Array signal processing in wireless communication has been a topic of interest in research for over three decades. In the fourth generation (4G) of the wireless communication systems, also known as Long Term Evolution (LTE), multi antenna systems have been adopted according to the Release 9 of the 3rd Generation Partnership Project (3GPP). For the fifth generation (5G) of the wireless communication systems, hundreds of antennas should be incorporated to the devices in a massive multi-user Multiple Input Multiple Output (MIMO) architecture. The presence of multiple antennas provides array gain, diversity gain, spatial gain, and interference reduction. Furthermore, arrays enable spatial filtering and parameter estimation, which can be used to help solve problems that could not previously be addressed from a signal processing perspective. The aim of this thesis is to bridge some gaps between signal processing theory and real world applications. Array processing techniques traditionally assume an ideal array. Therefore, in order to exploit such techniques, a robust set of methods for array interpolation are fundamental and are developed in this work. In this dissertation, novel methods for array interpolation are presented and their performance in real world scenarios is evaluated. Problems in the field of wireless sensor networks and vehicular networks are also addressed from an array signal processing perspective. Signal processing concepts are implemented in the context of a wireless sensor network. These concepts provide a level of synchronization sufficient for distributed multi antenna communication to be applied, resulting in improved lifetime and improved overall network behaviour. Array signal processing methods are proposed to solve the problem of radio based localization in vehicular network scenarios with applications in road safety and pedestrian protection.