Contextual Reasoning-based Semantics-Native Communication

This chapter conducts a thorough investigation into semantics-native communication (SNC), a methodology crafted to optimize communication systems by augmenting efficiency and task effectiveness. It is particularly tailored for semantic communication scenarios where the sender and receiver independently develop their encoder and decoder, respectively. SNC leverages contextual reasoning (CR), especially in situations where communicators share a semantic framework or context. Drawing inspiration from linguistic and psychological methodologies inherent in human communication, CR seamlessly integrates into communication systems to implement SNC. The chapter commences with an overview of the SNC model, followed by a comprehensive examination of various CR methods applicable to SNC. These methods encompass single-sided CR, double-sided CR, inverse CR, and information bottleneck CR. A key emphasis is placed on the theoretical background and mathematical analysis of SNC throughout this exploration. The ensuing insights promise to shed light on the potential and possibilities of effectively deploying SNC in practical communication systems.