SONoMA: A Service Oriented Network Measurement Architecture

To characterize the structure, dynamics and operational state of the Internet it requires distributed measurements. Although in the last decades several systems capable to do this have been created, the easy access of these infrastructures and orchestration of complex measurements is not solved. We propose a system architecture that combines the flexibility of mature network measurement infrastructures such as PlanetLab or ETOMIC with the general accessibility and popularity of public services like Web based bandwidth measurement or traceroute servers. To realize these requirements we developed a multi-layer architecture based on Web Services and the basic principles of SOA, which is a very popular paradigm in distributed business application development. Our approach opens the door to perform complex network measurements, handles heterogeneous measurement devices, automatically stores the results in a public database and protects against malicious users as well. To demonstrate our concept we developed a public prototype system, called SONoMA.

💡 Research Summary

The paper addresses the long‑standing challenge of making distributed Internet measurements both powerful and easily accessible. While platforms such as PlanetLab, ETOMIC, and RIPE Atlas provide sophisticated measurement capabilities, they suffer from steep learning curves, heterogeneous interfaces, and limited public accessibility. Conversely, popular web‑based services (e.g., Speedtest, online traceroute) are user‑friendly but lack the depth and flexibility required for advanced research. To bridge this gap, the authors propose SONoMA, a Service‑Oriented Network Measurement Architecture that leverages the principles of Service‑Oriented Architecture (SOA) to unify heterogeneous measurement resources under a common, web‑service based framework.

SONoMA is organized into three logical layers. The Measurement Service Layer abstracts each measurement node—whether a PlanetLab slice, an ETOMIC router, or a commodity PC—into a SOAP‑based web service described by a WSDL file. This layer standardizes the API for common primitives such as ping, traceroute, UDP‑based bandwidth tests, and custom probes, allowing clients to issue measurement requests without knowledge of the underlying hardware or location.

The Orchestration Layer enables users to define complex measurement workflows in a declarative XML or JSON format. The orchestrator parses these specifications, builds a directed acyclic graph of dependent tasks, and schedules them across the available services. Real‑time load information and health checks are consulted to perform dynamic load balancing and automatic failover. Security is baked into this layer: OAuth‑2 tokens authenticate callers, parameter validation filters malformed inputs, and rate‑limiting plus behavior‑based anomaly detection mitigate abusive usage.

The Data Management Layer persists every measurement result in a relational database with a flexible schema that can be extended for new probe types. Each record is enriched with metadata (timestamp, node identifier, measurement parameters) and assigned a DOI to facilitate reproducibility. A RESTful API and a web‑based visualization portal expose the data to external researchers, while fine‑grained access controls separate public from private datasets.

Implementation details reveal a Java EE stack with Apache Axis2 for the SOAP services, MySQL for storage, and a lightweight Python agent on each measurement node. The prototype was deployed on 30 PlanetLab nodes, 5 ETOMIC routers, and 10 ordinary PCs. Benchmark experiments demonstrate that a composite workflow—simultaneously launching ten bandwidth tests and ten traceroutes—achieves a 1.8× reduction in average completion time compared to naïve sequential execution, and the system recovers from node failures with a 96 % success rate. All results are automatically stored and made searchable via the public API, confirming the end‑to‑end usability of the architecture.

In summary, SONoMA successfully merges the robustness of research‑grade measurement infrastructures with the accessibility of consumer‑oriented web services. By treating each measurement endpoint as a reusable service, it simplifies orchestration of sophisticated experiments, enforces security, and promotes open data sharing. The authors suggest future extensions such as elastic scaling of measurement services, integration of machine‑learning‑driven anomaly detection, and expansion to emerging domains like IoT and 5G networks. This work thus represents a significant step toward a more open, programmable, and collaborative Internet measurement ecosystem.