Orchestrating tomorrow

We take traditionally difficult-to-work-with protocols and make them easy to use.

Counter loading...

Subscriber Management (AAA)

A RESTful JSON API integrates seamlessly into existing OSS/BSS systems, offering a scalable solution without concerns about storage or redundancy. Designed for communication service providers, such as ISPs, it enables them to concentrate on their core business rather than developing and maintaining complex AAA systems.

Interstellio IO
NebularStack Login

Platform Overview

Interstellio provides a fully managed, carrier-grade Subscriber Management platform featuring a 100% in-house RADIUS implementation.

Unlike most AAA providers, we do not rely on third-party vendors or open-source projects such as FreeRADIUS. Most AAA solutions depend on:

  • Community-driven open-source projects
  • Third-party support contracts
  • External update cycles

This creates dependency risks that directly affect SLA guarantees and operational certainty.

Owning the entire RADIUS protocol stack from RFC 2865 onward allows us to deliver:

Full protocol-level control
Deterministic performance
Independent security patching and updates
Guaranteed roadmap alignment
End-to-end accountability
There is no upstream vendor. We are the vendor.

NebularStack Dashboard

NebularStack

Our AAA service is built on NebularStack, a distributed SaaS platform with an API-first, cloud-native architecture that simplifies service deployment and removes infrastructure constraints.

The name reflects its philosophy:

Nebulaexpansion, growth, transformation
Stacklayered technologies

Architecture

Architecture Highlights

NebularStack is built to run distributed services such as our Subscriber Management (AAA).

Horizontal scalability
Vertical scalability
Distributed processing
Built-in redundancy
API-driven orchestration
Integrated telemetry systems

Multi-Tenant Architecture

Multi-Tenant Architecture & Role-Based Access Control

Interstellio is designed for multi-tenant environments, enabling operators to manage multiple business units, partners, or customer segments within a single platform while maintaining strict isolation and control.

Tenant Isolation: Each tenant accesses only its own subscribers, policies, sessions, and telemetry.
RBAC: Fine-grained permissions per tenant or function.
Centralised Oversight: Global visibility while tenants remain securely scoped.
Auditability: All actions logged for traceability and compliance.
Operational Flexibility: Wholesale, enterprise, or commercial segmentation without duplicating infrastructure.

Eliminating AAA Bottlenecks

Eliminating AAA Bottlenecks

Traditional RADIUS deployments often suffer from backend limitations that degrade user experience and operational efficiency:

Delayed AuthenticationSlow login responses frustrate subscribers
Timeout FailuresAccess-Requests and Accounting-Requests fail under load
Scalability ConstraintsPeak usage, reconnect storms, and subscriber state surges overwhelm monolithic systems
Operational Blind SpotsFragmented subscriber logs hinder troubleshooting
Interstellio removes these bottlenecks by decoupling authentication from monolithic systems and distributing processing across resilient infrastructure.

Core Architectural Advantages

Core Architectural Advantages

Fully redundant, distributed architecture
Transparent horizontal scalability
No physical hardware required
No single point of failure
24/7 NOC monitoring and Critical (P1) support
Data replicated at least three times, with separate DR archival
Real-time log visibility without debug activation
Full API access to all platform capabilities
Scheduled and custom report generation
Direct integration support via tickets, Teams, or Zoom

Virtual AAA Instances & Scale

Virtual AAA Instances & Scale

A Virtual is a fully distributed logical instance containing subscribers, policies, session data, and telemetry. This means the Virtual represents your subscribers everywhere it is deployed, rather than being limited to a single location.

Extended to any POP: One Virtual spans multiple Points of Presence with fully consistent subscribers, policies, and telemetry.
No single point of failure: State replicated across all nodes, clusters, and DR sites.
Operational & commercial flexibility: Multiple Virtuals coexist for wholesale, enterprise, or acquired networks.
Centralised management: Provisioning, session control, policies, reporting, and telemetry via a unified interface.

Secure On-Premise Proxy Option

For additional control or network isolation, we provide a lightweight on-premises proxy:

Runs on your own server or VM
Communicates with NAS/BNG devices using standard RADIUS
Establishes TLS-encrypted communication to our distributed platform
Uses a proprietary transport protocol for enhanced control
Requires minimal CPU and memory
This enables local RADIUS endpoints while leveraging our distributed intelligence.

Why Virtual Instances Matter?

Subscribers are always globally consistent, no matter where the service is delivered
Operational complexity is minimised. A single logical entity manages subscribers, sessions, and policies across multiple POPs
Commercial segmentation is simple without sacrificing control or consistency
Infrastructure resilience is built in. No isolated failure can impact subscriber sessions
In essence, a virtual is your subscribers, policies, and sessions. Everywhere, consistent, resilient, and centrally managed, giving ISPs freedom.

Subscriber State & Lifecycle Management

Subscriber State & Lifecycle Management

Our platform maintains a persistent, distributed subscriber state, including:

Session state tracking
IP allocation history
CDR records
Time-series telemetry (bandwidth usage)
Product and service associations
Policy enforcement status
State lives everywhere across all parts of the platform. Inactive subscribers continue to consume system resources and are billed separately to ensure complete tracking and replication.

