IEEE Network
Adaptive Quantum Search for Service Path Optimization in Edge-Cloud and 6G Networks
Engin Zeydan, Madhusanka Liyanage
Abstract
Service path optimization in large-scale edge–cloud 6G networks is computationally intensive due to multiple constraints, including latency, bandwidth, and energy. This article presentsGrover Adaptive Search (GAS)-enhanced Dijkstra, a hybrid classical-quantum approach that embeds an adaptive Grover schedule in a Dijkstra-based service path computation. By progressively narrowing the candidate set at each search step, GAS-enhanced Dijkstra reduces the oracle query count beyond what fixed-iteration quantum search achieves, while preserving path optimality. We evaluate it against seven classical routing algorithms and a fixed-iteration Grover baseline over six topology families. Under an idealized marked-item sampling model of the Grover subroutine, on 50-node graphs at edge density 0.7, GAS-enhanced Dijkstra needs 55% fewer oracle queries than a fixed-iteration Grover baseline running the identical threshold-descent loop, and a 4.8-fold lower quantum oracle-query count than the number of edge relaxations performed by classical Dijkstra, while retaining those classical relaxations itself. Benefits are limited in small or sparse networks.