Multihop Routing for Maximum Semantic Spectral Efficiency

Semantic communications promise to spearhead the contextual data processing and routing in future networks. To this end, and to the best of our knowledge for the first time, this paper addresses the problem of finding the route with maximum spectral efficiency in multihopsemanticcommunication systems. Our formulation relies on the recently introduced notion of semantic spectral efficiency, and a closed-form expression thereof that relies on data regression and a generalized logistic function approximation. We present a polynomial-time algorithm that providesprovablyoptimal solutions to the problem. The proposed algorithm relies on recent advances in non-isotonic routing metrics optimization. Our numerical results further illustrate the performance of the proposed approach.