3GPP Release 19: A Guide to 5G-Advanced Features

3GPP Release 19 is the second major release of 5G-Advanced (following Release 18), the final release fully dedicated to 5G evolution, and a bridge toward early 6G studies in subsequent releases.

It prioritizes commercial performance improvements, energy efficiency, non-terrestrial network (NTN) maturity, AI/ML integration, mobility enhancements, extended reality (XR) support, and IoT/RedCap advances, while consolidating prior work. Specifications reached full implementability by the end of December 2025 (on-schedule completion), with functional/Stage 3 freeze around September 2025 and protocol stability/ASN.1 freeze by December 2025 (SA#110). As of mid-2026, the release is frozen; only corrections and error fixes are permitted.

Release content was scoped at the December 2023 TSGs (#102) after extensive workshops. The authoritative summary appears in TR 21.919 (“Release description; Summary of Rel-19 Work Items”). Release 20 continues 5G-Advanced commercial focus while introducing the first 6G studies (requirements, architecture, security, radio evolution).

Timeline and Status

  • Scoping: Content decisions finalized December 2023 (TSG#102).
  • RAN work start: Q1 2024 (RAN1), staggered for RAN2/3/4.
  • Functional freeze: Approximately June–September 2025 (RAN1 earlier; RAN2/3/4 later).
  • Full implementable specifications / end date: December 2025 (SA#110).
  • Status (2026): Frozen. Work has shifted to Release 20 (open, targeting ~2027) and early Release 21/6G activities.

Official portal and 3GPP materials confirm the freeze and position Rel-19 as the last release fully dedicated to 5G.

Core Priorities and Feature Areas

Rel-19 balances continuity (refining commercial 5G deployments) with innovation. Major themes drawn from TR 21.919 summaries, RAN/SA reports, and industry analyses include:

Non-Terrestrial Networks (NTN) / Satellite (5GSAT) Phase 3

  • Downlink coverage enhancements (repetitions of PDCCH/PDSCH/SIB1, extended SS/PBCH periodicity).
  • Uplink capacity improvements via multiplexing techniques (e.g., inter-slot orthogonal cover codes).
  • Regenerative payloads (full gNB onboard satellite).
  • Multicast/broadcast support with intended service-area signaling.
  • RedCap and enhanced RedCap device support in FR1 NTN bands.
  • IoT-NTN Phase 3 (including TDD mode, store-and-forward for discontinuous coverage).
  • Mobility (inter-RAT TN to NR-NTN), positioning, new bands (Ku, additional S-band, Extended L-band), PWS support, GNSS assistance (including BDS, NavIC), and UAS/aerial enhancements.
  • Multi-orbit interoperability focus (LEO/MEO/GEO + terrestrial).

These build on Rel-17/18 foundations for direct-to-cell and broader coverage.

Energy Efficiency and Savings

  • Network energy savings enhancements (on-demand SSB/SIB1, adaptive common signals/channels, cell DTX/DRX refinements).
  • Low-power wake-up signal and receiver (LP-WUS/WUR) for idle/inactive UEs to extend battery life.
  • Energy efficiency treated as a service criterion.

AI/ML Integration

  • AI/ML for the NR air interface (one-sided models at gNB or UE for beam management, positioning, CSI prediction/compression; lifecycle management framework).
  • Core-network enhancements (NWDAF support for policy/QoS, vertical federated learning, model transfer, signaling storm mitigation).
  • Management and application enablement for AI/ML services.
  • Dual aspects: “AI for the network” (optimization) and “network for AI” (efficient transport of AI data).

MIMO and Radio Enhancements (NR MIMO Phase 5 and related)

  • CSI support extended toward 128 ports; codebook enhancements.
  • UE-initiated, event-driven beam reporting for lower latency/overhead.
  • Support for 3TX uplink (codebook- and non-codebook-based).
  • Mobility improvements, including L1/L2-triggered mobility (LTM) extensions across gNBs and dual connectivity.
  • Sub-band non-overlapping full duplex (SBFD) evolution.
  • Multi-carrier, RRM, demodulation, and high-power UE (HPUE) refinements.

