Innovative Approaches for Optimized Performance in Time-Modulated Linear Arrays

In time�]modulated arrays, time is exploited as an additional degree of freedom for the array synthesis in order to better control the radiated beam. More specifically, by properly turning on and off the array elements according to suitable time sequences, the synthesis of patterns with low side lobe levels (SLLs) [1�]2] is obtained and the opportunity of optimizing the array performances in time�]varying wireless scenarios enabled. Unfortunately, time�]modulated arrays generate undesired harmonics, the so�]called sideband radiations (SR), which represent a non�]negligible loss of radiated power [3]. In order to overcome such a drawback and to improve the performances of the array, evolutionary optimization approaches have been recently and successfully applied. In [4] and [5], a Differential Evolution (DE) algorithm has been used to optimize the static�]mode coefficients as well as the durations of the time pulses leading to a significant reduction of the sideband level (SBL). In [6], the minimization of the SLL at the carrier frequency and of the SBL in uniformly excited time�]modulated arrays has been performed by means of a Simulated Annealing (SA) technique. In such a framework, a Genetic Algorithm (GA) based strategy has been considered in [7] to optimize the time sequences where the modulation period was subdivided in shorter time steps. This paper presents an innovative approach based on a Particle Swarm Optimizer (PSO) aimed at further enhancing the performances of time�]modulated linear arrays in terms of SBL reduction. From the mathematical formulation of the problem at hand, it follows that two parameters controls the synthesis process: the switch�]on intervals (i.e., the durations of the rectangular time pulses of the time modulating sequence) and the switch�]on instants (the times when the excitation coefficients switch on). More in detail, since the principal pattern at the carrier frequency is a function of the switch�]on intervals, whereas the SBL depends by both the parameters, the target of the synthesis procedure is the optimization of the switch�]on instants to reduce the SBL for a given principal pattern (i.e., fixed switch�]on intervals). In the following, the problem is briefly formulated and some selected results are reported to assess the effectiveness of the proposed approach