As wireless networks push into higher frequency bands like millimeter-wave (mmWave) and 60 GHz, they encounter a fundamental physical hurdle: severe path loss. High-frequency signals attenuate rapidly over distance and are easily blocked by obstacles. Intuitively, one might expect that overcoming this requires boosting transmitter power.
However, a powerful counter-intuitive mechanism emerges from the physics of electromagnetic waves: as carrier frequencies increase, wavelengths shrink. Because the physical size of an antenna element scales proportionally with wavelength, a fixed physical aperture can pack exponentially more antenna elements at higher frequencies. This phenomenon, known as frequency-dependent antenna packing density, fundamentally changes how we manage signal propagation and network interference.
Compensating for Path Loss via Packing Density
Why does packing more antennas solve the path loss problem? When multiple antenna elements are arranged in an array, their signals can be phased to constructively interfere in a specific direction, creating a highly focused beam. As shown by Yi et al. and Yi et al., the shorter wavelengths of higher frequencies allow for a massive expansion of the antenna array scale within a compact footprint.
This large-scale beamforming creates an enormous directional gain. Instead of radiating energy in all directions, the array focuses power into a narrow pencil beam. Yi et al. demonstrate that this directional gain scales up with the number of elements in a planar array, effectively neutralizing the severe propagation losses inherent to high-frequency bands.
Reshaping Interference through Spatial Filtering
In traditional lower-frequency networks, omnidirectional or wide-beam antennas cause signals to spill over large areas, creating widespread interference. In contrast, high-frequency directional arrays alter the geometry of network interference entirely.
Instead of modeling interference as overlapping circles, the authors of Zhang et al. and Gan et al. utilize a sectored antenna model to capture how directional beamforming restricts signal overlap to narrow angular sectors. This phenomenon introduces a powerful spatial filtering effect described by Yi et al..
The intuition behind why this works is purely geometric: as the scale of an antenna array expands, its communication beamwidth narrows. As highlighted by Yi et al., narrowing the beamwidth drastically reduces the mathematical probability that a transmitter's beam will physically align with an unintended receiver or an external interferer. By filtering out the surrounding spatial noise, dense directional networks can achieve superior coverage and area spectral efficiency compared to traditional omnidirectional systems.
Exploiting Local Interference Minima via Beam Refinement
The extreme spatial resolution of scaled antenna arrays unlocks an even more surprising capability: the ability to eliminate self-interference and near-field coupling through tiny angular adjustments. In full-duplex or multi-panel systems, a major challenge is that a device's own transmitter blinds its receiver.
Conventional wisdom suggests that complex, expensive hardware cancellation is required to isolate these paths. However, a series of empirical studies by Roberts et al. revealed that self-interference at 28 GHz behaves like localized "coupling clusters" rather than uniform spherical waves. Because the beams are so sharp, self-interference is incredibly sensitive to steering direction: shifting a beam by a single degree can alter interference levels by up to 30 dB.
This sensitivity creates localized low-interference pockets within a tiny spatial neighborhood. To exploit this, Roberts et al. proposed a methodology called neighborhood-based beam searching (or beam refinement optimization). By exploring minor angular deviations—typically just 1 to 2 degrees—away from the absolute peak signal path, the system can find a configuration that places the self-interference path into a deep null while keeping the primary communication link firmly within the high-gain main lobe of the beam. As demonstrated by Roberts et al., fine-grained angular searches dramatically increase the availability of high-isolation beam pairs without sacrificing primary signal gain.
Architectural Density over Transmitter Power
Ultimately, this shift toward high-frequency directional beamforming changes how entire wireless networks are architected. Because high-frequency propagation is highly localized and relies on line-of-sight paths, trying to cover a city using a few centralized, high-power macrocells is highly inefficient.
Instead, Hall et al. show that architectural density—deploying a high concentration of low-power distributed access points—is a far more sustainable and effective solution. By combining network densification with the spatial filtering of scaled antenna arrays, modern wireless networks can achieve full coverage and multi-gigabit speeds while minimizing the total radiated power required.
Key Foundational Work
Analysis of capacity improvement by directional antennas in wireless sensor networks (2012) — contributed 1 concept elements to this direction.
Radio-over-Fibre access for sustainable Digital Cities (2013) — contributed 1 concept elements to this direction.