Summary
International Symposium on Antennas and Propagation
2009
Session Number:1C2
Session:
Number:1C2-4
Minimising the Sideband Patterns of Time-Modulated Arrays
Alan Tennant,
pp.129-132
Publication Date:2009/10/21
Online ISSN:2188-5079
Summary:
In 1959 H E Shanks and R W Bickmore published a paper entitled ‘Four-dimensional electromagnetic radiators' [1] in which they considered time as an extra variable to control the radiation pattern from array antennas. This paper and subsequent publications [2,3] demonstrated how 'time-modulated' (or 'time-switched') arrays could be used to provide a simple and low-cost system for electronic beam steering in antenna arrays. In addition these early papers also demonstrated how time-modulated arrays could be used to synthesize low sidelobe radiation patterns using simple binary switching of the array elements [2,3]. The concept of a timemodulated array may be explained with reference to Figure 1, which shows a conventional linear array topology to which switches have been added in the feed network connecting the outputs of the array elements to the summer. If all the element switches are closed the array behaves as a normal linear array with uniform element amplitude weighting. To synthesize an effective timeaverage weighting function across the face of the array, the elements of the array may be switched on for a period corresponding to a conventional element amplitude weight. For example, if a particular array element has a relative amplitude weight of i a , then in the time-modulated array implementation the element would be switched on for a period corresponding to i 0 a T , where T0 is the overall switching period. Hence, the technique can be used to synthesize any conventional array weighting function. Furthermore, as the switches can be controlled electronically the modulating sequence can be easily reprogrammed to provide arbitrary weighting functions in real time and hence provides an adaptive array system. A fundamental problem associated with time-modulated arrays is that they generate harmonics, or sidebands, at multiples of the switching frequency. So, for example if the array operates at a CW frequency of and the array elements are time switched by a sequence with a period T0 then harmonics will be generated at frequencies given by +/- n T0 , where n represents the number of the harmonic. These harmonics are generally unwanted as they waste energy and may cause interference in other parts of the radio spectrum. Consequently, much of the recent work on time-modulated arrays has investigated methods for minimising sideband levels using various adaptive optimisation techniques to modify the switching period of individual elements [4-6]. Almost all the schemes reported in the literature use a similar approach in which each element of the array is time-switched using a common switch-on time. However if the element switch-on time is arbitrary then this introduces an extra degree of freedom in the design process. This type of approach was recently investigated by Yang et al [7] in which the elements of a 16 element array could be assigned multiple ‘ON-OFF’ states during a switching period. The concept of a variable element switch on time combined with sub-array switching strategies was has also been investigated by the author [8]. In this paper we extend the approach presented in [8], by investigating a method for minimising the sidebands generated by time-switched arrays by controlling the time-domain output from the array.