LAMBO: Large AI Model Empowered Edge Intelligence

TL;DR

LAMBO introduces a large AI model-based framework for edge intelligence that employs novel architectures and learning methods to address heterogeneous constraints, improve decision-making, and adapt to dynamic environments.

Contribution

The paper presents a new LAMBO framework with an asymmetric encoder-decoder architecture, actor-critic pre-training, and active learning for edge offloading tasks, enhancing flexibility and performance.

Findings

LAMBO outperforms traditional offloading methods in simulations.

The AED model achieves better global perception and decision-making.

Active learning improves adaptation in dynamic environments.

Abstract

Next-generation edge intelligence is anticipated to benefit various applications via offloading techniques. However, traditional offloading architectures face several issues, including heterogeneous constraints, partial perception, uncertain generalization, and lack of tractability. In this paper, we propose a Large AI Model-Based Offloading (LAMBO) framework with over one billion parameters for solving these problems. We first use input embedding (IE) to achieve normalized feature representation with heterogeneous constraints and task prompts. Then, we introduce a novel asymmetric encoder-decoder (AED) as the decision-making model, which is an improved transformer architecture consisting of a deep encoder and a shallow decoder for global perception and decision. Next, actor-critic learning (ACL) is used to pre-train the AED for different optimization tasks under corresponding prompts, enhancing the AED's generalization in multi-task scenarios. Finally, we propose an active learning from expert feedback (ALEF) method to fine-tune the decoder of the AED for tracking changes in dynamic environments. Our simulation results validate the advantages of the proposed LAMBO framework.