An Adaptive Multiobjective Genetic Algorithm with Multi-Strategy Fusion for Resource Allocation in Elastic Multi-Core Fiber Networks

Abstract

Core switching on different links in optical networks enables network operators to allocate network resources more flexibly, so as to reduce the network request blocking ratio under limited resources. Facing a differentiated network environment and diversified user demands, network operators need to optimize multiple objectives that are independent and diversionary of each other, and to provide multiple resource allocation schemes whose objective values do not dominate each other. For the static routing, spectrum, and core assignment (RSCA) problem in elastic optical networks with multi-core fiber (MCF-EONs), there is no literature that simultaneously considers core switching and multiobjective optimization algorithms. This paper improves the existing models and algorithms to adapt to the RSCA problem. In this paper, the RSCA problem is formulated as an integer linear programming model to minimize both network request blocking and crosstalk ratios simultaneously by considering core switching and inter-core crosstalk. To solve the model efficiently, we, therefore, design a joint routing and core coding scheme supporting core switching and propose a multiobjective evolutionary algorithm based on decomposition with adaptation and multi-strategy fusion (MOEA/D-AMSF), which integrates the new mechanisms of hybrid initial population generation, adaptive crossover, and double-layer and multi-point mutation in different iteration stages. These new mechanisms accelerate algorithm convergence and enhance solution diversity. Simulation results show that the proposed algorithm can obtain more dominated and diverse solutions compared with the existing multiobjective algorithm without considering core switching.

Interpreting multi-objective reinforcement learning for routing and wavelength assignment in optical networks

“…( 1 8 ) 300 (19) 400 (20) 4 0 0 (2 1 ) 400 (22) 4 0 0 (2 3 ) v j ∈ V (i ̸ = j); the uniform traffic matrix T unif is defined as [1]…”

“…To the best of our knowledge, this is the first study to tackle this challenging problem, which involves optimizing network throughput, cost, and resilience simultaneously. Traditional meta-heuristic [21][22][23] and exact [24] multi-objective optimization approaches are not practical in this setting due to their high running times, which make them impractical for online servicing of new traffic requests. In contrast, our approach utilizes RL to rapidly provide optimal RWA solutions, making it a practical and efficient solution for this problem.…”

Analytical Modeling of Groups of Links in Elastic Optical Networks

“…The authors also presented a new method of determining a parameter that takes into account the specific structure of the link group, which affects the way new calls are accepted into the system. The new model proposed by the authors makes it possible to determine the occupancy distribution in a group of parallel links between two neighboring EON network nodes or in links using multi-core fibers [34]- [36]. Therefore, the proposed model is applicable to systems in which we increase the data transfer rate by paralleling transmission.…”