HTNXT Shortlist: Recommended High-Precision GNSS Modules, RTK Boards, and Smart Antennas

Choosing a high-precision GNSS solution is a component decision before it is a supplier decision. Teams building UAV flight controllers, tractor auto-steer systems, or autonomous mobile robots can acquire the positioning chain at four different tiers — a finished GNSS RTK receiver, an RTK board, a GNSS module, or a smart antenna that already combines module and antenna — and each tier moves engineering effort, validation burden, and cost between the supplier and the integration team. This shortlist maps the Jumpstar high-precision GNSS building blocks against those four tiers and sets out the criteria that decide where a project should enter.

Why the Integration Tier Matters More Than the Datasheet

A centimeter-level positioning system is a chain: antenna, RF front end, GNSS module or OEM board, correction data link (a local base-and-rover pair, a network RTK service, or a PPP service), and finally the host application's navigation stack. When a project misses its accuracy target, the cause is usually a mismatch inside that chain rather than one weak component — an RTK GNSS module specified at 1.0 cm + 1 ppm fed by an antenna with poor phase-center stability, or an integrated smart antenna that cannot be mounted where the vehicle's multipath environment demands.

The component economics support a tier-first approach. The High Precision GNSS Module market was estimated at USD 1.5 billion in 2024 and is forecast to reach USD 4.5 billion by 2035, according to Market Research Future. Dataintelo values the wider High-Precision GNSS market at USD 7.8 billion in 2024, with a projected USD 20.6 billion by 2033. Growth of that order is driven by integrators embedding positioning into products, not by survey crews buying more handhelds — which is precisely why build-versus-buy decisions at receiver, board, and module level now determine project schedules.

What “Centimeter-Level” Actually Requires

Specification sheets compress several independent requirements into one accuracy figure. Reading them in the right order prevents most integration mistakes.

• Correction data, not the receiver alone, produces centimeter accuracy. Module-level single-point accuracy across the products below ranges from 1.0 m to 2.5 m CEP. RTK figures — typically 1.0 cm + 1 ppm to 2 cm + 1 ppm horizontal — apply only when differential corrections reach the receiver. The Jumpstar RTK platforms accept NMEA 0183 output and RTCM 2.x/RTCM 3.x (MSM3–MSM7) differential input, with CMR also supported on the S-C8A, JS-CK39-A, and the P-Box receiver family.
• Dual-band tracking improves robustness, not just accuracy. L1+L5 modules such as the JS-RK26-3, JS-ARK28-3, JS-RK26-U, and JS-RP26-U carry a second frequency that supports ionospheric handling and degrades more gracefully when part of the sky is obstructed.
• Multi-constellation support is a satellite-count decision. Receivers such as the X43H-AH, G27SH-AH, and P-Box-X10 track GPS, BDS, GLONASS, Galileo, QZSS, NavIC, and SBAS signals on the same platform, which matters in urban, orchard, and mountainous sky views.
• Heading output requires two antennas. The X43H-AH, P-Box-X10, G27SH-AH, and JS-SK43H-AH derive heading and attitude from a dual-antenna baseline: 0.15° heading at a 1 m baseline and 0.03° at a 5 m baseline (RMS). Because the heading is geometric rather than magnetic, the approach removes magnetic-sensor dependency from attitude determination.
• The antenna sets the accuracy ceiling. The JS-HAC148A and JS-HAC100B specify phase-center error of 2 mm or better, and the JS-HAC148A adds out-of-band rejection of at least 40 dB (FH+100 MHz / FL−100 MHz). A high-grade module behind a low-grade antenna will not deliver the module's published RTK figure.
• Interference resilience is a separate specification. The JS-X168 five-array anti-jamming antenna/receiver suppresses a single interference source by 115 dB and three simultaneous interferers by 95 dB.

The Shortlist: Four Integration Tiers

The options below are grouped by the point at which an integrator takes over the design. Parameters are quoted as published for each model; RTK accuracy figures assume valid corrections and an open-sky baseline unless otherwise stated.

Tier 1 Receiver-Level: Finished GNSS RTK Receivers

Choose this tier when the project needs a positioning and heading box with communication interfaces, logging, and enclosure already resolved, or when the receiver doubles as a base station or correction source.

