3D beamforming is a promising technology for millimeter wave (mmWave) communications, due to the demand for higher capacity and the need to overcome strong atmospheric absorption at high mmWave frequencies. By forming narrow beams in both the azimuth and elevation directions, 3D beamforming allows efficient spatial division and leads to higher user capacity, better coverage, less interference, and better energy efficiency. The true promise of this technology, however, has not yet been realized due to immense challenges in both hardware and algorithms. Existing beamforming approaches rely on phased array antennas with bulky, power-hungry phase shifters and related amplifiers and size not suitable for practical handsets. True mmWave beamforming requires thin, lightweight antennas with the ability to produce sharp beams that are electronically steerable over a wide range of angles in three dimensions. For effective beamforming, there is an added difficulty in estimating the channel accurately at both the receiver and transmitter because of strong directionality of mmWave propagation, its high path loss and vulnerability to shadowing and blockage. Such channel conditions call for robust 3D beamforming algorithms that are tightly coupled to novel antenna designs in order to integrate them into efficient communication systems.
This proposal puts forth several key innovations spanning across signal processing, antenna design and integrated system levels. (1) The signal processing innovation lies in robust 3D beamforming algorithms for a planar antenna array, taking into account errors in channel estimation, particularly in angles of departure and arrival specific to mmWave propagation, and also variation in antenna beam formation caused by inherent non-idealities in hardware and process mismatch specific for the proposed metamaterial antenna. Both of these sources of error have not previously been considered. (2) The hardware innovation lies in the use of metamaterial-based resonant cavity transmit/receive antennas that are ultra-thin for electronic beam steering, achieved by embedding active components such as varactors in the metamaterial that are biased intelligently in software. The proposed antenna is a fundamental shift from the existing phase-array antennas with bulky mechanical moving parts or phase shifters and related amplifiers, thereby also saving significant power, while providing flexible narrow beam-width of wide steering angles. (3) These two innovative aspects are then integrated in a multi-level evaluation plan, including detailed testing and evaluation at component (antenna), system (transceiver pair), and network (cellular) levels. Characterized antenna beam profile is integrated in a network simulation that implements the proposed robust 3D beamforming with detailed mmWave channel model. Project results are expected to have direct impact on mmWave communications in meeting the demand for high data rate connectivity in cellular, vehicular, and IoT systems.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.