XR / Immersive Services and Edge Computing

  • XR for NR Phase 3 (capacity/QoS improvements, reduced power consumption, cancellable measurement gaps, rate adaptation, tighter 5GC coordination).
  • Avatar communications in AR calls, split rendering over IMS, immersive voice/audio (IVAS) enhancements.
  • Edge computing Phase 3 (architecture, industrial scenarios, satellite edge).

IoT, RedCap, and Ambient IoT

  • Power-class 2 RedCap in FR1, management aspects, NAS overhead reductions.
  • Ambient power-enabled IoT (new feature area) with associated charging and security.

Other Notable Areas

  • Sidelink (multi-hop U2N/UE-to-UE relays, ProSe Phase 3, ranging/positioning).
  • Mission-critical / public-safety / FRMCS enhancements.
  • Network slicing, SBA/protocol improvements, QoS/policy, multi-access (ATSSS Phase 4).
  • Verticals/NPN security and interconnect.
  • Various RF, band, CA/DC, and performance requirements.

Most features are evolutionary (“Phase n” iterations). Ambient IoT stands out as relatively new.

Context Within the Broader Evolution

  • Rel-15–17: Foundational 5G (eMBB, URLLC, mMTC expansions, initial NTN/RedCap).
  • Rel-18: First 5G-Advanced release (AI/ML studies, broader verticals, initial energy/XR/NTN advances).
  • Rel-19: Commercial maturation of 5G-Advanced + last pure-5G focus.
  • Rel-20 onward: Continued 5G-Advanced commercial needs + formal 6G studies (requirements, architecture, radio, security).

Rel-19 strengthens the foundation for multi-orbit connectivity, intelligent RAN, lower energy footprints, and immersive/low-power use cases that will carry into 6G discussions. Industry demonstrations (e.g., NR-NTN end-to-end links in new bands, immersive IMS calls) already show early implementation momentum.

Practical Implications

Operators and vendors can treat Rel-19 as a stable baseline for commercial feature rollouts (energy savings, NTN expansion, XR capacity, AI-assisted optimization). Device and network implementations benefit from clearer RF/RRM requirements, regenerative NTN architectures, and reduced-power mechanisms. Testing and conformance work continues for the frozen specifications.

1) Non-Terrestrial Networks (NTN) / Satellite (5GSAT) Phase 3

Non-Terrestrial Networks (NTN) / Satellite (5GSAT) Phase 3 in 3GPP Release 19 marks a significant maturation of satellite integration into 5G-Advanced systems.

It builds on the foundational NR-NTN and IoT-NTN frameworks introduced in Release 17 (primarily transparent payloads for LEO and GEO) and the mobility, regenerative, and coexistence enhancements of Release 18. Phase 3 focuses on practical performance gains under real satellite constraints (limited payload power, large footprints, long delays, and feeder-link limitations), architectural flexibility via regenerative payloads, broader device support, and system-level capabilities for intermittent or multi-orbit operations.

These advances support direct-to-device (including smartphones), IoT, broadband, multicast/broadcast, and public-safety use cases while improving seamless terrestrial–non-terrestrial interoperability. Specifications were completed as part of the frozen Rel-19 work (implementable by end-2025).

Evolution Context

  • Release 17: Initial NR-NTN framework (transparent bent-pipe payloads dominant) and IoT-NTN (NB-IoT/eMTC over satellite). Focus on basic connectivity, Doppler/delay compensation, and timing.
  • Release 18: Mobility enhancements, initial regenerative support, coexistence with terrestrial networks, uplink coverage improvements, and expanded bands.
  • Release 19 (Phase 3): Performance optimization (especially downlink for handsets), full regenerative gNB-onboard architectures, uplink multiplexing for capacity, RedCap/eRedCap support, multicast/broadcast service-area signaling, Store-and-Forward for intermittent links, GNSS-independent options, positioning enhancements, new bands, and system architecture/security/charging refinements.

Satellite aspects appear under both RAN (NTN) and SA (5GSAT) work items, with related efforts for UAS/aerial and air-to-ground networks.

Key Features of NR NTN Phase 3 (NR_NTN_Ph3)

The core RAN work item (led primarily under RAN2 with RAN1/RAN4 contributions) targets NG-RAN enhancements for FR1 and FR2 NTN bands. Major elements include:

Downlink Coverage Enhancements

Targeted at handset-type terminals (e.g., smartphones with limited antenna gain around –5.5 dBi) under power and footprint constraints.