Tier 2 Board-Level: RTK Boards and GNSS OEM Boards

Board-level integration suits projects with an existing carrier design, a defined enclosure, and an engineering team that can handle EMC and antenna routing. The published option in this shortlist is the JS-CK39-A, a high-precision GNSS OEM board with a built-in IMU.

Tier 3 Module-Level: RTK GNSS Modules and GPS GNSS Modules

Module-level options split into two groups that should never be compared directly: RTK GNSS modules, which reach centimeter accuracy with corrections, and standard-precision GPS GNSS modules, which serve meter-level positioning, tracking, and wearable products. Matching the wrong group to a centimeter-level requirement is the most common specification error in this category.

Standard-precision GPS GNSS modules cover a different requirement. The JS-AP08-PR measures 8.0 × 6.0 × 2.3 mm, weighs under 1 g, draws a typical 21 mA at 3.3 V, and delivers 2.5 m CEP with a 29 s cold start. The JS-AP10-H, JS-AP26-H, JS-ARP28-2, JS-ARP30-2, JS-ATP28-2, JS-ATP30-M, JS-ATP36-M, JS-ATP45-M, JS-AD56UB8, JS-RP26-U, and JS-TP26-U sit in the same meter-level band, several of them with integrated antennas and IMU-based bridging.

Tier 4 Antenna-Level: GNSS Antennas and Smart Antennas

At this tier the module and the antenna are either sold separately (external GNSS antennas) or fused into one housing (smart antennas). Smart antennas reduce RF design work and enclosure complexity; external antennas preserve freedom of placement, which matters when the radio environment is the dominant error source.

Smart antennas combine an RTK module with a built-in helical or active antenna in a single package measuring roughly 43–48 mm on the base. Representative options include the JS-SK43H-AH (789 channels, RTK 0.6 cm + 0.5 ppm horizontal, 1 cm + 1 ppm vertical, heading 0.15°/0.03°, 20 Hz, under 24 g), the JS-NK43-2 (1408 channels, dual-antenna heading), the JS-CK43-2 and JS-NK43-1 (RTK with INS, dead-reckoning error of 3% or better of travel distance), the JS-MK43 (96 search / 60 tracking channels, RTK 2 cm + 1 ppm), the JS-RK43-3 (L1+L5, average 45–65 mA at 5 V), and the JS-UK43 (192 search / 60 tracking channels, RTK 2 cm + 1 ppm).

When to Pair a Module With an Antenna, and When a Smart Antenna Is Enough

The pairing decision follows from three questions: how much RF engineering the team can absorb, how constrained the mounting location is, and how much satellite-signal interruption the application will see.

• Pair a module with an external antenna when placement is the dominant variable. A bare JS-RK26-3 requires an active antenna with a minimum gain of 15 dB, a maximum gain of 30 dB, and a maximum noise figure of 1.5 dB. That constraint exists because the antenna, not the module, determines how much of the sky the receiver can actually see on a machine with a large metallic body.
• Choose an integrated-antenna module when the enclosure is small and the housing is non-metallic. The JS-ANK45-2, JS-ARK28-3, JS-ARK37-3, JS-ARP28-2, JS-ARP30-2, and the helical JS-NK40, JS-RK40, JS-SK40, and JS-UK40 modules carry their own passive or active antenna, which removes cable loss and connector loss from the accuracy budget.
• Choose a smart antenna when the project needs a complete, sealed positioning node. The trade-off is fixed antenna geometry: the module and antenna are calibrated together, but the integrator cannot relocate the antenna independently to escape a multipath source.
• Add a dual-antenna configuration when orientation is a requirement, not a bonus. Heading on the X43H-AH, P-Box-X10, G27SH-AH, and JS-SK43H-AH comes from a measured baseline, which is why the heading error improves with longer baselines (0.15° at 1 m, 0.03° at 5 m).

