TABLE OF CONTENTS Although physical reasons differ with frequency and environment, the characteristics of signals received over most radio channels vary from instant-to-instant, and location-to-location (i.e. over time and space). If the responsible mechanisms change in an unpredictable manner, received signals resulting from deterministic transmissions, such as channel soundings, exhibit a random nature. Since the randomness is imparted by the channel, radio channels are often modelled as Stochastic Processess (StPs), herein referred to as Channel Processes (CPs). Statistical measures of a CP can be used in a number of ways, including the estimation of achievable performance on the channel of interest, and the generation of realisations of the CP for use in simulations.
On MIMO Links (MLks), where there are M antenna elements at the TX, and N elements at the RX, there are M N possible Physical Links (PLks) between different elements of the TX and RX arrays. When there is no motion of the antennas, or in the Critical Region (CrR) [3], randomness among transfer function models for the different PLks results because of the spatial separation of the antennas in the multipath field created by the presence of randomly located Interacting Objects (IOs). Variations in this randomness occur as the RX or TX arrays (or both) move, IOs in the CrR move, or as a result of both occurrences. Six types of randomness can be identified:
Type I: random changes that result when any of...
