Toward Semantic Communication Protocols: A Probabilistic Logic Perspective
[IEEE JSAC: SI on Beyond Shannon Communication, Catalysts for 6G] Toward Semantic Communication Protocols: A Probabilistic Logic Perspective
Authors: Sejin Seo, Jihong Park, Seung-Woo Ko, Jinho Choi, Mehdi Bennis, Seong-Lyun Kim
Classical medium access control (MAC) protocols are interpretable, yet their task-agnostic control signaling messages (CMs) are ill-suited for emerging mission-critical applications. By contrast, neural network (NN) based protocol models (NPMs) learn to generate task-specific CMs, but their rationale and impact lack interpretability. To fill this void, in this article we propose, for the first time, a semantic protocol model (SPM) constructed by transforming an NPM into an interpretable symbolic graph written in the probabilistic logic programming language (ProbLog). This transformation is viable by extracting and merging common CMs and their connections, while treating the NPM as a CM generator. By extensive simulations, we corroborate that the SPM tightly approximates its original NPM while occupying only 0.02% memory. By leveraging its interpretability and memory-efficiency, we demonstrate several SPM-enabled applications such as SPM reconfiguration for collision-avoidance, as well as comparing different SPMs via semantic entropy calculation and storing multiple SPMs to cope with non-stationary environments.
1) SPM Construction: We propose a novel method to construct a ProbLog-based SPM from an NN-based NPM, which occupies only 0.02% of the NPM memory usage by extracting and merging semantically equivalent vocabularies.
2) SPM Reconfiguration for Collision Avoidance (CA) and fairness: Without re-training, we demonstrate that an SPM is reconfigurable for CA by identifying rules that cause collision and instantly manipulating their connections written in ProbLog. We demonstrate the extent to which reconfiguration improves the performance in a new environment, e.g. CA and fairness constrained.
3) Best SPM Selection via Semantic Entropy: We empirically show that minimizing the average semantic entropy of an SPM (i.e., mean uncertainty of the SPM operations) achieves the highest goodput or equivalently the highest reward, allowing one to select the best SPM in a stationary environment.
4) SPM Portfolio for Non-Stationary Environments: By exploiting the memory efficiency of SPMs, we propose an SPM portfolio storing a set of SPMs, each of which is the best SPM for a specific environment.
S. Seo, J. Park, S. -W. Ko, J. Choi, M. Bennis and S. -L. Kim, "Towards Semantic Communication Protocols: A Probabilistic Logic Perspective," IEEE Journal on Selected Areas in Communications, Vol. 41 Iss. 8, Aug. 2023.