Stratosphere-First Telecoms #2 - Stratospheric RAN: Designing a gNB-in-the-Sky

Executive Brief

• Treat HAPS as a first-class RAN layer that your core can see and control like any other RAT.

• Design around four priorities:

  (1) mission-critical first,

  (2) simple payload stack,

  (3) sovereign/security-by-design,

  (4) KPIs you can fly and verify.

• Build for everyday coverage and surge capacity—but prove value on the worst day (floods, fires, fibre cuts, power loss).

Architecture Overview

Payload (air segment)

• LTE/5G eNB/gNB with high-gain, multi-beam antenna system

• Optional airborne MEC (edge compute) for MCX, video, GIS, and analytics

• LMR/TETRA interworking gateway for legacy radio integration

• Cell Broadcast agent interfacing with the national/enterprise CBE

Backhaul (choose two for diversity)

• Microwave to high ground or metro hub

• Ka/Ku satellite to teleport

• LEO relay for path redundancy

Core Integration (ground segment)

• Standard EPC/5GC interfaces

• Policy (PCF) rules for HAPS steering & pre-emption

• NWDAF analytics fed by HAPS telemetry for closed-loop assurance

Security & Compliance

• Zero-trust ground segment, in-country key custody, signed SBOMs, lawful-intercept alignment

Radio & Beam Design

• Bands: Prefer existing licensed LTE/5G bands for handset compatibility; add mid-band for capacity, low-band for reach.

• Beams: One macro umbrella for command/control plus sectorised beams for towns, transport corridors, ports, or hospitals.

• Link budget: High-gain arrays, beamforming, adaptive coding; in disaster modes, bias the scheduler to voice/PTT and messaging to protect control plane and public-safety services.

• Mobility: Larger timing-advance windows; conservative handover thresholds; pre-computed beam footprints for likely AOIs.

Backhaul & Gateways (resilience first)

• Primary path: Microwave to a hardened site with protected power.

• Secondary path: Satellite/LEO relay; router auto-failover with QoS preservation.

• Residency & LI: Place core-facing gateways in-country; enforce data residency and lawful-intercept.

• QoS mapping: Maintain QCI→5QI mappings end-to-end so public-safety traffic keeps its class through the backhaul.

Core & Policy Control — Make HAPS “just another RAT”

• Registration: Advertise HAPS cells with distinct TAC/TAI to target policy and slices precisely.

• Policy (PCF):

  • Normal ops → prefer terrestrial; steer only priority UEs to HAPS.

  • Outage/disaster → pre-empt consumer classes and promote safety slices automatically.

• Slicing (5G): Define S-NSSAIs for (1) Public Safety MCX, (2) Health & Emergency Ops, (3) Utility/SCADA, (4) General Public.

• Analytics: Feed HAPS KPIs into NWDAF; trigger actions (drop video bitrates, expand PRBs for MCX, shift beams).

Public-Safety Slice Blueprint (Mission-Critical First)

Target services: MCX (PTT/video/data), voice fallback, secure messaging, location.

Design

• Dedicated S-NSSAI with guaranteed PRBs and pre-emption.

• Admission control: whitelist responder SIM/eSIM; cap UEs per cell for MCX integrity.

• QoS: Tight loss/jitter; low-bitrate resilient video for bodycams; PTT latency ≤300 ms (95th).

• Inter-agency roaming: Temporary IMSIs/eSIMs for mutual aid; rapid provisioning at the EOC.

Edge (optional)

• Airborne MEC for group comms, video transcoding, object detection (smoke/fire/person), and local GIS layers to reduce backhaul load.

LMR/TETRA Interoperability

• Gateway (air or ground) bridging MCX talkgroups ↔ legacy LMR/TETRA.

• Map talkgroups and keys via the EOC directory; log for after-action review.

• Fallback: If MCX degrades, maintain voice-only interop.

Cell Broadcast & Public Warning

• Connect the national/enterprise CBE to HAPS cell broadcast.

• Pre-build multilingual templates: flood, fire, storm, evacuation, shelter, medical.

• Broadcast is out-of-band relative to user traffic; schedule alerts even when consumer traffic is throttled.

