This technology is a wireless communication control device that selects k nodes to participate in analog cooperative beamforming and transmits control commands. It uses a weight matrix trained on the 3D coordinates (x, y, z) of a receiver to manage n sensor nodes distributed in a 3D space that includes height information. The nodes are assumed to be arranged in a 3D Gaussian distribution within a spherical space of a defined radius.
Individual node transmissions in wireless sensor networks are vulnerable to eavesdropping and suffer from low receive diversity gain. In environments where nodes are distributed vertically—such as high-rise buildings, underground facilities, or aerial vehicles—selecting cooperative nodes based solely on planar coordinates fails to account for vertical distance and channel differences, making optimal node selection difficult.
This technology constructs a 1×3 matrix of the receiver's 3D coordinates and applies a weight matrix and softmax function to calculate probabilities for each node. By machine-learning the selection of nodes with the maximum SNR, it performs three-dimensional node selection that accounts for height differences. Based on a Gaussian distribution model within a spherical space, it effectively handles real-world volumetric layouts where nodes are dense at the center and sparse toward the periphery. It is applicable to multi-story IoT, drone swarm communications, and underwater/underground sensor networks, ensuring secure cooperative transmission paths even in environments with significant vertical or altitude variations.
This invention was developed with research support from the Ministry of Science and ICT for magnetic resonance-based beamforming technology for selective wireless power transmission systems.
N/A