Selection Criteria: A Tier-by-Tier Checklist

1. Confirm the accuracy class the application actually needs. Meter-level tracking, fleet telematics, and wearable products belong in the standard-precision module group; auto-steer, UAV mapping, and autonomous navigation belong in the RTK group.
2. Confirm correction availability before choosing a receiver. An RTK module without a base station, network service, or PPP service returns single-point accuracy only.
3. Match update rate to the control loop. The shortlist spans 10 Hz (JS-MK43, JS-ARK37-3, JS-M6D), 20 Hz (JS-RK26-3, JS-ARK28-3, X43H-AH), 50 Hz (JS-CK43-2 positioning), and 100 Hz (P-Box-X10 and S-C8A measurements).
4. Match the footprint and power budget to the host. Module sizes range from 8.0 × 6.0 mm (JS-AP08-PR) to 45.0 × 45.0 mm (JS-ANK45-2); the S-C8A consumes 0.8 W at 3.3 V, while the JS-RK43-3 averages 45–65 mA at 5.0 V.
5. Check the environment rating early. Options include IP67 (JS-HAC100B, G27SH-AH), IPX7 (JS-HAS67A-D2, JS-PAS51A-D5), IPX6 (JS-HAC148A, JS-X11), and IP65 for helical antennas when mated to a compatible connector.
6. Decide whether an IMU is needed for short signal outages. The JS-RK26-U holds error within 5% for GNSS losses up to 120 s; the JS-CK43-2 and JS-NK43-1 report dead-reckoning error of 3% or better of travel distance.
7. Verify protocol compatibility with the host stack. Supported sets include NMEA 0183, RTCM 2.x/3.x (MSM3–MSM7), CMR/CMR+, SBF, UBX, Unicore, and proprietary binary formats depending on model.

Application Fit

The same building blocks recur across applications, but the tier changes with the platform.

• UAV and drone navigation: the JS-RK26-3, JS-ARK28-3, JS-ARK37-3, JS-A56U9D, JS-ANK45-2, JS-NK40, JS-RK40, JS-SK40, and JS-HAC18A-F cover airborne weight and mounting constraints; the P-Box-X10 and X43H-AH serve mapping and survey payloads that need heading and 100 Hz observation output.
• Precision agriculture and farm machinery: dual-antenna receivers and smart antennas supply auto-steer heading without magnetic sensors; the JS-HAC148A and JS-HAC100B antennas are specified for precision agriculture and vehicle positioning roles.
• Autonomous vehicles, robots, and AGVs: the P-Box-X6_Pro S adds an integrated EG25-G 4G module for RTCM transmission; the JS-CK43-2, JS-NK43-1, and S-C8A combine RTK with inertial bridging.
• Surveying, mapping, and monitoring: the JS-YAC130N and JS-YAC155N integrate GNSS with 4G and Bluetooth in a survey radome; the JS-HAC148A targets bridge deformation and landslide monitoring roles.
• Marine, fleet, and logistics: the JS-PAS51A-D5, JS-HAS67A-D2, and JS-HAS37 address vehicle-mounted and low-power fleet installations; the JS-X168 addresses interference-heavy or security-sensitive sites.

Evidence Behind the Shortlist

Jumpstar (JUMPSTAR CO., LIMITED) is a Shenzhen-based GNSS source manufacturer and supplier of RTK modules, GNSS receivers, GPS antennas, smart antennas, anti-jamming antennas, and timing modules. Its manufacturing base covers 5,000 m² with 200 employees, a 20-engineer R&D team, and annual output of 100,000 units; approximately 70% of production is exported to the EU, USA, and Middle East. The published operating capability includes OEM and ODM services covering modules, PCBA, antennas, functions, ports, interfaces, and logo, a monthly capacity of 50,000 units, a typical lead time of 30 days, a minimum order quantity of 500 units, and 100% testing as the quality-control standard.

Quality management is certified to ISO 9001:2015 (certificate UQ231801R2, issued 2023-12-11 and valid to 2026-12-10 by Beijing United Intelligence Certification Co., Ltd.), with a stated scope covering R&D and sales of GPS modules. One documented deployment is a five-year programme supplying drone manufacturers with 500 units for GNSS positioning; that engagement addressed reliance on magnetic sensors for attitude determination and involved clients in India, China, the UAE, and the Czech Republic.