• Geo-target with beam polygons; log delivery for compliance.

Security & Sovereignty by Design

• Zero-trust: mTLS everywhere, hardware roots of trust, immutable logging.

• Key custody: In-country KMS/HSM for SIM/eSIM and network keys; strict emergency credential workflows.

• Supply chain: Signed firmware (SBOMs), attestation on boot, secure update pipelines.

• Lawful intercept: Intercept points at in-country gateways; auditable processes.

Disaster Playbook (Tech Flow)

1. Pre-event

   • Load policy packs (pre-emption rules, throttle tables).

   • Whitelist responder UEs; stage bilingual broadcast templates.

   • Test dual backhaul; confirm generator autonomy and spares.

2. Activation (H+0 → H+6)

   • Launch; attach umbrella beam; auto-register to core.

   • Public-safety slice up first; push initial cell broadcast.

   • Enable LMR/TETRA interop; verify MCX talkgroups.

3. Stabilisation (H+6 → H+72)

   • Add beams for hospitals, ports, depots, EOCs.

   • Run MEC analytics for situational video; throttle public traffic if needed.

   • Reroute backhaul as restoration proceeds.

4. Handover & Stand-down

   • Offload to repaired terrestrial sites; keep HAPS as a guardrail.

   • Export logs, CB history, and KPIs to the incident record.

Disaster KPIs (Executive Dashboard)

Availability & Speed

• Time-to-air (TTA): launch → first attach (target: hours)

• Time-to-alert (T2A): CBE trigger → handset broadcast (target: minutes)

Mission-Critical Performance

• PTT latency (95th): ≤300 ms

• MCX call setup success; handover success under load

• Bodycam/field video success rate at 256–512 kbps

Coverage & Capacity

• Population under coverage (dynamic estimate)

• PRB utilisation by slice; beams at capacity

Continuity & Impact

• Sites restored with HAPS support

• Service credits avoided (currency)

• Responder attach success; device density supported

Pilot Bill of Materials

• Air: HAPS platform; eNB/gNB; high-gain antennas; optional MEC; telemetry & flight control

• Backhaul: Microwave kit; sat/LEO modem & antenna; routing with auto-failover; crypto

• Core: EPC/5GC integration; PCF policies; NWDAF hooks; slice definitions; IMS/MCX stack

• Interop & Alerts: LMR/TETRA gateway; CBE connector; multilingual templates

• Security: HSM/KMS; LI mediation; SIEM feeds; SBOM verification toolchain

• Ops: Ground equipment; spares; test devices; flight/drive-test kits

90-Day Test Plan

Phase 1 — Bring-up

• Attach umbrella beam; verify registration and PCF enforcement.

• Baseline RSRP/RSRQ/SINR; attach time; idle→connected latency.

Phase 2 — Mission-Critical

• Whitelist responder UEs; measure PTT latency, MCX call setup, cell broadcast delivery time.

• Validate LMR/TETRA voice quality & group mapping.

Phase 3 — Stress & Failover

• Load consumer traffic; pre-empt to protect MCX; verify QoS integrity.

• Hard-cut primary backhaul; confirm automatic failover with KPI stability.

Phase 4 — Report

• Executive dashboard with the KPIs above; rollout recommendations.

Risks & Mitigations

Risk	Mitigation
Airspace / weather	Seasonal planning; dual launch sites; pre-cleared corridors; high-wind abort rules
Spectrum conflicts	Early interference studies; coordination with regulator; dynamic power/tilt control
Backhaul fragility	Dual path; LEO relay; QoS marking end-to-end
Security incident	Zero-trust; HSM; SBOM; continuous monitoring; red-team drills
Interop complexity	Lab integration for MCX + LMR; end-to-end runbooks; operator/EOC exercises

Conclusion: Engineer for the Worst Day

A stratospheric RAN that defaults to public-safety first, broadcasts on command, and fails over by policy transforms HAPS from novelty to necessity. Build the payload and integration around these principles, measure the disaster KPIs, and you’ll have a platform that delivers everyday coverage—and proves its worth when everything else fails.