Quota (Capped Subscribers)

Quota (Capped Subscribers)

A credit-based model manages capped subscriber accounts by allocating individual data allowances. Multiple credits can operate simultaneously on a single account, automatically renewing or expiring based on time-driven rules. Features like data rollover ensure unused data remains accessible until expiry. All credit usage is recorded in the telemetry backend for complete auditability, and credits can be managed via Credit Profiles in the UI or added through the API for external integrations.


Data Retention & Subscriber Removal

When a subscriber is purged:

Usage history, CDRs, and IP logs remain retained according to Lite (12 months) or Carrier (5 years) policies
Historical data incurs no additional cost
Automated inactivity policies, suspension, and purge rules can be applied and customised

Advanced Reporting

By default, two monthly reports are provided containing:

Active subscriber counts
Inactive subscriber counts
Data usage summaries
Platform metrics
Additional capabilities:
Custom report design and automated distribution
Raw data exports for investigations
Market research analytics (e.g., broadband usage trends)

Detailed RADIUS Processing

Detailed RADIUS Processing

We provide near real-time insights:

Per-BNG/NAS RADIUS packet counts (access-requests, access-accepts, etc.)
Request volumes and failure analysis
Available globally and per NAS/BNG, providing operational visibility rarely found natively in AAA platforms.

Just as launch vehicles depend on real-time telemetry to monitor propulsion performance, structural loads, guidance data, and environmental conditions during ascent so engineers can analyse anomalies and improve future missions, we believe continuous, high-fidelity telemetry is vital to our platform: measure everything, learn from every signal, and keep improving.


Detailed Data Usage & Bandwidth Graphing

Detailed Data Usage & Bandwidth Graphing

True 5-year data retention for bandwidth usage, CDRs, and IP logs
15-minute granularity preserved across the retention window
Compare FTTH vs FTTB usage, per-product, or per-service trends
Track active sessions for capacity planning and commercial insight

Full-Text Searchable RADIUS Logs

Full-Text Searchable RADIUS Logs

Every RADIUS transaction is captured and instantly searchable:

Full attribute capture (request & reply)
Real-time indexing
Full-text search across all log attributes
Immediate troubleshooting without debug or archived retrieval

API-First Integration Model

Documented RESTful JSON API
Complete access to subscriber provisioning, policy management, session management, reporting, telemetry, and administration
Integration support included

Intelligent Session Management

Unlike pure RADIUS solutions, our platform actively manages sessions:

Subscribers can have data caps, expiration times, and policy rules defined
Upon expiration, sessions are automatically disconnected
PUT or PATCH API calls automatically trigger re-authorisation, session disconnects, or policy updates
This removes the need for external orchestration, particularly for prepaid environments.

Under the Hood: Infrastructure

Delivering carrier-grade AAA at scale requires far more than RADIUS:

Redundant compute nodes
Dedicated backup servers
Multi-node SQL clusters
Distributed time-series database clusters
Distributed in-memory caching
Distributed scheduling
Distributed queueing both sync and async
Object storage clusters
Elastic indexed search
Layered load balancing
Distributed coordination (e.g., Apache ZooKeeper)

In-House Sharding & Partitioning

Custom sharding and partitioning logic
Predictable scaling of subscriber state, logs, and telemetry
No re-architecting required as subscribers grow
Engineering Distributed Systems

Designing large-scale distributed systems requires careful consideration of interdependent components and data flows. Latency, throughput, asynchronous processes, and resource contention interact in complex ways, much like flows in fluid or queuing networks studied in physics and operations research. By applying these principles, modelling pipelines, predicting bottlenecks, and optimising resource distribution, we ensure that our platform scales efficiently, performs predictably under stress, and remains stable even under extreme conditions.

AAA Service OptionsCarrierLite
Telemetry Granularity15 MinutesDepends on Retention
15 Minute Aggregation Data Retention5 years60 days
Hourly Data Retention6 months
6 Hour Aggregation Data Retention5 years
CDR Records12 months
IP Records
Bandwidth & Data GraphsHourly, Daily, Weekly, Monthly, Yearly
Multi-TimezoneFully supported
Max Sessions Per Subscriber10
Unlimited Session SubscribersIncluded (No CDR or IP Logs)
Webhook CallbackIncluded
Credits (Data Usage Caps)
Fair Usage Policies
Subscriber Aliases
Time-based bandwidth boost
RADIUS clients/nas/bng100025
Custom Scheduled Report Run1 IncludedAdditional Charge
Multi-TenancyIncluded 1 per SubscriberNot included
Multiple Platform Logins250 + 1 per Subscriber25
Active/Inactive Session Stats5 years
Multi-Vendor SupportIncluded
Radius Statistics
Live Sessions
Per product bandwidth usage
Per service bandwidth usage
Per radius client bandwidth usage
Extensive Subscriber Logs
Audit Logs
Password Auth (CHAP / PAP)
MAC Authentication
PPPOE / IPOE Support
Hotspot Support
RESTful JSON API
24/7 Session monitoring + notifications
Support24/7/365
P1 SLA Response time1 hour

Why Interstellio

24/7 Support and Monitoring
Fully owned AAA protocol implementation
Distributed, cloud-native infrastructure
Carrier-grade redundancy
True five-year granular telemetry retention
Intelligent subscriber lifecycle control
Operational simplicity without sacrificing control

We remove the infrastructure burden of AAA, allowing ISPs to focus on network growth, service innovation, and customer experience.

Book a Meeting