X2X Channel Gain Map Estimation Based on Bidirectional Virtual Scatterer Model

Abstract

This letter studies the construction of any-to-any (X2X) channel gain maps (CGMs), which predict the large-scale channel power gain for arbitrary transmitter (TX) and receiver (RX) locations in a region of interest. Unlike fixed-TX CGM methods that require intensive measurements for varying TXs, we propose a bidirectional virtual-scatterer-based X2X CGM construction framework that models dominant environmental objects as shared power-transfer virtual scatterers. Specifically, a priori coarse object geometry is used to determine the visibility and propagation loss of each multipath component, while sparse channel gain measurements are used to estimate bidirectional scatterer responses via a convex quadratic program. Simulation results show that the proposed framework accurately reconstructs X2X CGMs for different TX locations, substantially reduces the measurement requirement over fixed-TX baselines, and further benefits from a geometry-guided spatial sampling strategy.