Velocity-Dependent Doppler-Beam Correlation in Millimeter-Wave V2V Communications: Characterization and Implications

Abstract

In millimeter-wave V2V communications, traditional exhaustive codebook searches cannot meet the sub-10 ms latency constraints required for safety applications. Sensing-aided approaches have been widely proposed to reduce this overhead however, well-known methods rely on radar, LiDAR, or GPS modalities that require hardware external to the millimeter-wave receiver chain. The Doppler shift, by contrast, is inherently available from standard channel estimation at zero additional hardware cost. This paper presents a systematic empirical characterization of the statistical relationship between the relative Doppler shift and the optimal beam sector in 60 GHz vehicle-to-vehicle channels. The central finding is that this relationship is strongly velocity-dependent: the Pearson correlation coefficient transitions from negative at pedestrian speeds to strongly positive at highway speeds, with a practical threshold near 10 m/s. The transition is explained analytically through a geometric decomposition of competing scattering and motion-alignment contributions and bounded via Fano’s inequality. A Doppler-Guided Transition predictor is introduced that maps the observed Doppler shift to a distribution of candidate beam transitions, conditioned on the previous optimal beam. Fused with temporal persistence under strict chronological train/test separation, the predictor achieves a statistically significant improvement over persistence alone at a beam probing overhead below 4%, without requiring any auxiliary hardware modality. The characterization, analytic model, and predictor are validated on 112,172 real-world measurements from the DeepSense 6G dataset and cross-validated on 8,588 ray-tracing samples from Raymobtime s007.

More by Assel Mukasheva

• Performance Evaluation of WPA2 and WPA3 Security Protocols in Private Wi-Fi 6 Environments2026
• HYBRID GAT-AUTOENCODER-BILSTM-ATTENTION ARCHITECTURE FOR MULTI CLASS CYBERATTACK DETECTION IN HETEROGENEOUS IOT NETWORKSВестник Алматинского университета энергетики и связи · 2026
• DISTRIBUTED SYSTEM FOR DATA SHARING AND JOINT PROCESSING BETWEEN RESEARCH CENTERS AND FARMSВестник Алматинского университета энергетики и связи · 2026
• Comparative Study of Regression Models Using Machine Learning to Predict Real Stock PricesInternational Conference on Electronics, Engineering Physics and Earth Science · 2026