Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61108: Maritime GNSS Receiver Standards — GPS, GLONASS, Galileo, and BeiDou Engineering Guide
The IEC 61108 series is arguably the most consequential standard family for maritime electronic navigation. It establishes the performance requirements, test methods, and required test results for Global Navigation Satellite System (GNSS) receiver equipment used aboard SOLAS-class vessels. Published in four parts covering GPS (Part 1), GLONASS (Part 2), Galileo (Part 3), and BeiDou (Part 4), the standard directly implements IMO Resolution MSC.112(73) and its revisions. For engineers designing, certifying, or integrating maritime navigation equipment, a thorough understanding of IEC 61108 is essential — not merely as a compliance checkbox, but as the foundation for receiver architecture decisions.
📡 1. Standard Architecture and Multi-Constellation Landscape
IEC 61108 adopts a multi-part structure where each part addresses one satellite constellation, yet they all share a unified performance baseline derived from IMO’s Performance Standards for Shipborne GNSS Receivers. This modular approach means that a receiver certified under one part (e.g., 61108-1 for GPS) can be extended to other constellations without re-running the full test suite for shared requirements such as environmental robustness or EMC immunity.
| Part | Constellation | Orbit / Satellites | Civil Signals | Position Accuracy (95%) |
|---|---|---|---|---|
| IEC 61108-1 | GPS (USA) | MEO, 31 operational | L1 C/A, L2C, L5 | < 5 m |
| IEC 61108-2 | GLONASS (Russia) | MEO, 24 operational | G1 (FDMA), G2 (FDMA), G3 (CDMA) | < 7 m |
| IEC 61108-3 | Galileo (EU) | MEO, 26 operational + spares | E1, E5a, E5b, E6 (CS) | < 4 m |
| IEC 61108-4 | BeiDou BDS (China) | MEO + IGSO + GEO, 44 total | B1I, B1C, B2a | < 5 m |
✅ Engineering Insight: A critical RF front-end consideration is that GPS L1 (1575.42 MHz), Galileo E1 (1575.42 MHz), and BeiDou B1C (1575.42 MHz) all share the same center frequency. This enables a shared LNA and SAW filter chain, significantly reducing BOM cost and PCB footprint for multi-constellation receivers. GLONASS G1 at 1602 MHz, however, requires either a separate RF path or a wideband front-end covering 1559–1610 MHz with sufficient linearity to handle the additional noise. For production designs targeting global markets, a wideband RF front-end with a digital down-converter (DDC) is the recommended approach — it adds roughly 15–20% to the RF BOM cost but eliminates the risk of constellation lock-out in mixed-fleet deployments.
🔧 2. Core Performance Requirements and Design Implications
IEC 61108 defines a comprehensive set of performance thresholds. While all four parts have slight variations reflecting constellation-specific characteristics, the baseline requirements are remarkably consistent. The following areas carry the most engineering weight:
2.1 Position Accuracy and Dynamic Tracking
The standard mandates horizontal position accuracy better than 10 meters at 95% confidence — a threshold that comfortable modern receivers surpass by a factor of 2-3x under open-sky conditions. However, the requirement for sustained tracking at vessel speeds up to 30 knots (approximately 55 km/h) introduces real design constraints. At this speed, Doppler shifts on L-band carriers can reach ±5 kHz during turns (acceleration up to 0.3 g). Conventional second-order phase-locked loops (PLLs) with 10–20 Hz bandwidth will struggle to maintain lock during rapid heading changes. The engineering solution is an FLL-assisted PLL architecture where a frequency-locked loop provides coarse frequency tracking (loop bandwidth 40–80 Hz) while the PLL refines carrier phase for precision ranging. For high-dynamic platforms such as patrol boats or fast ferries, a vector DLL/PLL architecture using a centralized navigation filter to cross-couple tracking loops across satellites offers significantly better robustness at the cost of increased computational load.
2.2 Acquisition Time and Sensitivity
Three startup modes are defined: Hot start (valid ephemeris, almanac, position, and time) — acquisition within 10 seconds; Warm start (almanac and approximate position available, ephemeris expired) — within 35 seconds; Cold start (no valid aiding data) — within 5 minutes. The cold-start requirement is the most demanding from a signal-processing perspective. At a signal power of −130 dBm (the specified minimum for acquisition), the received carrier-to-noise density ratio (C/N₀) is approximately 33 dB-Hz — only 3 dB above the GPS L1 C/A code’s typical tracking threshold.
