Robust Fronthaul-Aware Link Adaptation for UAV Communication With Lossy Wireless Fronthaul

This paper addresses loss of packets in a wireless fronthaul link being interpreted as channel errors for the service link. This may occur if, for instance, User Equipment (UE) is served by a radio unit in an Unmanned Aerial Vehicle (UAV) via a wireless fronthaul link to a baseband unit (BU) on the ground. The fronthaul packet loss leads to a mismatch between the real and estimated channel state, which in turn results in degraded system performance. The paper proposes a novel strategy to mitigate fronthaul packet loss impact and regain network performance without requiring complex packet-to-PHY data mapping. The method identifies fronthaul packet losses by detecting missing inputs to the Inverse Fast Fourier Transform (IFFT) in an OFDM modulator and maps this information to a codeword loss metric, which is fed back and processed at the BU. The loss metric is compared against thresholds by the BU to differentiate between fronthaul and channel-related losses. Performance evaluation is conducted using a 5G New Radio (NR) downlink system with a UAV serving a UE under full-buffer traffic conditions. Simulation results demonstrate that the proposed method achieves up to 64.1% gain in throughput compared to conventional link adaptation schemes under various fronthaul packet loss rates and service link signal-to-noise ratios.