From GEO Incumbents to LEO Challengers: The Innovation Timeline

Satellite internet terminal innovation has unfolded across three distinct phases—foundational concept establishment, ecosystem maturation, and the current surge driven by LEO mega-constellation deployment. Patent publication dates in this dataset span from 1998 to 2026, providing a clear view of how the technology has evolved and which architectural choices have endured.

The foundational period (1998–2013) established the baseline concepts still relevant today. Globalstar’s 1998 patent on satellite-controlled power control for personal communication user terminals introduced uplink power control loops via LEO constellations—a concept that directly anticipates modern non-terrestrial network (NTN) power control challenges. Alcatel’s hybrid satellite/terrestrial routing patents (2006 and 2013) established the architectural precedent for latency-aware traffic steering between satellite and ground paths. ViaSat’s improved spot-beam GEO terminal system (CN, 2010) defined the high-capacity GEO VSAT modem reference point, describing approximately 1 Gbps per spot beam using 300 MHz carriers with TWTAs at full saturation.

The growth and standardization period (2014–2021) saw GEO VSAT ecosystems mature into commercial products. Hughes Network Systems’ patents covering bandwidth allocation and high-capacity satellite systems reflect this maturation. Rockwell Collins’ airborne LEO connectivity patent (US, 2017) signalled the first commercial push for LEO-integrated mobile terminals with steerable antenna elements actively tracking proximal LEO satellites. The highest density of recent filings is concentrated in the 2022–2026 window, dominated by CN and KR jurisdictions—reflecting intense LEO mega-constellation buildout and 3GPP NTN standardization activity.

Phased Array vs. Fixed Antenna: The Core Hardware Tradeoff in Satellite Internet Terminals

The fundamental design tension in satellite internet terminal hardware is between phased array (steerable) antennas and fixed or simple antennas. Phased arrays enable seamless handover across LEO constellations and cross-satellite measurements, but impose significant cost and power burdens. Fixed antennas constrain terminals to opportunistic access windows aligned with satellite orbital passes, but dramatically reduce hardware cost—a tradeoff that is reshaping the IoT terminal market.

The GEO terminal ecosystem, represented by ViaSat’s spot-beam VSAT system and Hughes Network Systems’ high-capacity satellite systems, operates against fixed geostationary targets—making phased arrays unnecessary. ViaSat’s dynamic beam assignment patent (BR, 2020) shows how GEO terminals can be managed via ground processing nodes that monitor beam drift and dynamically reassign user terminals to maintain signal quality, incorporating load balancing—without any antenna steering at the terminal. XMW’s ultra-compact VSAT with dual RF front-end (KR, 2024) demonstrates how GEO terminal design has shifted toward cost and form-factor reduction: a single antenna, modem, and power supply enabling operation across two adjacent microwave bands simultaneously.

For LEO terminals, the picture is more complex. Rockwell Collins’ 2017 airborne LEO connectivity patent represents the first-generation approach: steerable antenna elements actively tracking proximal LEO satellites from an aircraft platform, delivering latency values within video conferencing tolerances. This architecture—essentially a miniaturised phased array—remains the reference design for aviation and maritime mobility terminals. However, a distinct innovation direction has emerged that trades tracking capability for cost reduction entirely.

“Phased array dominance is not inevitable for all terminal classes. IoT and low-mobility endpoints can achieve LEO connectivity with fixed or discrete-state antennas using ephemeris-scheduled access, dramatically reducing bill-of-materials cost.”

Gilat Satellite Networks’ non-tracking fixed-antenna IoT terminal (CA, 2023) explicitly identifies that LEO pass durations—measured in minutes—are sufficient for IoT data volumes accumulated between passes, eliminating the need for beam steering. Tsinghua University’s pan-synchronous orbit multi-access patent (CN, 2026) takes this further: terminals point in a fixed direction at scheduled times using pre-planned time-frequency resource allocation broadcast from the constellation, explicitly avoiding phased array or mechanical steering antennas. The patent argues this approach is analogous to what LoRa achieved for terrestrial IoT—enabling mass-market deployment at low per-unit cost.

Huawei’s antenna state switching method (CN, 2025) represents a cost-performance compromise between these two extremes. The terminal pre-computes which antenna state—from a discrete set of fixed-direction states—provides maximum gain toward the next satellite based on ephemeris prediction, enabling handover without continuous beam steering. According to 3GPP‘s NTN specifications, this class of terminal architecture is increasingly relevant as direct-to-device satellite connectivity is standardised in Release 17 and Release 18.

NTN Protocol Stack: Where the Satellite Internet Modem IP Wars Are Being Fought

The non-terrestrial network (NTN) protocol stack—covering timing advance compensation, uplink power control, MAC-layer delay management, and random access procedures—has become the primary IP battleground for Korean and US chipset vendors. Qualcomm, LG Electronics, and Samsung are building deep patent moats in this space, and entrants designing LEO terminal modems without freedom-to-operate analysis face high litigation risk.

