Collaborative Safe Bayesian Optimization

TL;DR

This paper introduces CoSBO, a novel safe collaborative Bayesian optimization algorithm tailored for mobile networks, enhancing online parameter tuning efficiency while ensuring safety constraints are met.

Contribution

It presents the first application of safe Bayesian optimization to mobile networks and develops CoSBO, which leverages multiple tasks and safety constraints for improved efficiency.

Findings

CoSBO outperforms SafeOpt-MC in early optimization stages.

The method enables safe, online network parameter tuning with few iterations.

Sample efficiency is significantly improved in mobile network scenarios.

Abstract

Mobile networks require safe optimization to adapt to changing conditions in traffic demand and signal transmission quality, in addition to improving service performance metrics. With the increasing complexity of emerging mobile networks, traditional parameter tuning methods become too conservative or complex to evaluate. For the first time, we apply safe Bayesian optimization to mobile networks. Moreover, we develop a new safe collaborative optimization algorithm called CoSBO, leveraging information from multiple optimization tasks in the network and considering multiple safety constraints. The resulting algorithm is capable of safely tuning the network parameter online with very few iterations. We demonstrate that the proposed method improves sample efficiency in the early stages of the optimization process by comparing it against the SafeOpt-MC algorithm in a mobile network scenario.