Imitation Learning-based Implicit Semantic-aware Communication Networks

Semantic-aware communication has shown promising potential in enhancing communication efficiency and improving quality-of-experience (QoE) of the users. Previous works focus primarily on the transmission and recovery of explicit messages observed by the source user. In this chapter, we introduce the concept of implicit semantic-aware communication (iSAC) in which the implicit semantic information, e.g., hidden relations and implicit meaning, can be interpreted and recovered by the destination user. We propose a novel generative imitation-based reasoning mechanism learning (G-RML) solution. G-RML allows the destination user to successfully infer the implicit meaning based on the received explicit semantics. We also provide algorithm and architecture for multi-user computing network in the chapter.