Reconfigurable Holographic Surface-Aided Distributed Edge Computing.

The recently advocated reconfigurable holographic surface (RHS) is anticipated to bring substantial improvement in communication throughput, especially for the multi-node scenarios. From this viewpoint, RHS is a natural fit to distributed edge computing, where the computation task is offloaded to multiple edge nodes for faster processing. Nonetheless, existing pioneering works on RHS mainly focus on its application in conventional wireless communications, and, to the best of our knowledge, its potential for elevating distributed edge computing performance remains unexplored. In this work, an RHS-aided distributed edge computing scheme is proposed, where the mobile device performs part of the task locally and exploits the RHS to distribute the rest to multiple edge nodes simultaneously. Considering both local computing latency and offloading latency, a latency minimization problem is formulated by jointly optimizing the task partitioning, the local computing frequency, the holographic beamforming of RHS, and the digital beamforming. To tackle this non-trivial non-convex problem, a joint communication and computation optimization algorithm is developed based on the generic block coordinate descent (BCD) framework. Particularly, the original problem is divided into three subproblems, and the task partitioning and local computing frequency, the holographic beamformer, and the digital beamformer are optimized efficiently and iteratively in each subproblem. In addition, simulations are conducted to validate the effectiveness of the proposed scheme.