Plausibility of assuming ideal arrays for direction of arrival estimation

In order to fulfil the increasing demands on capacity and coverage of mobile communication systems smart or intelligent antennas have been suggested. These antennas increase the spectral efficiency of a wireless system by using arrays of antenna elements to shape RF signals in particular directions. Generally, smart antennas can be divided into two approaches: switched multi-beam and adaptive. Adaptive antennas utilize sophisticated signal processing algorithms to continuously distinguish between the desired signal, multipath and interfering signal as well as to calculate their direction of arrival (DOA). This makes it possible to smoothly track users with main lobes and interferers with nulls and thereby constantly maximizing the SINR (signal-to-interference and noise ratio). The DOA is one of the most important parameter for an adaptive antenna system to estimate. It is also the main subject of this paper. We establish whether or not it is plausible to assume the array to be ideal (which is often assumed in array signal processing theory) when the received signals are actually collected with a physical array consisting of standard folded dipoles. In doing so we first measured the element patterns (or steering vectors) of a twelve element antenna array. The results were then used in array signal processing simulations. The performance of two DOA estimation algorithms, MUSIC and ESPRIT, are studied and compared.