  • Repetitions of PDCCH in the common search space.
  • Repetitions of PDSCH carrying Msg4 and SIB1 (within 20 ms).
  • Extended SS/PBCH block periodicity up to 160 ms for initial cell selection (FR1-NTN and FR2-NTN).
  • Support for additional reference satellite payload parameters (power-sharing among beams, alternative energy configurations for GSO/NGSO). These improve link margin and system-level throughput while accounting for large footprints and limited feeder bandwidth.

Uplink Capacity and Throughput Enhancements (primarily FR1-NTN)

  • Inter-slot orthogonal cover codes (OCC) for PUSCH transmission, enabling multiplexing of multiple UEs (e.g., length-2 sequences supporting two UEs).
  • Reduced signaling overhead for certain procedures. This addresses the challenge of many users under a wide satellite beam with constrained spectrum and power.

Regenerative Payload Support

A major architectural step: full gNB hosted onboard the satellite (beyond simpler regenerative processing).

  • Enables onboard demodulation, processing, switching/routing, and inter-satellite links (Xn interface support).
  • Benefits include lower round-trip time for Uu procedures (random access, HARQ), inter-gNB mobility without always routing via ground, resource coordination, energy saving features, and greater resilience.
  • Supports “data centers in the sky” concepts and direct UE-to-UE communication via satellite in some scenarios.
  • Contrasts with transparent (bent-pipe) payloads, where the satellite mainly frequency-converts and amplifies; regenerative places more intelligence in orbit. Related stage-2 descriptions appear in TS 38.300 updates.

Multicast/Broadcast (MBS) Support

  • Signaling of the intended service area (via enhanced SIB and core-to-NG-RAN interfaces) when the satellite footprint exceeds the actual service region.
  • Enables efficient delivery of broadcast services over large coverage areas without unnecessary resource use.

RedCap and Enhanced RedCap Device Support (FR1-NTN)

  • Adaptations for half-duplex UEs, including handling of timing-advance mismatches and collision scenarios in paired spectrum (semi-static and dynamic scheduling).
  • Enables wider-band services for lower-complexity, lower-cost, or power-constrained devices while maintaining NTN operation.

Performance requirements, RF aspects, RRM, and testing methodology were also refined (RAN4).

Broader 5GSAT Phase 3 and Related System Aspects

SA-level work (Satellite access Phase 3 / 5GSAT_Ph3) complements the RAN features with system architecture, requirements, security, and charging:

  • Store-and-Forward (S&F) operation: Supports delay-tolerant services (especially IoT) under intermittent or temporary satellite connectivity (e.g., NGSO constellations without continuous feeder links). Data can be stored onboard and forwarded when a ground link becomes available. Regenerative eNB/gNB payloads facilitate this.
  • GNSS-independent operation: Enables access for UEs without GNSS receivers or in GNSS-denied/degraded conditions (building beyond Rel-17/18 assumptions).
  • UE-to-UE communication via satellite: Direct communication between UEs under the same satellite coverage (with or without inter-satellite links).
  • Positioning enhancements: Gap analysis against terrestrial methods and support for satellite-only scenarios; additional GNSS systems (e.g., BDS B2b, NavIC L1 SPS).
  • On-demand GNSS assistance data broadcast.
  • Public Warning System (PWS) support over satellite E-UTRAN and NG-RAN (including NB-IoT NTN).
  • Security and charging aspects tailored to Phase 3 satellite access.
  • Inter-RAT mobility: Support from E-UTRAN terrestrial to NR-NTN.
  • New bands and spectrum: Additional NR NTN bands (e.g., Ku-band, further S-band options, Extended L-band combinations), HAPS-related bands, and specific LTE bands for 5G broadcast over geosynchronous satellites (e.g., in Region 3).

IoT-NTN Phase 3 (for LTE/NB-IoT) parallels many of these with Store-and-Forward using regenerative eNB payloads, uplink capacity enhancements (OCC multiplexing), TDD mode introduction, and PWS support.

Related items cover Uncrewed Aerial System (UAS) Phase 3, air-to-ground network enhancements for NR, and management aspects.