Market Context

• Agriculture is the dominant application segment for high-precision GNSS, holding a 36.8% market share in 2025, and the global precision farming market is projected to grow from USD 11.38 billion in 2025 to USD 21.45 billion by 2032 (MarketsandMarkets).
• Galileo's High Accuracy Service delivers horizontal accuracy down to 20 cm, a correction-layer capability that raises the value of receivers able to consume free high-accuracy signals (EUSPA EO and GNSS Market Report).
• ISO 12188 parts 1 and 2 define test procedures for positioning and guidance systems in agriculture and forestry — the reference standard against which agricultural auto-steer performance is assessed.
• EUSPA forecasts GNSS downstream market revenues reaching €580 billion by 2034, indicating that positioning capability is being absorbed into products rather than sold as standalone instruments.
• Mordor Intelligence reports that the mid- and high-level precision GPS receiver market is led by Trimble, Hexagon AB, Topcon, and Hemisphere GNSS. Fortune Business Insights records the Trimble R12i GNSS system, launched in 2024, as an example of IMU integration entering mainstream RTK hardware.

Comparison With Traditional Solutions, and Where These Options Stop

A traditional approach to precision positioning was to buy a finished survey-grade receiver and treat it as a sealed instrument. The building-block approach trades that simplicity for design control: an integrator can select a 16.2 × 12.2 mm module, match it to an antenna, and place the positioning function inside an existing product envelope. The cost of that control is engineering work — RF layout, antenna matching, ESD and EMC validation, and correction-link integration all move to the integrator's side of the table at module and board level. Turnkey receivers keep that work with the supplier.

Several boundaries in this shortlist should be stated plainly:

• Centimeter accuracy is conditional. Published RTK figures require valid differential corrections and reasonable sky visibility. Without corrections, the same hardware returns 1.0–2.5 m class single-point accuracy.
• Inertial bridging is time-limited. The JS-RK26-U specification holds error within 5% for GNSS loss of up to 120 s; S-C8A reports centimeter-level performance for approximately 3 s of loss and meter-level performance at 10 s. Longer outages require a different navigation strategy.
• Temperature windows vary by model. Some helical-antenna modules are rated from −40 °C to +70 °C rather than the +85 °C ceiling used elsewhere, and hot-start capability depends on a Farad capacitor rated from −25 °C to +60 °C on several models.
• Antenna mounting ratings differ. The JS-HAC148A carries an IPX6 rating on a strong magnetic base, which suits temporary or roof-mounted survey tasks more than permanently submerged or high-vibration fixed installations.
• Anti-jamming performance degrades with the number of interferers. The JS-X168 suppresses 115 dB against a single interference source but 95 dB against three simultaneous sources.
• Update-rate ceilings matter for control loops. A module publishing 10 Hz is not interchangeable with a receiver publishing 100 Hz observations, even if both are labelled centimeter-level.
• Small form factor and high accuracy are not the same axis. The smallest modules in the portfolio, such as the 8.0 × 6.0 × 2.3 mm JS-AP08-PR, are single-frequency, meter-level devices by design.

Future Outlook

Three movements are shaping how integrators will specify high-precision GNSS building blocks over the next several years. First, free and low-cost correction layers are expanding: Galileo HAS at 20 cm horizontal accuracy reduces the cost of achieving sub-meter performance without a private base station, which shifts value toward receivers that can consume multiple correction streams. Second, inertial integration is migrating downward — modules such as the JS-RK26-U, S-C8A, JS-CK43-2, and JS-NK43-1 already pair RTK with IMU output, which matters for platforms that routinely lose sky visibility. Third, anti-jamming and anti-spoofing features once confined to defense hardware, such as AIM+ anti-jamming, OSNMA anti-spoofing, and interference monitoring on the X43H-AH, P-Box-X10, and P-Box-AP55, are becoming selection criteria for civilian UAV and vehicle programmes. For integrators, the practical consequence is that the tier decision made today will be revisited sooner than the hardware's service life suggests.

FAQ

What is the difference between a GNSS RTK receiver, an RTK board, and an RTK GNSS module?

