A investigation on the correlation-coefficient and power metrics for MIMO antennas in a reverberation chamber

Abstract

Many MIMO antenna specifying aspects have been investigated. These aspects include capacity, diversity gain, BER between multiple antennas, K-factor and their mutual relations in both a reverberation chamber and an anechoic chamber. In this presentation, the correlation-coefficient (ρCC) of the transmission coefficients will be compared to the complex transmission-coefficient squared (TCS) by investigating the behavior of the coefficient of variance (C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">v</sub> ) for the two measures. The C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">v</sub> is calculated by normalizing the standard deviation by the mean for the different positions in the reverberation chamber. The transmission coefficients S <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sub> and S <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">23</sub> are obtained from a three-port vector network analyzer measurement done between two monopole antennas (connected on ports 1 and 2, on a single ground-plane) and a brass-horn antenna (on port 3). Three sets of monopoles are at fixed separation distances on a ground plane (called fix separation parallel-monopole antenna (FSPA)), and are representative of good, bad, and average multiple-input multiple-output (MIMO) antenna's. Another antenna with an adjustable separation distance between the monopoles (called variable separation parallel-monopole antenna (VSPA)), allows variation of the MIMO antenna over a range from good to bad. For both types of MIMO antennas, the ρCC as function of the monopole separation follows the trend of the analytic model (Bessel-function). The ρCC is better (i.e., a smaller value) at larger separations as can be seen in Figure 1a and 1b. With increasing separation the C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">v</sub> increases to values of approximately 0.4 to 0.5, which implies that the ρCC is becoming less accurate. However, if we look at the case were there are 6 typical RF absorbers-sections added for loading, the C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">v</sub> of the TCS almost doubles while the C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">v</sub> of the ρCC stays more-or-less the same. In conclusion, for a small separation between the monopoles, especially in a heavily loaded environment, the ρCC would likely exhibit less uncertainty as a performance metric for a MIMO system, than the TCS measurement.