KOMSENS-6G final report

Abstract

The ever-increasing demand for ultra-wide bandwidths in future 6G and sensing systems pushes communication and sensing applications into the D-band (110–170 GHz). At these frequencies the free-space path loss is severe and the available output power of individual MMICs is limited, which makes highly directive, beam-steerable antennas a prerequisite. Classical half-wavelength phased arrays become impractical at these frequencies because the element spacing shrinks dramatically, leading to extreme integration density and thermal-management problems. In the frame of KOMSENS-6G, IMST proposes an active dielectric-lens array as an alternative beam-forming architecture. Each 1 cm-diameter lens is fed by a focal-plane array of SIW (substrate-integrated waveguide) antennas, enabling fully electronic 2D beam steering without mechanical movement. The lens-array approach reduces the number of required active transceiver channels, relaxes inter-element spacing, and adds flexibility for adapting beamwidth and EIRP by selecting different illumination configurations. A modular unit cell (2 × 2 lenses with respectively 4 × 4 SIW feeds) can generate up to 16 independent beams or, alternatively, combine several lenses coherently for a high-EIRP single beam. Compared with a conventional half-wavelength patch array, the lens array needs roughly ten times fewer active elements to achieve the same EIRP (e.g., 4 × 5 patches versus a 2 × 2 lens array for 40 dB EIRP), simplifying PCB layout, reducing thermal load, and easing signal routing. The implementation uses SIW feeding antennas, aerosol-jet printed interconnects for low-loss MMIC-to-PCB transitions, and a 4-channel D-band transceiver MMIC (PA with 8 dBm P1dB, 16–22 dB gain). Measured S-parameters and far field patterns of single- and four-row SIW prototypes (with and without dielectric lenses) agree well with full-wave simulations, confirming ≈ 27 dBi gain and a ±25° field-of-view. A demonstrator comprising a 2 × 2 lens array, two front-end boards (each hosting two MMICs), and a dedicated mainboard for voltage regulation and SPI programming has been assembled. The targeted system specifications are 130–150 GHz operation, 27 dBi antenna gain, ±3° half-power beamwidth, 25° azimuth scan, 4 simultaneous TX/RX beams and ≤ 6 W TX power consumption. Functional tests verify correct on-chip LDO operation, SPI-based beam-state programming, and basic TX/RX functionality. Preliminary TX measurements at 144 GHz demonstrate proper carrier suppression and side-band behavior. Future work will focus on improving RX conversion gain, calibrating the vector-modulator phase shifters, and performing far-field measurements to validate multi-beam and high-EIRP operation. The active lens-array concept thus provides a compact, scalable solution for high-gain, electronically steerable D-band antenna systems, addressing the key challenges of integration density and power consumption in next-generation wireless and sensing applications.