Antenna placement for line-of-sight MIMO communications and sensing applications

Satisfying the demand for higher data rates is one of the long-standing goals of wireless communication systems. A shift towards higher frequency bands, such as millimeter wave (mmWave) and sub-THz bands, is one of the main enablers for this, due to the large amounts of available bandwidth in these bands. In addition, multiple-input multipleoutput (MIMO) technology has been one of the cornerstones of wireless communication systems, since it can also offer a tremendous increase in the capacity. However, as multi-path propagation is very limited at mmWave and sub-THz, the channel may be dominated by the line-of-sight (LoS) path only, thereby leading to rank deficient MIMO channels. Yet, as the rank of a LoS MIMO channel depends on the antenna positions and the distance between the transmitter and receiver, it has been shown that a spatial multiplexing gain can still be extracted by optimizing the antennas’ placement at the two arrays, by assuming a fixed transmit distance (i.e., the distance between the arrays’ centers) and fixed array orientations. Nonetheless, for mobile systems, where the location and orientation of the arrays is varying, it is very challenging to sustain a satisfactory performance in a LoS MIMO system deployment. Consequently, in the first part of this thesis, the goal is to investigate how the capacity of a mobile LoS MIMO can be optimized over a range of varying array locations and orientations. To this extent, different methods are proposed that focus on optimally placing the antennas at the transmit and receive arrays towards achieving a robust performance. Namely, regarding a uniform circular array (UCA)-based LoS MIMO system, a suitable array-of-subarrays is considered. By leveraging a simplified hybrid digitalanalog beamforming architecture, that is based on dynamically selecting the optimal subarray, a robust capacity is attained over a range of array locations and orientations. Furthermore, it is shown that a varying orientation (i.e., tilting) at the transmit or receive UCA can be also exploited towards extracting close to the maximum capacity at varying transmit distances for a 4 × 4 system. Then, two novel methods for jointly designing non-uniform transmit and receive arrays that comprise arrays of arbitrary geometry are presented. The proposed arrays outperform uniform and non-uniform arrays of the literature in terms of the minimum capacity over a range of varying array locations and orientations. Besides LoS MIMO communication systems, as the the shift of wireless communication systems to higher frequency bands results in an exceedingly high utilization of the spectrum by different technologies, the integration between radar and communication systems has emerged as a prominent research direction. In the second part of this thesis, integrated sensing and communication (ISAC) systems are investigated, where the focus is again on the optimal antenna placement towards improving the performance. In particular, joint antenna selection and covariance matrix optimization is considered for an array that is transmitting communication symbols to a user equipment and probing signals to several radar targets. It is shown that, by combining convex optimization with dynamic programming, the system can yield large communication capacity, while also properly focusing the transmit power towards the targets’ direction, which results in improved estimation performance. Finally, a distributed ISAC system is considered, where the transmit antennas are assumed to be widely separated. By employing a genetic algorithm, the problem of optimally selecting a subset of the active transmit antennas, towards maximizing the communication capacity and minimizing the Cram´er–Rao bound with respect to the target’s location, is efficiently solved.