ABot 4G/5G Network Test Solution

Test your 4G/5G network on demand.
Pay for the test, not the lab.

One software platform to emulate the network, generate the traffic, write the test cases and explain the result. It spins up when a test starts and releases the compute when it ends.

Schedule a demo Download brochure
Software only, runs on COTS Kubernetes · OpenShift · EKS · VMs On Red Hat Marketplace Vendor-neutral
1MUEs on a single ABot instance
1200-1500control plane CPS per instance
< 20 minsto bring the test bench online
0idle emulator nodes between test runs
On-demand Test BenchEphemeral emulators that scale from zero and back Traffic GeneratorVideo, VoIP, gaming, IoT, URLLC and more Scenario FactoryAI-generated test cases for every new release Core Digital TwinReplay field issues and prove the fix safely
The platform

Simulate. Automate. Analyze.

Any node in the topology can be emulated by ABot or be the real system under test. Swap in a real AMF, UPF or MME without rebuilding the lab.

Not state-machine bound

Run any call flow, including collisions, corner cases and negative paths.

Fault injection

Delays, drops, timeouts and procedure collisions on demand.

Security testing

Simulate DDoS and signalling storms, then verify the fix.

Artefacts that explain

Call flow diagrams, PCAPs and KPIs from every run.

On-demand Test Bench

A test bench that only exists while the test runs

The ABot framework stays on. The 4G/5G emulators don't. They launch when a test tranche starts, and the nodes are released as soon as it ends.

ALWAYS ON ABot framework · 16 vCPU / 32 GB · results and artefacts stay here ELASTIC NODES scale 0 → N → 0 0 nodes 0 nodes NF emulator + SUT pods running Functional pod 2 vCPU / 1 GB · Load pod 4 vCPU / 6 GB Trigger GUI or CI/CD < 20 mins nodes join cluster Execute and capture 3GPP procedures + traffic vs the SUT 20-60 min 5G Core sanity, small to medium pack ~15 min pods deleted, nodes drain Release compute back to 0

Typical timings for a 5G Core sanity tranche on AWS EKS with node groups set to a minimum and desired capacity of zero.

Cost follows the test, not the calendar

One day of emulator compute, illustrated.

Always-on lab
ABot on-demand, three test tranches
00:0012:0024:00

Results outlive the setup

PCAPs, logs, KPIs and reports stay on persistent storage after the nodes are gone. Each cycle is stored against the NF software version, so you can compare this release with the last one at a glance.

Per-release result matrix Call flow diagrams PCAPs, logs, KPIs

On-prem

Beside the network under test, on hardware you already own.

Any cloud

Public, private, hybrid or edge. Same images everywhere.

Any runtime

Kubernetes, OpenShift, EKS or VMs, matching how the SUT ships.

CI/CD triggered

Start a tranche from the ABot GUI or straight from your pipeline.

Read more: ABot Test-as-a-Service on Red Hat OpenShift · AI-driven zero-touch telco testing
Traffic Generator

Real-world traffic, end to end

ABot plays both ends of the conversation: the data client on the UE side and the application server on N6. Use real handsets over Wi-Fi or ABot-simulated UEs, against a real or ABot-emulated core.

ABot DATA CLIENTS Mobile / Web Video / IPTV VoIP / Gaming IoT / CIoT Real UE / CPEWi-Fi (FWA) gNodeBsystem under test N3 5G Corereal or ABot-emulated N6 ABot App Serverthe other end of every flow Throughput · latency · jitter · loss · call setup, per traffic type

Traffic types

HTTP / web browsing Social media Streaming video / OTT IPTV VoIP VoLTE / VoNR Gaming URLLC CIoT / NIDD RedCap Control and user plane load

Pick your topology

Real everything
UERANCore

ABot drives tests and traffic on your live lab with real phones. No simulators needed.

ABot UE/RAN vs your core
UERANCore

Load of 100+ UEs without buying handsets, plus a ready 5G sanity suite.

Hybrid
UERANCore

Mix ABot and real core NFs. Opens up conformance and E2E tests together.

ABot-emulatedYour real network
Read more: MEC Testing Needs · MWC LA 2021: network slicing with video traffic
Scenario Factory

New release in, test cases out

An agentic AI workflow reads your release notes and the matching 3GPP sections, finds what your current test repository misses, and designs the tests to close the gap.

