Joint Non-Line-of-Sight Predictive Beamforming and Power Allocation for ISAC-Assisted Vehicular Networks

In this paper, we propose a joint non-line-of-sight (NLoS) predictive beamforming and power allocation (JNPB-PA) scheme to enhance the power efficiency of the road side units in the integrated sensing and communication (ISAC)-assisted vehicular networks. This scheme decouple the spatial and amplitude components in power allocation by exploiting angular domain discretization of a novel modulation technique–spatially-spread orthogonal time frequency space (SS-OTFS). Specifically, we first develop an auxiliary target method to achieve predictive beamforming in NLoS scenarios, which initially determines the power allocation vector’s non-zero positions corresponding to discrete angles of the vehicles. Then, we further refine the power allocation by solving a multi-objective optimization problem (MOP) aimed at minimizing both the age of information (AoI) for communication and the Cramér-Rao bound (CRB) for sensing. A low-complexity algorithm based on the proposed reconstruction-contraction-constraint (RCC) approach is developed to solve the formulated MOP based on its inherent features. Simulation shows that our proposed JNPB-PA scheme can achieve higher power utilization rate, lower AoI, and lower CRB in comparison with benchmark schemes. Besides, RCC solves the formulated MOP more efficiently by avoiding iterative searching of traditional methods.