💡 Design Note: The cold-start specification dictates the correlator architecture. Traditional serial search across the 32-code × 20-kHz frequency uncertainty space would require approximately 90 seconds on a 32-channel receiver (assuming 1 ms coherent integration per cell). Modern receivers use FFT-based parallel acquisition engines capable of searching the entire frequency domain simultaneously. A 16K-point FFT on a 1 ms C/A code period compresses the cold-start time to 25–35 seconds, well within the 5-minute requirement. For multi-constellation receivers, the search space grows proportionally — GLONASS FDMA adds 24 frequency channels on top of the code-phase search — making FFT acquisition essential rather than optional.
2.3 Receiver Autonomous Integrity Monitoring (RAIM)
RAIM is arguably the most technically sophisticated requirement in IEC 61108. It demands that the receiver autonomously detect a satellite fault (or localized signal anomaly) within 10 seconds and alert the user with a “don’t use” indication. The underlying principle is statistical consistency checking among redundant pseudorange measurements. Fault detection (FD) requires a minimum of 5 visible satellites; fault detection and exclusion (FDE) requires 6 or more.
The standard implementation uses the weighted least-squares residual method:
1. Compute pseudorange residual vector: r = (I − H(HᵀWH)⁻¹HᵀW) Δρ 2. Calculate the sum-of-squared errors: SSE = rᵀWr 3. Apply a threshold from the χ² distribution (k = N−4 degrees of freedom) with a false-alarm rate P_fa = 10⁻⁵ per independent sample 4. If SSE exceeds threshold: satellite fault declared 5. For FDE: iteratively remove one satellite, recompute, until SSE passes
⚠️ Critical Engineering Consideration: RAIM availability is geometrically constrained. In low-latitude regions with only 5 visible satellites, the dilated GDOP significantly reduces the fault-detection probability. The receiver must continuously compute the Horizontal Protection Level (HPL) and compare it against the Horizontal Alert Limit (HAL), which IMO defines as 25 meters for ocean navigation. When HPL exceeds HAL, the receiver should flag the condition as “RAIM not available” — even if no fault is detected. Failing to implement this HPL/HAL comparison is a common compliance gap during type-approval testing.
🧭 3. Maritime-Specific Design Challenges
The maritime operating environment differs fundamentally from aviation or land-mobile GNSS applications, and IEC 61108 includes specific provisions that directly influence receiver hardware and software design.
3.1 Multipath Mitigation at Sea
The sea surface presents one of the most challenging multipath environments in GNSS. The L-band reflection coefficient of calm seawater can exceed 0.8, producing specular reflections that interfere constructively and destructively with the direct line-of-sight signal. The resulting pseudorange error can reach 15–30 meters for code-based measurements — enough to push a receiver past the 10-meter threshold. Several engineering countermeasures are available: choke-ring antennas suppress gain below 10° elevation, reducing sensitivity to sea-surface reflections; dual-polarization antennas exploit the fact that reflected signals undergo a polarization reversal (right-hand circular to left-hand circular); and at the baseband level, narrow correlator spacing (0.1 chip versus the traditional 1.0 chip) combined with a double-delta discriminator can suppress multipath-induced ranging errors below 3 meters.
3.2 Electromagnetic Compatibility in Shipboard Environments
A modern vessel is an electromagnetic nightmare for sensitive GNSS receivers. X-band (9.4 GHz) and S-band (3 GHz) navigation radars, satellite communication terminals (Inmarsat, VSAT, Starlink), and high-power variable-frequency drives for propulsion and thrusters all generate significant in-band or out-of-band interference. IEC 61108 requires the receiver to maintain full functionality under 100 V/m field strengths from 10 kHz to 30 GHz. Meeting this specification demands: LNA stages with IIP3 exceeding +15 dBm to prevent intermodulation; a SAW filter cascade with at least 60 dB out-of-band rejection; and rigorous PCB layout partitioning with grounded via fences separating the RF front-end, digital baseband processor, and power regulation stages.
