Performance of MIMO technology is highly dependent on the surrounding propagation environment and antenna configuration. In order to benefit from MIMO technology inside underground mines with low angular spread, careful antenna design and deployment strategies are required. In this regard, we studied the effect of uniform-linear-array (ULA) configuration on the MIMO channel capacity inside a short underground service tunnel. With its extensive infrastructure, this tunnel represents a typical midsize underground-mine tunnel. We simulated and measured channel frequency response at 2.49 GHz for grids of transmitters and receivers in different parts of the tunnel while taking into account practical considerations for access point (AP) deployments. All the ULA configurations were deployed either close to the sidewalls or under the ceiling. Among them, two configurations, which show the best performance for AP communications, were identified. For these configurations, we found that interelement separation requires to be four times longer than in conventional indoor environment with rich multipath (i.e., 2λ) to provide decorrelation among MIMO subchannels. We also found that infrastructure which has been neglected in previous studies, significantly degrades MIMO performance. Insightful results and guidelines for employing MIMO technology in underground mines can significantly benefit mining industry.

One of the biggest challenges facing wireless designers is determining how to configure multiple-input-multipleoutput (MIMO) antennas in underground-environments, such as mines, so that they deliver best performance. Here, we showed that angular dispersion of the signal, a phenomenon that greatly impacts channel capacity, can be accurately predicted in both near-field and far-field of underground tunnels using a multimode waveguide. Not only this analytical model is much faster than other methods, but also it does not have other shortcomings of measurement-based modeling (e.g., poor resolution for far distances in the tunnel). Further, we characterized angular-spread for different tunnel sizes and showed that the power-azimuthspectrum can be modeled by a zero-mean Gaussian distribution whose standard-deviation (angular-spread) is dependent on the tunnel dimension and the transmitter-receiver distance. Angular-spread is found to be a decreasing function of axial distance. After a certain distance, it becomes very small (about 4°) and independent of tunnel transverse dimension. Our results are useful for design of MIMO systems in underground tunnels and sufficient to permit the correlation-matrix required to extend the IEEE 802.11n MIMO channel model to underground environments.

Low angular spread of propagation environments and small interelement spacing within mobile terminals are two main factors degrading the capacity of multiple-input-multiple-output (MIMO) systems. This is particularly more evident in underground tunnels and mines that have lower angular spread compared to conventional environments. In such environments where colocated MIMO (C-MIMO) systems may not perform properly, we have shown that for access-point-to-mobile communications with limited and fixed antenna spacing at the mobile end, capacity close to that of an independent and identically distributed Rayleigh channel is achievable by employing distributed MIMO (D-MIMO) systems. To find and optimize C-MIMO and D-MIMO average capacities, we employed the waveguide-multimode model and particle-swarm-optimization method. We found that D-MIMO systems show desirable performance while significantly outperforming C-MIMO configurations. Nevertheless, D-MIMO desirable performance depends on the antennas' placement and is achievable for optimized locations. Based on the results, by uniformly distributing D-MIMO antennas along the tunnel axis and optimizing the transverse location of antennas afterward, suitable performance can be achieved. This paper is concluded by providing implications for D-MIMO antenna deployment in underground mines.

We show by simulation that when the fading signals observed on orthogonally polarized diversity branches follow Ricean statistics, the distribution of polarization states on the Poincare sphere is well-approximated by a Fisher distribution. Further, we show that the Fisher concentration parameter is: (1) completely determined by the corresponding Ricean Kfactors and the cross-correlation coefficient between the diversity branches, both of which can be estimated from simple measurements of received power vs. time, and (2) a good indicator of the level of cross-polar discrimination (XPD) on the channel. The insights gained are potentially useful to those engaged in the development and validation of schemes that use either polarization re-use or polarized MIMO.