Talaia · Reference · Network QoS APIs
Route prediction and network QoS APIs
ETSI GS MEC 030 [1] predicts network quality along a vehicle route. Network QoS APIs such as CAMARA [2] work over an area, from the operator. This sheet is the boundary between them, and the place they meet.
Two ways to predict network quality
A vehicle application can ask two different systems what the network will do. ETSI GS MEC 030 [1] answers for a route, from a MEC application at the edge. CAMARA [2] answers for an area, from the operator network. Both predict. They divide the problem by geometry, by vantage point, and by the moment they are useful.
CAMARA is an open-source suite of operator network APIs, hosted by the Linux Foundation and backed by the GSMA Open Gateway programme. Four of its APIs touch this ground. Quality on Demand requests priority for a flow, QoS Profiles names the available service levels, and Connectivity Insights checks whether a session's requirements can be met; these three are stable, in a CAMARA meta-release. Predictive Connectivity Data, which forecasts connectivity for an area over a future interval, sits at the sandbox stage, CAMARA's earliest maturity level.
The boundary, on six axes
The clearest way to read the two is side by side. The comparison holds the MEC 030 predicted-QoS resource (§6.2.6) against the CAMARA Predictive Connectivity Data API, the point where both forecast network quality. One caveat on maturity: MEC 030 is a published normative specification, while that CAMARA API is at the sandbox stage. Its stable QoS APIs, Quality on Demand and Connectivity Insights among them, sit in a meta-release.
| Axis | ETSI MEC 030 §6.2.6 | CAMARA Predictive Connectivity Data |
|---|---|---|
| Geometry | A route. Ordered waypoints, at least two, origin and destination | An area. A polygon split into geohash cells, with an optional vertical layer |
| Output | Quantitative KPIs per stream, each with a value and a confidence | Qualitative class per cell: Good, Marginal, No, or No Data |
| Vantage point | The MEC platform, consumed by an application co-located at the edge over Mp1 | Northbound from the operator, consumed by an app anywhere |
| Moment | During the journey, to adapt | Ahead of it, to plan (its own material frames it as decision support) |
| Domain | V2X-specific. Sits alongside PC5/MBMS provisioning and V2X message distribution | Domain-agnostic. Gaming, streaming, drones, fleets |
| Governance | ETSI ISG MEC, a normative Group Specification | Linux Foundation and GSMA, open source on meta-releases |
route of ordered waypoints and answers with quantitative KPIs. CAMARA predicts over an area of geohash cells and answers with a class per cell.How they work together
The two layers stack. A vehicle application at the edge reads the MEC 030 prediction for the route ahead and decides where on that route it needs a guaranteed level of service. For those stretches it calls the operator through CAMARA Quality on Demand and requests the profile it needs. One layer says where quality matters along the journey, the other secures it from the network.
Positioning Talaia as a replacement for CAMARA would be inaccurate. By October 2025 CAMARA reported 60 published APIs, more than 1,300 contributors and 476 member organisations, carried by the GSMA and dozens of operators. The MEC 030 interface answers a question those APIs leave open, the quality along a specific route from inside the edge, and it hands off to CAMARA when an application needs the network to act.