Inputs
Release notes 3GPP TS sections Vendor specs Existing test repo Call models from field data
Agentic workflow
1DiffWhat changed since the last release
2Find gapsAgainst the current test repository
3Design3GPP-aware test case designs
4PreparePre-requisites to make them executable
Outputs
Test designs Executable ABot tests Or scripts for any tool Labelled experiment data

Ready-to-run 5G suite, 16 categories and growing

Interface managementRegistrationDeregistration Xn and N2 handoverSession managementUser profile management Connection managementNetwork slicing with dataSMS over NAS VoNR (IMS)5G NF conformanceCell broadcast (CBCF) NIDD and CIoTRedCapUE policy (URSP)DDoS and fix verification

Every category covers negative cases, corner cases and boundary conditions, not only the happy path. Also supports NSA, SA and 4G/5G interworking.

The same factory produces labelled training data for predictive AI models. See how on the ABot AI page →
4G/5G Core Digital Twin

Find it in the field. Replay it in a sandbox. Prove the fix.

ABot AI recovers the exact call model from lab, pre-prod or live data. ABot Test replays it on an on-demand twin built from your own blueprint. Nothing touches live subscribers.

Digital Twin closed loop: 1 Ingest, 2 Recover, 3 Build, 4 Replicate, 5 Validate

Issue replication

Reproduce a field fault safely, on demand.

RCA confirmation

Make it happen again under control.

Mitigation check

Try the patch on the twin first.

Upgrade impact

Run your real call model on a new release.

Capacity what-ifs

Signalling storms and peak load at scale.

Resilience

DDoS, collisions, drops, then verify the fix.

Why build the twin on ephemeral resources?

One twin per blueprint

Each customer or release gets a twin that matches its NFs, config and versions. No shared, drifting lab.

Break it freely

It is disposable. Inject any fault, then tear it down and start clean in minutes.

Zero idle cost

Compute exists only while you reproduce and validate. Artefacts are kept.

In production today. A hyperscaler's telco support team uses the ABot Digital Twin to replicate CSP issues and validate fixes before rollout. The result: lower MTTR and safer network changes.

Read more: Security Validation of 5G Core Functions / DDoS Attacks · Accelerating 5G standalone rollout with continuous testing
Who uses ABot

Built for every team that ships a network

Service providers

Release validation, capacity planning, network audits, and field-issue replication on a twin.

Network equipment vendors

Test any NF against emulated peers in CI. Auto-generate tests for every new release.

System integrators

Multi-vendor, multi-cloud integration and interop testing. Works with your existing test tools.

80%

less testing time through automation

70%

less manual QA for a tier-1 equipment vendor

60%

faster deployment cycles with continuous testing

60%+

lower cost of testing on COTS and cloud

Licensing

Three ways to license ABot. Pick the one that matches how often you test.

Fits the on-demand bench

Usage based

Pay per test run

You pay only when a test runs. Nothing to pay for while the bench sits idle.

  • No idle licences between test cycles
  • Cost scales with your test volume
Best forRelease-driven or occasional testing

Time based

Subscription for a fixed term

A predictable fee for the length of your contract.

  • Fixed, predictable budget
  • Test as often as you need during the term
Best forContinuous testing in CI pipelines

Perpetual hybrid flex

Own the licence, flex when needed

A permanent licence for your baseline, with extra capacity on demand for peaks.

  • Own your core test capacity
  • Add capacity for large campaigns
Best forLabs with a steady baseline and peak demand

Not sure which model fits? Talk to us about licensing

Insights & Blogs

Go deeper

BlogABot Test-as-a-Service on Red Hat OpenShiftCloud-native, on-demand 5G testing BlogThe future of AI-driven zero-touch telco testingAutomation that runs itself BlogAccelerating 5G standalone rolloutContinuous interop testing with Red Hat WhitepaperSecurity validation of 5G Core functionsDDoS, signalling storms and QoS overloads WhitepaperReadiness of network slicingSLA, scale and interop validation Blog5G Core deployment optimizationTest, measure and right-size your core

See the on-demand bench in action

From zero nodes to a full 5G test run, then back to zero.

Schedule a demo