LEO satellites impose Doppler shifts and propagation delay variations of up to tens of milliseconds depending on orbital pass geometry. This is not a marginal engineering challenge: the entire 3GPP random access procedure, designed for terrestrial base stations with propagation delays measured in microseconds, must be re-engineered for LEO. Qualcomm’s MAC-CE delay patent (CN, 2022) defines MAC-CE delay durations calibrated to channel propagation delay in NTN. User terminals defer new communication parameter implementation until after the MAC-CE delay expires following satellite acknowledgment—a fundamental change to how MAC control elements are processed.

LG Electronics has filed two complementary patents addressing the same problem from different angles. The first (KR, 2023) enables terminals to select power control settings from a plurality configured by the NTN based on satellite orbit information—LEO versus GEO—enabling orbit-adaptive uplink power management. The second (KR, 2023) derives terminal-specific timing advance from terminal GPS location and satellite ephemeris to achieve precise random access timing in LEO NTN. Samsung’s NTN terminal communication method (KR, 2025) covers synchronization via multiple broadcast messages, while a separate LG filing (KR, 2025) covers timing advance update periods for NTN uplink repetition. As documented by ITU, these protocol adaptations are foundational to enabling satellite broadband under international spectrum coordination frameworks.

Satellite Selection and Handover in LEO Mega-Constellations

As mega-constellations deploy hundreds to thousands of LEO satellites, terminal firmware must dynamically select among LEO, MEO, and GEO options, execute seamless handover, and optimize for latency, bandwidth, and coverage continuity. The patent dataset reveals three distinct algorithmic approaches to this challenge, ranging from policy-based selection to ML-native predictive handover.

ZTE’s satellite selection method (CN, 2024) establishes the baseline: terminals apply network-provisioned satellite selection policies to choose among GEO (maximum coverage, highest latency), MEO (moderate latency and coverage), and LEO (minimum coverage area, lowest latency, highest bit rate) based on service QoS requirements. This policy-based approach is straightforward but reactive—the terminal selects based on current conditions rather than predicting future degradation.

Huawei’s satellite access method (CN, 2026) advances this by integrating real-time attitude data, position data, antenna beam pattern, satellite ephemeris, and obstruction data to identify a target satellite and optimal attitude angle for access. This addresses LEO fast-pass acquisition—the challenge of acquiring a satellite signal quickly enough during a short orbital pass window. A companion Huawei filing on antenna state switching (CN, 2025) pre-computes which discrete antenna state provides maximum gain toward the next satellite based on ephemeris prediction, enabling handover without continuous beam steering.

Galaxy Space’s ML-native handover architecture (CN, 2026) represents the most advanced approach in this dataset. A pre-trained predictive model combines SNR, remaining satellite bandwidth, terminal position, and time delay to determine which satellite each terminal connects to at the next time slot—shifting from reactive to predictive handover management. This is consistent with the direction that WIPO‘s Global Innovation Index has identified as a key differentiator in next-generation satellite systems: AI integration at the access layer.

Hughes Network Systems’ MEO-LEO integrated satellite communication system (BR, 2024) introduces another layer of complexity: user terminals routing data over optical inter-satellite links (ISLs) across a combined MEO-LEO system, with direct UT-to-UT IP routing. This implies terminal firmware capable of resolving routing decisions across two orbital shells simultaneously—a significant step beyond current single-orbit terminal designs. Hughes’ smart network edge system (BR, 2025) extends this to the enterprise edge: a single customer premise device with a steerable antenna and multiple modems that selects among available satellite constellations and terrestrial networks, avoiding inter-constellation interference.

Geographic and Assignee Concentration: Who Holds the Satellite Terminal IP

The geographic distribution of satellite internet terminal patents in this dataset reveals a clear strategic picture: Chinese entities dominate the recent LEO mega-constellation terminal ecosystem, Korean entities concentrate on 3GPP NTN protocol stack adaptations, and Western operators show strength in system architecture and GEO VSAT hardware—but with fewer recent filings compared to CN/KR volumes.

Among 70+ retrieved results with identifiable jurisdictions, the distribution is heavily weighted toward CN (approximately 35% of relevant satellite-specific records), KR (approximately 28%), BR (approximately 12%), ES (approximately 5%), and scattered entries in US, CA, CO, DE, MX, and PT. Brazil’s prominence as a filing jurisdiction reflects regional IP protection strategy by international operators—Hughes, ViaSat, and Ericsson—rather than domestic innovation originating in Brazil.

