Starlink Direct-to-Cell in Africa: MTN Zambia test and the QoE validation challenge

On March 7-9, 2026, MTN Zambia completed the first Starlink Direct-to-Cell (D2C) test on the African continent. A standard, unmodified smartphone connected to a Starlink satellite and sent a text message. No specialized antenna. No satellite phone. A regular handset, communicating through low Earth orbit infrastructure as if it were connecting to a terrestrial cell tower.

This test is not a curiosity. It is the beginning of a fundamental shift in how Africa’s 1.4 billion people access mobile connectivity, and it introduces validation challenges that existing field testing methodologies were never designed to handle.

What happened in Zambia

The technical setup

MTN Zambia, in partnership with SpaceX, conducted a controlled Direct-to-Cell trial over a 48-hour window. The test parameters:

Parameter	Detail
Dates	March 7-9, 2026
Location	Rural Zambia (undisclosed site)
Technology	Starlink Direct-to-Cell (LTE Band 4 emulation)
UE	Standard commercial smartphone (unmodified)
Services tested	SMS, basic data
Satellite constellation	Starlink V2 Mini (D2C-capable)
Orbit altitude	~550 km LEO

The test confirmed that a standard LTE-capable handset could register on the satellite-based cell, complete authentication, and exchange data. This is functionally equivalent to what T-Mobile demonstrated in the United States in late 2024, but it is the first validation on African soil with an African operator.

Why Zambia matters

Zambia’s mobile landscape makes it a natural proving ground:

• Population: ~20 million
• Mobile penetration: ~56%
• 4G coverage: ~35% of land area
• Rural population: ~55% (largely unconnected or 2G-only)
• Terrain: mix of savanna, plateau, and river valleys that challenge terrestrial deployment

The economic case is straightforward. Deploying terrestrial macro sites to cover Zambia’s rural 745,000 km² is prohibitively expensive. At an average cost of $150,000-250,000 per macro site (including backhaul), achieving 90% geographic coverage would require thousands of additional towers and billions in capital expenditure. D2C provides a coverage floor with zero terrestrial infrastructure.

The broader African NTN landscape

Airtel’s 14-market D2C deployment

MTN is not alone. Airtel Africa announced plans to deploy Direct-to-Cell services across 14 African markets, including Nigeria, Kenya, Tanzania, Uganda, Rwanda, and the DRC. Airtel’s approach targets initial SMS and emergency service coverage, with data services following in 2027.

The 14-market rollout represents approximately 450 million potential subscribers who currently lack reliable coverage in at least part of their daily movement patterns. Even providing basic SMS and emergency calling to these subscribers fundamentally changes the connectivity equation.

NTN market trajectory

The Non-Terrestrial Network market is projected to reach $9.66 billion by 2030, growing at a CAGR of approximately 59% from 2024. This growth is driven by:

Segment	2024	2030 (projected)	Primary driver
Direct-to-Cell (consumer)	$0.2B	$3.8B	Coverage extension
IoT satellite backhaul	$0.8B	$2.4B	Agriculture, logistics
Enterprise NTN	$0.3B	$1.9B	Mining, oil & gas
Government/defense	$0.5B	$1.5B	Sovereign connectivity

Africa represents the largest addressable market for consumer D2C because it has the largest unconnected population with existing handset ownership. An estimated 300 million Africans own a mobile phone but experience regular coverage gaps in their daily lives.

QoE validation challenges for D2C networks

The fundamental problem: a moving cell tower

In terrestrial networks, the base station is fixed and the UE moves. In D2C, both the satellite (cell) and the UE may be moving, and the cell is traveling at approximately 27,000 km/h at 550 km altitude. A single Starlink satellite is visible from any ground point for approximately 4-6 minutes before handover to the next satellite.

This creates validation challenges that have no precedent in terrestrial field testing:

Temporal variability

The same geographic location will experience different performance at different times because the satellite constellation geometry changes continuously. Unlike terrestrial networks where a drive test at 10 AM and 2 PM yields comparable RF conditions (assuming traffic load is similar), a D2C measurement at 10:00:00 and 10:06:00 may connect to entirely different satellites with different elevation angles, path losses, and interference conditions.

Validation implication: single-pass drive tests are insufficient. D2C QoE validation requires multi-pass temporal sampling across multiple orbital periods to establish statistical coverage confidence.

Geographic variability

Terrestrial coverage is determined by antenna height, tilt, power, and terrain. D2C coverage is determined by satellite elevation angle, atmospheric conditions, and the satellite’s instantaneous beam pattern. A location with excellent D2C coverage at one moment may have degraded coverage minutes later as the serving satellite moves.

Elevation angle	Expected path loss (550 km LEO)	Typical RSRP range
90° (directly overhead)	~155 dB	-85 to -95 dBm
45°	~162 dB	-95 to -110 dBm
25°	~168 dB	-110 to -120 dBm
10° (near horizon)	~175 dB	Below sensitivity

Validation implication: coverage maps must be time-averaged across orbital periods rather than representing instantaneous snapshots.

Capacity sharing

A single Starlink D2C satellite illuminates a ground cell approximately 30-50 km in diameter. Every UE within that footprint shares the available bandwidth, which is significantly lower than a terrestrial macro cell. Early D2C implementations offer approximately 2-5 Mbps aggregate capacity per beam, shared across all connected users.

