How do I set up a network test platform for interoperability testing across mixed 5G protocol stacks?

To set up a network test platform for interoperability testing across mixed 5G protocol stacks, design it as a modular, multi-vendor, multi-domain lab with clear separation between radio, core, transport, and test orchestration.

1) Define the interoperability scope

Start by listing exactly what you need to validate:

• Access network: gNB, CU/DU splits, RU interfaces, NSA/SA
• Core network: 5GC functions such as AMF, SMF, UPF, PCF, NRF, UDM, AUSF
• Protocol layers:
  • RF / PHY
  • MAC / RLC / PDCP
  • F1 / E1 / Xn / N2 / N3
  • SBA interfaces in the core
  • Transport and timing
• Feature combinations:
  • SA vs NSA
  • VoNR / VoLTE interworking
  • slicing
  • roaming
  • QoS flows
  • handover
  • dual connectivity
  • carrier aggregation

Define a matrix of:

• vendors
• software versions
• feature sets
• topologies
• expected outcomes

That matrix becomes your test plan.

2) Build a layered lab architecture

A practical lab usually has these layers:

A. RAN emulation or live RAN

Use one of:

• Real gNB/RU/CU/DU equipment
• RAN simulators/emulators
• UE emulators or device farms

If you want broad interoperability coverage, combine:

• real vendor RAN nodes
• UE test devices
• protocol analyzers
• signal/RF emulator or call box

B. 5G core test environment

Use:

• a live 5GC from one or more vendors
• a cloud-native core instance
• open-source core elements for baseline testing
• a traffic generator for user plane validation

Keep core functions modular so you can swap AMF/SMF/UPF/NRF implementations.

C. Transport and timing

Ensure:

• synchronous timing: GNSS, PTP, or SyncE as needed
• L2/L3 transport for N2/N3/F1/E1
• QoS-aware switching
• VLAN/VXLAN/MPLS if your production network uses them

Interoperability failures are often timing or transport issues, not just protocol issues.

D. Test control and automation

Add:

• a test orchestrator to configure setups and launch cases
• automation for UE attach, PDU session setup, handover, and teardown
• result collection and correlation
• configuration version control

Examples of orchestration capabilities:

• traffic profile loading
• topology switching
• KPI collection
• packet capture trigger
• automated pass/fail rules

3) Include the right test tools

Typical tooling stack:

• Protocol analyzers
  • NGAP, XnAP, F1AP, E1AP, NAS, GTP-U
• RF test equipment
  • call box, channel emulator, spectrum analyzer
• Packet capture
  • SPAN/TAP on N2/N3/F1/E1 links
• Traffic generators
  • uplink/downlink throughput, latency, jitter
• UE/device management
  • SIM profile control, logging, remote scripting
• Core observability
  • logs, traces, metrics, container monitoring

4) Design for mixed protocol stacks

For mixed-stack interoperability, support combinations such as:

• vendor A gNB with vendor B core
• vendor C CU/DU with vendor A RU
• open-source core with commercial RAN
• SA core with NSA access
• different release levels, e.g. 3GPP Rel-15/16/17

Use a compatibility matrix:

• interface version
• encoding/IE support
• optional procedures
• security modes
• timer defaults
• roaming and slice identifiers

This helps isolate whether failures are:

• spec mismatch
• implementation bug
• parameter mismatch
• transport issue
• timing issue

5) Create a repeatable test topology

A good basic lab topology looks like:

• UE / UE emulator
• RAN node(s)
• switch fabric / transport network
• 5GC functions
• test controller
• logging and capture server
• traffic server
• time synchronization source

Add redundancy where needed:

• isolated VLANs for management and signaling
• separate paths for control and user plane
• mirrored ports for capture
• out-of-band management

6) Test the key interoperability cases

Prioritize these scenarios:

Registration and attach

• initial registration
• authentication
• security mode control
• PDU session establishment

Mobility

• intra-gNB handover
• inter-gNB handover
• Xn-based and N2-based mobility
• path switch handling

QoS and slicing

• 5QI mapping
• ARP handling
• slice selection
• dedicated vs default bearers / QoS flows

User plane

• GTP-U forwarding
• throughput under load
• packet loss and reordering
• MTU fragmentation

Resilience

• node restart
• link failure
• timer expiry
• mobility during congestion
• recovery after deregistration

Security and identity

• SUPI/IMSI handling
• authentication vector exchange
• ciphering/integrity negotiation
• certificate trust for SBA interfaces

7) Automate the test lifecycle

Automate:

1. environment provisioning
2. node configuration
3. test execution
4. log collection
5. KPI extraction
6. report generation
7. regression comparison

Automation frameworks often include:

• Python-based orchestration
• CI/CD integration
• config templates
• artifact storage for logs and pcaps

Make every test case produce:

• test ID
• topology
• software versions
• parameter values
• timestamps
• pass/fail evidence

8) Measure the right KPIs

Track:

• registration success rate
• PDU session establishment time
• handover success rate
• throughput
• latency
• jitter
• packet loss
• reconnect time after failure
• signaling message success/failure
• CPU/memory utilization on core and RAN nodes

9) Control lab isolation and safety

For a real interoperability lab:

• isolate it from production networks
• use dedicated IP ranges and SIM profiles
• control RF leakage and spectrum use
• ensure lawful and licensed frequency usage
• restrict access and secrets
• back up configurations and certificates

10) Recommended implementation pattern

A strong setup is:

• Layer 1/2: real or emulated RF + timing
• Layer 3/control plane: packet capture and protocol analyzers
• Core: modular 5GC with multiple vendor interfaces
• User plane: traffic generator and reflector endpoints
• Automation: centralized test runner and CI pipeline
• Observability: logs, metrics, traces, and PCAP archive

11) Practical starting blueprint

If you want a minimal but effective starting platform:

• 1 UE emulator or test handset
• 1 commercial or open-source 5GC
• 1 RAN stack or gNB emulator
• 1 packet capture switch/TAP
• 1 traffic generator
• 1 time sync source
• 1 orchestration server
• logging and dashboarding tools

Then expand with:

• second vendor RAN/core
• real UE devices
• channel emulator
• multi-cell and roaming scenarios
• automation for regression testing

12) Best practice for mixed-stack interoperability

Treat interoperability as a matrix problem:

• interfaces
• vendors
• software versions
• feature flags
• topology
• timing
• load conditions

And keep a disciplined workflow:

• baseline one known-good stack first
• change only one variable at a time
• capture every signaling exchange
• compare against expected 3GPP procedures
• maintain a regression suite

If you want, I can also provide:

1. a reference lab architecture diagram
2. a bill of materials
3. a sample interoperability test matrix
4. a step-by-step build plan for a small-budget 5G lab