The present deliverable “Pencil beamforming and device tracking algorithms and performance,” focuses on initial results for beamforming, effects of phase noise on beamforming performance, channel estimation, device discovery (finding out that a device is present), beam discovery (finding out the proper beamforming direction to use), and fast beam tracking (keeping track of proper beamforming direction when movement/rotation is present in at least one end) for realizing terabit/s wireless connectivity.
Section 2 presents a literature survey on beamforming, beam discovery and beam tracking, especially on mmWave band since many papers have been written for mmWave but only few for THz (lower THz band belongs also to mmWave band). Different beamforming architectures are presented, and their benefits and drawbacks are discussed. Beam discovery methods avoiding exhaustive full search are discussed. Beam tracking methods reducing the search space are also discussed.
In Section 3, hybrid beamforming with the array-of-subarrays architecture at THz is presented as our reference system. Transceiver model is presented, and two models are
presented for the phase noise. Correlation effects (that depend on the architecture) among different branches are modelled with a simple correlation coefficient. Phase shifter uncertainties are also discussed, since in addition to normal phase noise they can also have effect on the performance. I/Q imbalance and amplifier non-linearities are also modelled. The detailed simulations including these are left for D3.3, which is the version 2.0 of this deliverable due in month 18 of TERRANOVA project presenting more detailed performance evaluation/algorithms.
Here we focus on the effects of the number of antenna elements, beam direction, and phase noise and also on fast beam tracking. In Section 4, numerical results are presented for main lobe gain, main lobe pointing variation, and side lobe gain. Results are presented for both hybrid beamforming architecture and fully digital architecture. In Section 5, channel estimation methods that can assist in beamformer design and tracking are presented. Fast tracking method is presented, enabling more efficient tracking of users. Extensive numerical results are presented for fast tracking, including for both linear and crooked user movement. Numerical results are presented for device discovery when using a log likelihood detector.
The main outcomes of the deliverable are:
• A literature survey on beamforming, beam discovery and beam tracking is presented.
• The identification of reference beamforming method representing the SOTA relevant to TERRANOVA.
• Suitable models for phase noise, I/Q imbalance and amplifier non-linearities are presented.
• Extensive numerical results for beamforming and for fast device tracking. This deliverable forms the basis for the deliverable D3.3 that will later present the final algorithms and more detailed simulation results.
Download the PDF here: TERRANOVA_D3.1