I. Introduction
The purpose of over-the-air (OTA) testing is to assess the performance of wireless devices including antenna performance. Multiple-input–multiple-output (MIMO)-capable devices should be tested under realistic channel conditions, so that their true performance is assessed. The multiprobe anechoic chamber (MPAC) method is a promising solution to achieve this. In this method, a number of source antennas, referred to as probe antennas, are placed inside an anechoic chamber surrounding the device under test (DUT). Through the use of a channel emulator connected to the probe antennas, various spatial channels can be created [1]. The main disadvantage of the MPAC method is its cost. A sufficient number of probe antennas is required to accurately emulate the channel, which leads to costly designs [2]. In addition, the radius of the ring where the probes are placed is typically assumed to be sufficiently large so that waves radiated from the probe antennas are planar in the test area. Typical ring radius in practical OTA setups reported in the literature is, for example, 2 m in Aalborg University and ETS-Lindgren setups, to name a few [3]. A large multiprobe setup and consequently a large anechoic chamber are often cost-prohibitive.