Practical Implications and Industry Progress

Phase 3 makes NTN more commercially viable for multi-orbit (LEO/MEO/GEO) deployments and hybrid terrestrial–satellite networks. Regenerative architectures reduce dependency on continuous ground connectivity and improve latency/resilience. Coverage and capacity gains particularly benefit handheld and RedCap devices, while Store-and-Forward unlocks sparse IoT deployments.

Demonstrations (e.g., end-to-end NR-NTN links in bands such as n252 alongside earlier n255/n256, satellite-to-satellite mobility, and cross-vendor interoperability) have validated key aspects using commercial modem silicon under realistic multi-orbit conditions. Testing emphasizes multi-orbit handovers, regenerative behavior, and seamless TN–NTN transitions.

Challenges and Outlook

Satellite constraints (power, delay, Doppler, large beams) remain, so enhancements are carefully scoped. Regenerative payloads increase onboard complexity and power draw compared with transparent designs. Full multi-orbit interoperability, regulatory spectrum harmonization, and end-to-end performance under mobility continue to drive implementation and testing work.

These Rel-19 capabilities form a stable baseline for commercial NTN rollouts and feed into Release 20 (further 5G-Advanced refinements plus early 6G studies that will further evolve NTN). Primary references include TR 21.919 summaries, the NR_NTN_Ph3 and 5GSAT_Ph3 work items, TS 38.300 and related RAN specifications, and SA architecture documents.

In summary, NTN/Satellite Phase 3 transforms satellite connectivity from a coverage-extension tool into a more capable, integrated, and flexible component of 5G-Advanced systems, enabling broader device support, efficient multi-user operation, onboard intelligence, and resilient service delivery across orbits and terrestrial networks.

2) Energy Efficiency and Savings

Energy Efficiency and Savings in 3GPP Release 19 form a core pillar of 5G-Advanced, addressing both network-side and device-side power consumption to support sustainable deployments while maintaining performance.

These features build on earlier releases (Rel-16/17/18 power-saving mechanisms such as DRX, DCP, and PEI) and introduce more granular, adaptive, and low-complexity solutions. The work prioritizes reducing always-on transmissions, enabling deeper device sleep states, and treating energy efficiency as an explicit service criterion. This supports operator cost reduction, environmental goals, longer device battery life (especially for IoT and RedCap), and prepares the ground for 6G energy-efficiency design principles. Specs were completed as part of the frozen Rel-19 package (implementable by end-2025).

Key work items include Enhancements of Network energy savings for NR, Low-power wake-up signal and receiver for NR (LP-WUS/WUR), and Energy Efficiency as Service Criteria.

Network Energy Savings Enhancements for NR

Following Rel-18 studies and work items, Rel-19 introduces mechanisms that allow the network to reduce energy consumption from periodic control signaling without significantly impacting coverage, accessibility, or UE performance (for idle/inactive and connected UEs).

On-demand SSB operation for SCells

  • Secondary cells can operate with no always-on SSB or with always-on SSB plus additional on-demand SSBs.
  • On-demand SSBs are configured via higher layers and activated, deactivated, or adapted through RRC or MAC-CE signaling.
  • Center frequencies of always-on and on-demand SSBs may differ.
  • Supports both L1 and L3 measurements; on-demand SSBs are not cell-defining.
  • Particularly useful in multi-carrier / inter-site carrier aggregation scenarios, allowing secondary cells to transmit SSBs only when needed (e.g., for measurements or access).

On-demand SIB1 for UEs in idle/inactive mode

  • Reduces energy from continuous periodic SIB1 transmissions.
  • UEs can request SIB1 transmission (via a wake-up signal / random-access preamble) in the required beam direction.
  • Typically applied in multi-carrier scenarios with an “anchor” cell assisting configuration for legacy compatibility; the NES (network energy saving) cell may be barred for non-capable UEs.
  • Enables cells to enter sleep modes more frequently while preserving essential system information delivery.

Adaptation of common signal/channel transmissions

  • Dynamic adjustment of SSB, PRACH, and paging configurations (e.g., periodicity adaptation between sparse and dense values, frame/half-frame offsets).
  • Supports more agile cell discontinuous transmission/reception (DTX/DRX) and beam-level control.
  • Enables the network to scale common-channel overhead with actual load.

