Optimized Distributed Unit (DU) Placement under Physical Constraints for 6G x-Haul

The disaggregation of radio access networks (RAN) into radio units (RUs), distributed units (DUs), and central units (CUs) enables flexible and virtualized deployments, but imposes stringent constraints on the underlying transport network. In converged metro-access optical networks, DU placement must be carefully designed to ensure feasible RU–DU–CU connectivity under joint strict latency and quality-of-transmission (QoT) constraints. In this work, we investigate DU placement as an optimization problem aimed at maximizing feasible RU–CU routes. A GNPy-based optical simulation framework is used to evaluate the feasibility of RU–DU–CU paths under coherent DP- 16QAM transmission and functional split 7.2 latency constraints while comparing exhaustive, greedy, and genetic algorithm (GA)- based strategies to quantify the trade-off between feasible routes optimization and computational complexity. Results show that full RU–CU coverage can be achieved with only two optimally placed DUs using GA and Greedy-Coverage in our considered scenario. Among the evaluated approaches, a coverage-driven greedy approach achieves optimal performance with much lower complexity than genetic and exhaustive approach. These findings show that a network-aware DU placement heuristic provides an efficient and scalable solution for practical disaggregated RAN deployments.