RAN EVIDENCE LAB · TAKWA SEBAI

Evidence that survives the handoff.

I publish the evidence boundary, not a polished answer built after the fact.

Every entry is labelled before you read it: real case, controlled campaign or method note.

Observed case: Built from preserved records. Sensitive identifiers are removed, not replaced with invented facts.
Controlled campaign: Measured against a declared contract. A completed run can still be rejected.
Method note: A standards-based investigation sequence. Synthetic teaching data stays explicitly synthetic.

CASE FILES 001–002

Two field decisions. Two conclusions kept deliberately narrow.

CASE 001 REAL · ANONYMISED RECONSTRUCTION

Operational decision: did the requested LTE target become the active serving cell, and could the restriction be rolled back safely?

Observed: Control responses, before-and-after radio samples, registration checks and contemporaneous engineering notes.
Decisive proof: The target PCI became the active serving PCI during the controlled refresh.
Not claimed: No internal modem root cause and no portability claim across modem families.

CASE 002 REAL · CONTROLLED CAMPAIGN

Operational decision: could BBR and CUBIC be compared when the 5G NSA bearer state changed around the scored transfers?

Observed: Five BBR and five CUBIC uploads to one controlled endpoint, with aligned application, server and radio evidence.
Decisive proof: Only one run per branch met the declared radio validity contract.
Not claimed: No algorithm winner. The surviving sample was below the pre-declared minimum.

METHOD 001 STANDARDS-BASED · SYNTHETIC EXERCISE DECLARED

The note separates EN-DC Secondary Node procedures from SA preparation, command delivery, target access, path switch and service outcome.

Boundary: the failure sequence is a teaching exercise. It is not presented as an operator incident or a Takwa Sebai field case.

USE THE METHOD

The free 111-page field guide connects the physical site, active configuration, decoded protocol and performance evidence.