Deploying and executing large model inference on edge devices is challenging due to their limited computational power and memory resources. To address this challenge, we present a novel Split-and-Pipeline, a collaborative inference scheme that partitions a large model into multiple submodels and executes them across distributed edge devices in a pipelined manner. The scheme parallelizes data transfer across multiple CPU cores to avoid transmission bottlenecks. We build a real-world testbed using NVIDIA Jetson series edge devices to demonstrate the proposed scheme, achieving 1.2×–3.0× throughput improvement over state-of-the-art baselines.

By fully utilizing the caching space of nearby smart users, mobile user caching is a promising solution for efficient content delivery service. However, how to motivate mobile users to share their spare caching space and protect the privacy information during caching sharing is a challenging issue. In this paper, we consider a blockchain enabled mobile user caching on the basis of blockchain network, thus the security and privacy issues can be well addressed, and the contributions of each mobile user can be recorded in the blockchain. Specifically, the interactions between the content service provider and the mobile users are modelled as a Stackelberg game, and the equilibrium is analyzed. Simulation results demonstrate the effectiveness of the proposed scheme and it is shown that the optimal strategies of content service provider and caching users can be established with different cache cost and caching requirement.