5G Core & IMS Training

Six build-and-break courses that take an engineer from running a 5G core, to operating IMS voice, to modelling the whole network with machine learning — all on one low-cost cloud lab.

6Courses
615Runnable cells
~30Guided hours
3Learning tracks

The program is built on the open-source 5G stack — Open5GS (5G Standalone core), UERANSIM (simulated RAN), and a Kamailio + PyHSS IMS subsystem — running together on a single cloud VM. Every procedure is proven with real packet evidence: the habit trained throughout is never claim it works without the signalling to prove it.

Setup 5G Core track IMS / Voice track Data & ML track
00

Lab Setup

Prerequisite

Build the whole lab in your own cloud project — a VM, a 5G Standalone core, NAT, the kernel GTP module (with the toolchain pitfall pre-solved), a simulated RAN, a test subscriber, and a three-line smoke test.

39 cells build-your-ownlow-cost cloud VM
Cloud VMOpen5GS installGTP / NATUERANSIMFirst attach
01

5G Core Operations in Practice

5G Core

Operate, monitor and troubleshoot a live 5G Standalone core: network functions, PDU sessions, registration, reading the dashboard, and localizing faults by reference point.

97 cells ~4 hhands-on
NF lifecyclePDU sessions5G-AKAN1/N2/N3/N4Fault localization

Requires 00 · basic Linux + networking

View course →
02

IMS Core & VoNR Procedures

IMS · Voice

Register and call over a live IMS core on top of 5G — a P/I/S-CSCF chain, an HSS, and a media relay. IMS-AKA registration, a VoNR call between two UEs, and every procedure proven with packet evidence.

112 cells ~4.5 hcore lab
IMS-AKA REGISTERVoNR callCx: UAR/MAR/SAR/LIRSIP + Diameter decodeRTP / MOSSMS over IMS

Requires 01

View course →
03

Advanced 5G Core Engineering

5G Core

Change the network, not just watch it: network slicing end-to-end, QoS differentiation with Linux traffic control, TS 33.501 security hardening, signalling forensics via the NF discovery API, and an AI-RCA on-ramp.

70 cells ~5 hengineering
Network slicingQoS / tc shapingTS 33.501 securitySUCI / ECIESSignalling forensicsKPI pipeline

Requires 01 · 02 recommended

View course →
04

Advanced IMS Scenarios & Service Engineering

IMS · Voice

The service layer above the base procedures: initial Filter Criteria and Application Servers, MMTEL (hold, diversion, barring, identity), transfer and conference, emergency calls, access continuity, interconnect, advanced media and charging. Half is offline reference call-flows — runs with no lab.

119 cells ~6 h20 call-flows
iFC / AS / ISC / ShMMTEL servicesREFER / conferenceEmergency (sos)T-ADS / SRVCCENUM / IBCFRf/Ro charging

Requires 02

View course →
05

Data Engineering, Analytics & ML

Data · ML

Turn the telemetry every earlier course produces into a data-science lifecycle: ingestion, a feature store, analytics, visualization, and machine learning — anomaly detection, fault classification, and forecasting. Ships two notebooks: a synthetic edition that runs anywhere, and a live-lab edition that runs on real telemetry.

178 cells 100 + 78 cells~7 h + 4 h
Ingestion & feature storeSessionizationTS 28.554 KPIsIsolation ForestDTW classificationForecastingMLOps

Requires 01 · 02/03 for vocabulary

View course →

The one idea the whole path is built on

Every voice fault forks into the pipe (5G core) versus the application (IMS). Course 01 owns the pipe, Course 02 owns the application, the advanced pair deepens each, and Course 05 teaches the machine to read both. You leave able to answer that fork first — with evidence, not guesswork.

Build 00 Operate 01 Call 02 Engineer 03·04 Model 05

What each course ships

A self-contained package, not just a notebook.

  • Notebook — the instrument you run, cell by cell
  • Guide — a narrative companion, module by module
  • Assessments — quizzes, a practice test and a capstone
  • Slides — concept decks for the theory
  • Resources — quick-reference cards and a glossary

Cell-safety discipline

Every cell is labelled so you always know what it does.

  • 🟢 read-only — inspects, never changes the lab
  • 🟠 changes state — starts a service, provisions a UE
  • 🔴 instructor-guarded — fault injection; never fires on Run All