Routing-Led Evolutionary Algorithm for Large-Scale Multi-Objective VNF Placement Problems

TL;DR

<3-5 sentence high-level summary> The paper tackles the large-scale VNF Placement Problem by introducing a routing-led, parallel multi-objective evolutionary algorithm that leverages a genotype-phenotype mapping and memory-efficient data structures. It combines a decomposition-based search, an improved initialization operator, and compression techniques for distance and forwarding tables to scale to tens of thousands of servers. The study shows substantial memory savings, faster convergence, and competitive solution quality across very large data centers (up to ~64k servers), with routing-table construction identified as a key scalability bottleneck. These innovations enable practical optimization of QoS and energy tradeoffs in realistically sized data center networks, paving the way for deployment in large-scale environments.

Abstract

Modern data centers contain thousands of servers making them major consumers of electricity. To minimize their environmental impact, it is critical that we use their resources efficiently. In this paper we study how to discover the optimal placement of virtual network functions in large scale data centers. We propose a novel parallel metaheuristic, fast heuristic objective functions of the QoS and new memory efficient data structures for large networks. We further identify a simple, fast heuristic that can produce competitive solutions to very large problem instances. Using these new concepts, we are able to find high quality solutions for data centres with up to 64,000 servers.

Figures (16)

• Figure 1: Dcell$_1$ with 4 port switches GuoWTSZL08
• Figure 2: Fat Tree with 4 port switches Al-FaresLV08
• Figure 3: Leaf-Spine with 6 port switches Cisco19
• Figure 5: A high level overview of our proposed algorithm.
• Figure 6: An example of the components of the solution representation.