Machine Learning and the Secure Access Service Edge

The Secure Access Service Edge (SASE) framework integrates network and security functions in a cloud native architecture. The paper reviews the SASE framework to examine the applicability and necessity for the inclusion of Machine learning (ML). The feasibility of ML at the edge, and its measurement and performance aspects, are reviewed in considering the feasibility of ML functions in the SASE framework. A proof-of-concept implementation is discussed to provide concrete examples of ML based services related to the service edge including monitoring, predictive alerting, and service disruption avoidance. The applicability and necessity are ultimately driven by the enterprise's objectives for its infrastructure. Enterprises adoption of ML services in a SASE framework may be driven by considering whether the radical network decentralization of recent years is a transient event.