For QoE validation, this means that unloaded throughput measurements are misleading. The relevant metric is per-user throughput under realistic loading, which requires either coordinated multi-UE testing or statistical modeling based on expected subscriber density.

Satellite-terrestrial handover: the critical transition

The most challenging QoE validation point occurs at the boundary between terrestrial and satellite coverage. When a subscriber walks from a village with LTE coverage into an area served only by D2C, the handover must occur seamlessly. In practice, this involves:

1. Terrestrial signal degradation below a threshold
2. NTN cell search and synchronization (different timing advance due to propagation delay)
3. Registration on the satellite cell
4. Service continuity (for voice, this means no call drop; for data, no session interruption)

The timing advance alone is problematic. At 550 km altitude, the round-trip propagation delay is approximately 3.7 ms, compared to microseconds for a terrestrial cell at 500 meters. The UE must adjust its timing advance by orders of magnitude during handover.

Field validation of this transition requires:

• L3 message capture showing the complete handover signaling sequence
• Timing advance measurements before and after handover
• Service continuity metrics (packet loss, latency spike, session drops)
• GPS-referenced location marking of the exact handover boundary

Measurement methodology for D2C validation

A practical D2C field validation campaign in Africa should follow this structure:

Phase 1: Temporal coverage profiling

Deploy static measurement points at representative locations. Log continuously for 24+ hours to capture full orbital period coverage statistics. Key outputs:

• Coverage availability percentage (what fraction of time is service available?)
• RSRP/SINR distribution across satellite passes
• Service registration success rate per pass

Phase 2: Transition zone mapping

Identify and walk the boundaries between terrestrial and satellite coverage. Capture:

• Handover trigger conditions
• Handover completion time
• Service interruption duration
• Fallback behavior (what happens when handover fails?)

Phase 3: Application QoE under D2C

Test realistic applications over the D2C link:

Service	Metric	Acceptable threshold (D2C)
SMS	Delivery time	<30 seconds
Voice (if supported)	MOS	>2.5
Basic web	Page load (text-heavy)	<10 seconds
Messaging apps	Message delivery	<15 seconds
Emergency call	Setup time	<5 seconds

Note that D2C QoE thresholds are deliberately lower than terrestrial standards. The comparison is not D2C versus 5G; it is D2C versus no service at all.

Phase 4: Multi-UE capacity validation

Simulate realistic loading by deploying multiple UEs in the same satellite beam footprint. Measure per-user throughput degradation as active users increase. This is critical for operators planning commercial pricing and capacity dimensioning.

Africa telecom investment context

D2C does not exist in isolation. Africa’s telecom sector is experiencing significant investment:

Metric	Value
Projected 5G connections (Africa, 2026)	54 million
Tower market size (Africa, 2026)	$3.5 billion
Annual capex (top 10 African operators)	~$8 billion
Fiber-to-tower backhaul penetration	~25%
Average revenue per user (Sub-Saharan)	$3.20/month

The challenge for African operators is that terrestrial 5G deployment targets urban centers (Lagos, Nairobi, Johannesburg, Cairo) while the coverage gap is rural. D2C addresses the rural gap without cannibalizing urban 5G investment.

For field engineering teams, this means maintaining dual competency: terrestrial 5G validation in urban areas and NTN/D2C validation in rural zones. The tools must support both paradigms.

Implications for field diagnostic tools

D2C validation demands capabilities that most existing drive test tools do not have:

1. Extended logging duration: 24-hour static measurements versus 4-hour drive routes
2. Satellite-aware KPIs: elevation angle, Doppler shift, timing advance in the millisecond range
3. NTN L3 message parsing: 3GPP Release 17 NTN-specific NAS and RRC messages
4. Temporal aggregation: statistical coverage over orbital periods, not instantaneous snapshots
5. Offline capability: measurements in areas without terrestrial backhaul require local storage and deferred upload

Smartphone-based diagnostic tools have an inherent advantage for D2C validation. The test device is the same form factor as the subscriber device, meaning the measurements directly represent the end-user experience. There is no uncertainty about antenna gain differences or modem behavior differences between a scanner and a commercial handset.

The road ahead

MTN Zambia’s test is a proof of concept. Commercial D2C service in Africa likely launches in late 2026 or early 2027, beginning with SMS and emergency services. But the validation work starts now.

African operators planning D2C deployment need to:

1. Establish baseline terrestrial coverage maps to identify exact D2C target zones
2. Develop NTN-specific acceptance criteria that reflect realistic D2C capabilities, not terrestrial benchmarks
3. Build temporal measurement capability for statistical coverage assessment
4. Train field teams on satellite-terrestrial handover validation, which is a new skill set
5. Integrate D2C KPIs into existing network monitoring platforms for unified coverage visibility

The 300 million Africans in coverage gaps are not waiting for fiber. They are not waiting for macro towers. They are waiting for a signal. D2C provides that signal, and field validation ensures it actually works when it matters.

Validating Direct-to-Cell QoE requires a paradigm shift from instantaneous coverage snapshots to statistical temporal profiling. The measurement tools must capture L3 signaling, RF metrics, and application QoE simultaneously across multi-hour sessions in areas with zero terrestrial infrastructure.