H. Kawakami and G. Sato, Broad-band characteristics of rotationally symmetric antennas and thin wire constructions, IEEE Trans Antennas Propagat 46 (1998), 782–787. 3. K. Snyder and G.L. Peisley, Compact omnidirectional antenna, U.S. Patent 5140334. 4. K.A. Iskander, L. Shafai, A. Frandsen, and J.E. Hansen, Application of impedance boundary conditions to numerical solution of corrugated circular horns, IEEE Trans Antennas Propagat AO-30 (1982), 366 –372. 5. R.E. Collin, Field theory of guided waves, 2 nd ed., IEEE Press, New York, 1991. © 2003 Wiley Periodicals, Inc. ANALYSIS OF DIFFERENT GA STRATEGIES APPLIED TO ANTENNA FAR-FIELDS RECONSTRUCTION FROM PLANAR ACQUISITION J. R. Pe ´ rez and J. Basterrechea Department of Communications Engineering ETSIIT University of Cantabria Avda. de Los Castros s/n 39005 Santander, Spain Received 9 May 2003 ABSTRACT: The performance of a binary-coded microgenetic algo- rithm (GA) applied to planar near-field antenna characterization is ana- lyzed in this paper. The method replaces the antenna with an equivalent model consisting of short dipoles, whose optimization is carried out us- ing different GA strategies. Results of reconstructed far-field patterns from synthetic near-field data and a detailed analysis of the GA-based approach are included. © 2003 Wiley Periodicals, Inc. Microwave Opt Technol Lett 39: 422– 426, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/mop.11237 Key words: antenna measurements; near-field; far-field; equivalence principle; genetic algorithms 1. INTRODUCTION Genetic-algorithm (GA) optimizers are stochastic search methods based on the principles of natural evolution, which are particularly effective when searching for a global solution in a high-dimension function domain, less prone to converge to a weak local optimum than deterministic optimization methods, and used extensively in electromagnetics [1, 2]. For classical or conventional genetic algorithms (CGAs), a population or set of potential solutions is made to evolve as a result of the pressure exerted by selection, crossover, and mutation operators, using a fitness function to assess the accuracy of each individual. For computationally intensive design problems, the computational time depends mainly on the number of function evaluations, and therefore, on the population size. CGAs require large populations to reach near-optimal results, in order to preserve diversity among the solutions’ spaces, according to the mutation scheme used, and to avoid premature convergence. This fact makes this technique unacceptably slow when dealing with com- plex problems, because factors such as the population size and the number of generations place a considerable burden on computa- tional time and resources. A successful alternative approach for use with small population sizes is the microgenetic algorithm Figure 7 Radiation pattern at (a) 0.9 GHz and (b) 3.6 GHz: (____) optimized; (- - - -) conventional 422 MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 39, No. 5, December 5 2003