HF-MIMO Antenna Array Design and Optimization

MIMO as a new antenna array communication technology has been widely applied in modern communications, especially within UHF/VHF band. After several experimental campaigns, E.M. Warrington and S.D. Gunashekar proved that MIMO techniques could also be applicable within HF band. Further experiments showed that traditional widely spaced homogeneous antenna arrays could be replaced by co-located heterogeneous antenna arrays without significant reductions in data transmission rate. In other words, radiation pattern diversity can be used as a new kind of MIMO diversity to replace spatial diversity. In order to get a better understanding of this phenomenon, antenna modelling using numerical electromagnetics code (NEC) has been carried out in last three years. The study showed that phase difference difference (PDD) between collocated antenna array elements could be the key factor of radiation pattern diversity. The correlation level between array elements can be reduced with increased PDD between these elements. Several transmitting antenna arrays have been developed according to this study and tested via a 202 km radio link between Leicester and Lancaster. The experimental campaigns showed that the newly designed antenna arrays have a significantly increased de-correlation level between array elements compared with traditional antenna arrays. High performance computing (HPC) was used in the antenna modelling in order to investigate the relationship between antenna array geometry and phase difference difference. Several optimized antenna arrays with large phase difference difference were recommended from tens of million of different antenna geometries. This research gives the new direction of HF-MIMO antenna array design: in theory aspect, it clearly specifies what kind of pattern needs to be targeted in order to get decorrelation using pattern diversity; in antenna modelling aspect, HPC was for the first time applied in HF-MIMO antenna array design, which provided a brand new high level of computing platform for the future modelling work.