This paper investigates a Terahertz (THz)-enabled mobile edge computing (MEC)-assisted virtual reality (VR) system using reconfigurable holographic surfaces (RHS) as transceiver for multi-user beamforming and holographic-pattern division multiple access (HDMA). We develop an end-to-end model for the 3D field-of-view (FoV) generation pipeline and optimize content prefetching, rendering offloading under memory and power constraints, and beamforming accommodating user movement by adjusting holographic pattern weights for beamshaping and feeds power allocation for excitation amplitude adjustment. For homogeneous FoVs, we derive closed-form policies for prefetching 2D or 3D FoVs or direct transmission of 3D FoVs. For heterogeneous FoVs, we exploit the timescale separation between prefetching/rendering and fast RHS beamforming, decomposing the optimization into a rendering-prefetching combinatorial optimization problem and a short-timescale beamforming convex optimization problem. Simulations show significant latency reductions under tight resource constraints.

翻译：本文研究了一种采用可重构全息表面作为收发器的太赫兹移动边缘计算辅助虚拟现实系统，该系统支持多用户波束成形与全息图案多址接入。我们建立了三维视场生成流程的端到端模型，并在内存与功率约束下，优化了内容预取、渲染卸载，以及通过调整全息图案权重进行波束赋形、馈电功率分配进行激励幅度调节以适应用户移动的波束成形。针对同质视场，我们推导了预取二维或三维视场、或直接传输三维视场的闭式策略。针对异质视场，我们利用预取/渲染与快速全息表面波束成形之间的时间尺度分离特性，将优化问题分解为渲染-预取组合优化问题与短时标波束成形凸优化问题。仿真结果表明，在严格资源约束下，该系统能显著降低传输延迟。