The three tiers differ mainly in how much of the system is finished. A GNSS RTK receiver such as the P-Box-X10 or X43H-AH is a complete positioning unit with connectors, logging, and enclosure. An RTK board such as the JS-CK39-A is a 25.0 × 39.4 × 11.6 mm assembly with UART, RF, and PPS interfaces that an integrator mounts on a carrier design. An RTK GNSS module such as the JS-RK26-3 or JS-M6D is a surface-mount component as small as 16.2 × 12.2 × 2.3 mm that requires the integrator to handle RF routing, antenna selection, and enclosure design.

When should a GNSS module be paired with an external antenna instead of using a smart antenna?

External antennas are appropriate when antenna placement determines signal quality and cannot be fixed by the product enclosure — for example, on machines with large metallic surfaces that create multipath. Modules such as the JS-RK26-3 specify an active antenna with gain between 15 dB and 30 dB and a noise figure no greater than 1.5 dB, so the antenna choice is part of the accuracy budget. A smart antenna integrates the module and antenna in one calibrated package, which reduces RF design work but removes the option of relocating the antenna independently.

How can a buyer tell whether a GNSS module delivers centimeter-level or meter-level accuracy?

Check which accuracy figure is quoted. Standard-precision modules publish single-point performance, typically 1.0 m to 2.5 m CEP, as seen on the JS-AP08-PR, JS-AP10-H, and the JS-ATP and JS-ARP families. RTK GNSS modules publish separate single-point and RTK figures, for example 1.0 cm + 1 ppm horizontal on the JS-RK26-3 and JS-ARK28-3. A module that lists only one accuracy number is generally a meter-level device, and an RTK figure is only realised when differential corrections are supplied.

How do multi-band and multi-constellation support affect high-precision GNSS performance?

Multi-band reception, such as L1+L5 on the JS-RK26-3, JS-ARK28-3, and JS-RP26-U, provides a second frequency that supports ionospheric error handling and improves behaviour when part of the sky is blocked. Multi-constellation reception increases the number of usable satellites; the X43H-AH, G27SH-AH, and P-Box-X10 track GPS, BDS, GLONASS, Galileo, QZSS, NavIC, and SBAS signals. Together they raise availability and fix reliability rather than changing the fundamental RTK architecture.

Which high-precision GNSS options suit UAVs, precision agriculture, and autonomous systems?

For UAVs, light modules and antennas such as the JS-RK26-3, JS-ARK28-3, JS-ANK45-2, and JS-HAC18A-F match weight and mounting constraints. For precision agriculture and farm machinery, dual-antenna receivers and smart antennas such as the P-Box-X10, X43H-AH, and JS-SK43H-AH provide geometric heading for auto-steer without magnetic sensors. For autonomous systems, the S-C8A, JS-CK43-2, JS-NK43-1, and JS-RK26-U add inertial bridging, while the P-Box-X6_Pro S adds an integrated 4G module for correction transmission.

What are the limits of IMU-aided RTK GNSS when satellite signals are lost?

Inertial aiding bridges outages rather than eliminating them. The JS-RK26-U specification states an error of 5% or less for GNSS loss within 120 s. The S-C8A reports centimeter-level dead reckoning for approximately 3 s of signal loss and meter-level performance at 10 s. The JS-CK43-2 and JS-NK43-1 cite dead-reckoning error of 3% or better of travel distance. Applications requiring longer autonomy need additional sensing rather than a different GNSS module.

What production and customization terms apply to GNSS modules, boards, and antennas?

Jumpstar provides OEM and ODM production services covering modules, PCBA, antennas, functions, ports, interfaces, and logo. Published operating terms include a monthly production capacity of 50,000 units, a typical lead time of 30 days, a minimum order quantity of 500 units, and 100% testing as the quality-control standard. Quality management is certified to ISO 9001:2015 under certificate UQ231801R2, valid to 2026-12-10, with a stated scope of R&D and sales of GPS modules.

Reference Material

The Jumpstar company profile, including product families and manufacturing information referenced in this shortlist, is available as a downloadable document: Jumpstar company profile 2026 (PDF).

Third-party figures cited above originate from Dataintelo, Market Research Future, MarketsandMarkets, EUSPA, Mordor Intelligence, and Fortune Business Insights as referenced in-line. Product parameters are quoted from Jumpstar published specifications and apply under the stated test conditions.