Benchmarking Various ML Solutions in Complex Intent-Based Network Management Systems

Abstract

Intent-based networking (IBN) solutions to managing complex ICT systems have become one of the key enablers of intelligent and autonomous network management. As the number of machine learning (ML) techniques deployed in IBN increases, it becomes increasingly important to understand their expected performance. Whereas IBN concepts are generally specific to the use case envisioned, the underlying platforms are generally heterogenous, comprised of complex processing units, including CPU/GPU, CPU/FPGA and CPU/TPU combinations, which needs to be considered when running the ML techniques chosen. We focus on a case study of IBNs in the so-called ICT supply chain systems, where multiple ICT artifacts are integrated in one system based on heterogeneous hardware platforms. Here, we are interested in the problem of benchmarking the computational performance of ML technique defined by the intents. Our benchmarking method is based on collaborative filtering techniques, relying on ML-based methods like Singular Value Decomposition and Stochastic Gradient Descent, assuming initial lack of explicit knowledge about the expected number of operations, framework, or the device processing characteristics. We show that it is possible to engineer a practical IBN system with various ML techniques with an accurate estimated performance based on data from a few benchmarks only.

Toward Intent-Based Network Automation for Smart Environments: A Healthcare 4.0 Use Case

“…However, with the IBN problem, intents are presented with simple and short sentences; meanwhile, the LDA approach is generally unsuitable for document topic modeling if the dataset is too small, documents' lengths are too short, or there are too many topics within the dataset. Bensalem et al [22]- [23] presented intent-based networking in information and communications technology (ICT) supply chain networks. They extracted information from unstructured intents and stored them in JSON files inferred with an ML recommender that estimates the computational performance of different ML-based techniques (Singular Value Decomposition and Stochastic Gradient Descent).…”

“…Similarly, in [24], the authors proposed intent-based solutions for automatic network orchestration through the application of Graph Neural Networks (GNN) and Long Short-Term Memory (LSTM) techniques. However, in [24] and [22], the authors did not provide extensive analysis of the deployed intent datasets, IBN-engineering systems, and their corresponding performances. In [13], Ouyang et al reviewed the keyenabling technologies of the IBN while focusing on intent refinement schemes differentiated according to target users, input methods, and refinement approaches.…”