Smart Core Network Slicing for IoT and MVNOs: Unlocking 5G Potential in 3G/4G Networks

Network slicing is hailed as one of the most transformative 5G innovations, yet many operators overlook its immediate applicability. Bart Salaets, Systems Engineering Director at F5 Networks, points out that the technology can already be leveraged to deliver tailored network experiences today.

In essence, network slicing lets an operator partition a single physical infrastructure into multiple logical networks—each engineered to meet distinct performance criteria such as throughput, latency, and priority.

This capability is crucial as 5G networks evolve to support a broad spectrum of verticals—from automotive and agriculture to healthcare, logistics, and beyond. Dynamic end‑to‑end slicing enables services to scale network resources in real time, matching traffic demand with precision.

For instance, high‑definition video streaming requires substantial bandwidth but can tolerate moderate latency, whereas autonomous vehicles demand ultra‑low latency to ensure safety. By allocating dedicated slices, operators can guarantee the right mix of resources for each scenario.

Traditional IP transport networks achieve limited slicing through VPN technologies. In contrast, 5G slicing extends deep into the radio domain, allowing resources to be provisioned from the radio access network all the way to the core. This end‑to‑end isolation is what enables, for example, a self‑driving car to be mapped to a slice that guarantees low‑latency bearers from the base station to the core network.

Network Slicing Today

Facing growing demands from the Internet of Things (IoT) and Mobile Virtual Network Operators (MVNOs), operators are turning to core‑level slicing within legacy 3G and 4G infrastructures—networks that lack radio‑level slicing support.

Current approaches rely on DNS, leveraging location and Access Point Name (APN) data to route subscriber sessions to Serving Gateways (SGW) and Packet Data Network Gateways (PGW). However, these methods have key limitations: they cannot allocate groups of devices to specific SGWs based on location alone, and creating a new APN for each device cohort is cumbersome and hampers isolation of diverse IoT use cases.

F5 Networks offers a practical solution: the GTP Session Director, which replaces conventional DNS routing with granular, policy‑driven decisions at the GTP layer.

Every GTP session initiated by an end‑user device is intercepted by the Session Director. Local policies—such as IMSI ranges, device types, or application identifiers—direct the session to a chosen SGW or PGW. For example, a smart‑metering deployment might allocate all devices with IMSIs within a specific range to a dedicated set of SGWs, ensuring predictable performance.

The same principle applies to MVNOs sharing an operator’s APN. By mapping an MVNO’s IMSI range, the Session Director routes its traffic to a dedicated PGW that exclusively handles MVNO sessions, thereby providing clear isolation without the need for additional APNs.

While the 3GPP has introduced pre‑5G slicing concepts such as DECOR and eDECOR, the GTP Session Director stands out by delivering slice‑like granularity without requiring new attributes or core‑network upgrades. Operators can thus harness their existing SGW and PGW assets to support new use cases and maintain flexibility.

Author: Bart Salaets, Systems Engineering Director, F5 Networks