The strategic implications of this geographic concentration are significant. Chinese entities—Huawei, ZTE, Galaxy Space, Beijing University of Posts and Telecommunications, and Tsinghua University—collectively represent the largest filing cluster in satellite selection, handover, and access protocol innovation in this dataset. This is a strategic indicator of China’s national priority in indigenous LEO broadband infrastructure, consistent with the national space programme priorities documented by OECD in its space economy assessments. Korean filings from Samsung, LG Electronics, and KT concentrate heavily on 3GPP NTN protocol stack adaptations—timing advance, uplink power, and random access procedures—reflecting Korea’s position as a leading 5G chipset and handset manufacturer. US and Western entities (ViaSat, Hughes, Gilat, Telesat, Boeing) show strength in system architecture, GEO VSAT hardware, and NGSO routing, but with fewer recent filings in this dataset compared to CN/KR volumes. R&D teams should ensure PCT filings cover CN, KR, and BR jurisdictions explicitly, and freedom-to-operate searches must extend beyond USPTO and EPO databases to CNIPA and KIPO.

Emerging Directions: ML Handover, MEO-LEO Hybrids, and 5G NTN Direct-to-Device

Six distinct emerging directions are visible in the 2024–2026 filing cohort, each representing a technology vector that will shape satellite internet terminal design over the next planning horizon. Together they define the frontier of what satellite terminal hardware and firmware will need to accomplish by 2030.

ML-Driven Predictive Satellite Selection

Galaxy Space’s filed method (CN, 2026) uses pre-trained prediction models incorporating SNR, remaining bandwidth, and terminal priority to proactively switch satellite connections before degradation. This represents a shift from reactive to predictive handover management—a significant architectural change that requires on-device ML inference capability in the terminal modem. The approach is analogous to what predictive handover does in 5G terrestrial networks, but applied to a much faster-moving and more variable channel environment.

Ephemeris-Integrated Terminal Attitude Control

Huawei’s satellite access method (CN, 2026) and antenna state switching method (CN, 2025) both fuse satellite ephemeris, terminal GPS position, and terminal attitude/orientation data to pre-compute optimal antenna direction and time to switch—enabling LEO satellite acquisition without full phased-array beam steering. This ephemeris-first design philosophy reduces the need for real-time RF feedback loops, lowering modem complexity and power consumption.

MEO-LEO Hybrid Constellation Terminals

Hughes Network Systems’ MEO-LEO integrated satellite communication system (BR, 2024) describes user terminals routing data over optical ISLs across a combined MEO-LEO system, with direct UT-to-UT IP routing. This implies terminal firmware capable of resolving routing decisions across two orbital shells simultaneously. Telesat’s dual LEO constellation patent (CO, 2018) established the conceptual foundation for multi-shell terminal management. Together, these suggest that terminal hardware requiring multi-orbit simultaneous reception and optical ISL-aware routing will become necessary for high-reliability enterprise and government users within the 2026–2030 planning horizon.

Fixed-Antenna LEO IoT Terminals via Scheduled Access

Tsinghua University’s pan-synchronous orbit multi-access patent (CN, 2026) explicitly argues that terminals need only point in a fixed direction at predetermined times, eliminating phased arrays. China Star Network Research Institute’s IoT transmission method (CN, 2026) and Gilat’s non-tracking IoT terminal (CA, 2023) reinforce this direction. This architectural approach targets mass-market IoT deployments at low per-unit cost—a distinct terminal class from high-throughput broadband terminals that has its own IP landscape and competitive dynamics.

5G NTN Direct-to-Device Terminal Protocols

Multiple KR-jurisdiction filings from LG Electronics, KT, and Samsung (2025–2026) address discontinuous beam pattern reception, NTN uplink repetition with TA updates, and random access window adaptation. These signal active development of consumer handsets with built-in LEO satellite modems compliant with 3GPP Release 17/18 NTN standards. The convergence of terrestrial 5G and satellite NTN in a single consumer device represents the most commercially significant near-term development in this technology domain, as tracked by standards bodies including ETSI.

Smart Hybrid Terrestrial-Satellite Edge Nodes

Hughes’ smart network edge system (BR, 2025) describes a single customer premise device with a steerable antenna and multiple modems that selects among available satellite constellations and terrestrial networks, avoiding inter-constellation interference. This converged terminal architecture for multi-path connectivity represents the enterprise endpoint for the hybrid satellite-terrestrial integration trend that Alcatel’s routing patents established as a concept in 2006. The progression from concept to productized patent over two decades illustrates the long innovation cycles characteristic of satellite infrastructure.

“MEO-LEO hybrid architectures signal the next GEO-replacement cycle. Terminal hardware requiring multi-orbit simultaneous reception and optical ISL-aware routing will become necessary for high-reliability enterprise and government users within the 2026–2030 planning horizon.”