🔴 Critical Warning: Do not underestimate conducted interference through the power supply rail. Starting a bow thruster or deck crane can inject voltage dips of 300–500 mV with sub-millisecond edges onto the ship’s DC bus. The GNSS receiver’s power management unit must incorporate a wide-input-range LDO with ≥ 60 dB power supply rejection ratio (PSRR) at 100 kHz, supplemented by a common-mode choke (≥ 100 Ω at 100 MHz) and a bidirectional TVS diode rated for the ship’s transient profile. Poorly designed power regulation is the single most common root cause of intermittent GNSS lock-loss observed in post-commissioning failure reports — more frequent than RF interference or antenna degradation.
3.3 Antenna Siting and Installation Constraints
IEC 61108 does not prescribe antenna placement, but the standard’s performance requirements implicitly constrain it. The antenna must have a clear 360° field of view down to 5° elevation. In practice, this means mounting the antenna at the highest accessible point on the vessel, clear of radar beam paths, exhaust plumes, and crane swing arcs. The recommendation is to maintain a horizontal separation of at least 3 meters from any radar antenna and vertical separation of at least 2 meters from VHF/UHF communication antennas. The antenna cable should use low-loss RF coax (e.g., LMR-400 or equivalent) with total insertion loss not exceeding 3 dB, and the GNSS receiver should supply DC bias (typically 3.3–5 V) through the same coax for the active antenna’s LNA.
❓ Frequently Asked Questions (FAQ)
Q1: Can a receiver certified under IEC 61108-1 (GPS) legally substitute for a Part 4 (BeiDou) receiver on SOLAS vessels?
Not directly. While the performance baseline is harmonized, SOLAS equipment requirements typically reference specific parts. A GPS-only receiver (Part 1) is acceptable if it meets the carriage requirements for the vessel’s operating area. However, the trend in both IMO and national maritime authorities is toward multi-constellation mandating — China’s CCS and Russia’s RMRS already require BeiDou and GLONASS support, respectively, for vessels flagged under their jurisdictions. Future-proof designs should implement at least dual-constellation capability.
Q2: What is the practical difference between RAIM and SBAS augmentation?
RAIM is a receiver-autonomous algorithm requiring no external infrastructure — it uses measurement redundancy alone to detect faults. SBAS (e.g., WAAS, EGNOS, BDSBAS) provides differential corrections and integrity bounds via geostationary satellites. IEC 61108 does not mandate SBAS, but receivers combining both achieve superior integrity: SBAS delivers sub-meter accuracy with bounds, while RAIM provides a fallback integrity layer independent of external broadcast. Note that SBAS corrections are only valid within the broadcast service volume — an SBAS-dependent receiver operating outside WAAS or EGNOS coverage loses its integrity guarantees.
Q3: How does the cold-start time requirement scale for multi-constellation receivers?
The cold-start search space grows with each added constellation. For a GPS+GLONASS receiver, the total uncertainty is the sum of the GPS C/A code search (32 codes × 20 kHz) plus the GLONASS FDMA search (24 frequency channels × 511 code chips). Without parallel hardware, this could extend cold-start to 5–8 minutes. However, modern receivers with 192+ correlator channels and multi-channel FFT engines typically achieve full multi-constellation cold-start in 40–60 seconds. The key is to prioritize the strongest signals first — typically GPS L1 — then use the decoded time to seed the other constellations’ search spaces.
Q4: What are the most common type-approval failures for IEC 61108 compliance?
Based on published test house data, the top three failure modes are: (1) RAIM false-alarm rate exceeding the 10⁻⁵ requirement due to overly sensitive residual thresholding — often caused by unmodeled multipath in the receiver’s own measurement engine; (2) cold-start time exceeding 5 minutes under −130 dBm conditions, typically from inadequate coherent integration gain or suboptimal search strategy; and (3) conducted susceptibility failures during the 80 MHz–2 GHz immunity sweep, traced to inadequate filtering on the power input and data interface (NMEA 0183/2000) ports. Pre-compliance testing of these three areas before submitting to a notified body significantly increases first-pass approval probability.