Noncoplanar Geometry for Mobile NLOS MIMO Ultraviolet Communication With Linear Complexity Signal Detection

In this paper, we take a first step toward the mobile support for nonline-of-sight ultraviolet (UV) links by introducing a mesh UV network framework. Spatial multiplexing noncoplanar multiple-input-multiple-output (MIMO) UV system architecture is designed geometrically for the mobile mesh network to overcome the data rate bottleneck induced by scattered UV channel. Based on the MIMO UV channel characteristics, we further propose a modified sphere decoding (SD) method to accomplish MIMO signal detection in practical spatial correlated channels at a linear computational complexity. We evaluate mobile MIMO channel responses under various geometric parameters and the feasibility of this system geometry in $2 \times 2$ and $4 \times 4$ cases. Comparison of the bit error rates (BERs) between the modified SD, zero forcing, and SD is also done. Numerical results demonstrate that compared with the traditional linear array system structure, fully multiplexing is much easier via the noncoplanar MIMO geometry under very realistic parameters and the modified SD significantly improves the MIMO signal detection complexity with bit BER performance loss.