CafeTele5G / 6G Engineering CT-5G-NTN-500 Enroll now ↗
Course/Day 5/Planning & Link Budget
05
5
Day Five

NTN Planning: LEO / MEO / GEO Link Budget

Turn everything into a planning capability — a repeatable workflow, coverage design, full link-budget engineering with a live worked example, capacity & KPI planning. The capstone workshop follows on its own page.

Day objective — a repeatable NTN planning workflow that ends in a professional, customer-ready planning report.

5.1

NTN Planning Methodology

The end-to-end planning workflow — and the constraints that shape every orbit and payload decision.

  • End-to-end workflow — requirements → orbit / payload → coverage → link budget → core/RAN config → mobility → capacity → KPI / acceptance → report
  • Requirements capture — service, coverage, capacity & latency; UE / service profiles
  • Orbit selection strategy (LEO vs. MEO vs. GEO) & transparent vs. regenerative payload choice
  • Constraints — commercial, spectrum / regulatory & integration
5.2

NTN Coverage Planning

Footprint, beams and the elevation-angle trade-off that sets slant range, delay and margin.

  • Footprint & spot-beam planning — earth-fixed vs. earth-moving beam design
  • Service-link & feeder-link coverage — gateway placement & visibility
  • Minimum elevation angle & its effect on slant range, delay & link margin
  • Cell-edge evaluation & overlap for service continuity (rural, maritime, aviation, disaster)
5.3

Link-Budget Fundamentals

The link budget decides whether a given orbit, payload and terminal can deliver the target service. Participants build it term by term.

  • Transmit — EIRP = Tx power + antenna gain − losses
  • Path — FSPL, atmospheric / gaseous loss, rain fade (ITU-R P.618), scintillation, polarization loss, implementation margin
  • Receive — antenna gain, system noise temperature, noise figure, G/T figure of merit
  • Quality — C/N, C/N₀, Eb/N₀ and SINR; required Eb/N₀ for the target MODCOD & BLER
  • Availability margin for the target service-availability %
FSPL = 92.45 + 20·log₁₀(dkm) + 20·log₁₀(fGHz) C/N₀ = EIRP − Lpath + G/T − k (k = −228.6 dBW/K/Hz) Eb/N₀ = C/N₀ − 10·log₁₀(Rb)
5.4

Worked Example — LEO Downlink to Handheld (S-band)

Built on the TR 38.821 methodology. Drag the elevation angle — slant range, FSPL, C/N₀ and SNR recompute live with exact math.

  • Editable link-budget calculator — LEO-600 S-band handheld downlink; change EIRP, elevation, band, bandwidth
  • Worked example, line by line — C/N₀ = 66.1 dB-Hz, and why 5 MHz fails but 500 kHz closes
LEO S-band downlink link budgetlive compute
Carrier f2.0 GHz · FR1
Altitude h600 km · LEO
Satellite EIRP+34.0 dBW
UE G/T (handheld)−31.6 dB/K
Atm+scint+pol−2.8 dB
Shadow / impl. margin−3.0 dB
Slant range d1075 km
Free-space path loss−159.1 dB
Other losses−5.8 dB
UE G/T−31.6 dB/K
Boltzmann (−k)+228.6 dB
Resulting C/N₀66.1 dB-Hz
SNR @ 5 MHz−0.9 dB

EIRP, G/T & margins are TR 38.821-style planning placeholders, not a specific operator’s figures — but every derived value (d, FSPL, C/N₀, SNR) is computed exactly from the inputs above. C/N₀ is the bandwidth-independent headline metric; SNR scales with the reference bandwidth you choose.

5.5

LEO / MEO / GEO Link-Budget Design

What controls the margin in each orbit — and which direction (UL or DL, feeder or service) usually dominates.

  • LEO — short slant range & low FSPL, but high Doppler, short dwell & frequent handover; UL often limited by handheld power
  • MEO — balanced delay / coverage; satellite diversity & gateway-visibility planning; broadband example
  • GEO — highest FSPL & delay but a fixed, large footprint; rain-fade & availability dominate (Ku/Ka); VSAT / broadcast / delay-tolerant example
  • For each orbit — uplink & downlink, feeder-link & service-link budgets, and the controlling margin
5.6

Capacity, Frequency Reuse & KPIs

Dimensioning the beam/satellite/gateway, and the KPI acceptance criteria a customer signs off against.

  • Capacity — beam / satellite / gateway capacity; frequency reuse & beam isolation; spectrum reuse factor
  • Dimensioning — user-density & traffic-model prep; peak-hour dimensioning; QoS / slice-based capacity
NTN KPIs & acceptance criteria
KPI domainExample KPIsAcceptance intent
AccessibilityRACH success rate, PDU-session setup successUE attaches reliably over the NTN link
RetainabilityDrop rate, session continuityService sustained during beam / sat changes
MobilityBeam-HO & satellite-HO success rateHandover works under constellation motion
Latency / throughputOne-way delay, jitter, DL/UL throughputMeets service profile (incl. high-RTT)
Link / coverageLink-budget margin, C/N₀, edge coverage %Margin holds at target availability
5.7

Planning Tools & Simulation Workflow

The open-source toolchain that automates ephemeris, coverage, delay/Doppler and the link budget — and produces the report.

  • Ephemeris / TLE handling (Skyfield / Orekit); coverage & pass simulation
  • Delay & Doppler simulation; Python-based link-budget automation
  • System-level — ns-3 / 5G-LENA NTN; GIS-based coverage visualization
  • Reporting — planning-report template preparation
LAB

Hands-on Lab 5 — Link Budget & Pass Prediction

Plan coverage and close budgets like a real NTN engineer — reproduce a TR 38.821 reference link budget, derive the coverage geometry and its governing constraint, and produce a real 24-hour pass schedule.

Hands-on Lab 5

Link Budget & Pass Prediction (Planning)

  • 1Reproduce a TR 38.821 §6.1.3 reference link budget (SC-case) to ≤ 0.06 dB; vary elevation & band
  • 2Derive coverage geometry — beam footprint, cells, continuity, gateway visibility — and flag the binding constraint
  • 3Predict a real 24-hour satellite pass schedule with Skyfield/SGP4 on live TLEs
  • 4Estimate the minimum constellation size from the longest coverage gap
Skyfieldlink_budget.pycoverage.py
Deliverable — a one-page link-budget report + coverage table + pass schedule. Open the full lab guide →
QUIZ

Day 5 Assessment

Close the link and plan the system — FSPL/C-N0/G-T, the worked LEO example, orbit design, capacity/KPIs and the open-source toolchain. 42 questions; every number computed, not fabricated.