These features allow base stations to mute or sparsify transmissions during low-load periods, delivering measurable energy savings (evaluations in related studies show meaningful reductions depending on traffic and configuration) while preserving accessibility for active users.

Low-Power Wake-Up Signal and Receiver (LP-WUS/WUR)

This is a major device-side innovation. Traditional wake-up signals (DCP for connected mode, PEI for idle/inactive) rely on PDCCH and coherent detection, limiting power-saving gains due to the need for relatively accurate time-frequency tracking and higher baseband complexity.

Rel-19 LP-WUS/WUR introduces a dedicated low-power design:

  • Waveform: Based on DFT-s-OFDM, generating a superposed signal of On-Off Keying (OOK) and OFDM sequences in the time domain. This common design supports both OOK-based and OFDM-sequence-based wake-up receivers.
  • Detection: Enables non-coherent methods (envelope/energy detection for OOK; sequence correlation for OFDM), which relax time-frequency accuracy requirements compared with coherent PDCCH-based approaches.
  • Architecture: UE employs a low-power wake-up receiver (LP-WUR or LR) that monitors the LP-WUS while the main radio (MR) can remain fully or largely powered down. The main radio wakes only when needed (e.g., for paging or data).
  • Applicability: Covers both RRC idle/inactive and connected states. Supports duty-cycled monitoring. Target coverage aligns with Message-3 PUSCH.
  • Supporting signals: Includes Low-Power Synchronization Signal (LP-SS) for timing/RRM reference (with configurable periodicity and limited overhead).
  • Procedures: Configuration of monitoring windows, sub-grouping, entry/exit conditions, paging indication via LP-WUS, and relaxed measurements. RF and RRM requirements are specified (see related TR 38.774 and core specs).

Evaluations indicate substantial UE power-saving gains—often cited in the 70–90%+ range for idle/inactive modes under low paging probability—because the main radio stays in deep sleep far longer than with conventional DRX/PEI. Trade-offs exist between power savings, coverage, and hardware complexity (energy detector vs. coherent detector). The design is viewed as a strong baseline for future 6G wake-up mechanisms.

Energy Efficiency as a Service Criterion

Rel-19 elevates energy efficiency beyond an internal network optimization goal:

  • Enables energy-related information collection, KPIs, policy control, and exposure (e.g., via APIs/northbound interfaces).
  • Allows services or network slices to be monitored and enforced against defined energy constraints without degrading intended performance.
  • Supports intent-driven energy saving, multi-domain optimization, and exposure of energy metrics to applications or operators.
  • Related management and orchestration work (e.g., TS 28.310 updates) covers assessment of NG-RAN and slice energy efficiency, intra-/inter-RAT saving procedures, and compensation mechanisms.

This creates a framework where energy efficiency can be treated as a service attribute alongside traditional QoS parameters, facilitating green SLAs and end-to-end optimization.

Broader Context, Benefits, and Outlook

Energy features in Rel-19 complement other areas such as AI/ML-assisted optimization (for smarter on/off decisions and resource allocation), RedCap/ambient IoT (ultra-low-power devices), and XR (power-efficient traffic handling). Network-side savings reduce operational costs and carbon footprint; device-side mechanisms extend battery life for IoT sensors, wearables, and smartphones.

Challenges include balancing savings against latency/accessibility (especially for on-demand procedures and deep-sleep wake-up times), ensuring backward compatibility, managing additional signaling or hardware (LP-WUR), and accurate modeling of base-station power consumption. Evaluations typically use standardized power models that account for RF chains, baseband, and sleep states.

These capabilities provide a practical foundation for commercial green 5G-Advanced networks and directly inform 6G design goals, where energy efficiency is expected to be a first-class requirement from the outset. Primary references include the relevant sections of TR 21.919, the Netw_Energy_NR_enh and NR_LPWUS work items, associated RAN1/2/3/4 specifications, SA energy-efficiency studies, and supporting technical reports on RF requirements and procedures.

In practice, operators can combine on-demand common-channel control with LP-WUS-capable devices and energy-aware policies to achieve substantial, measurable reductions in both network and terminal energy consumption while sustaining service quality.

3) AI/ML Integration

AI/ML Integration in 3GPP Release 19 advances the dual paradigm of “AI for the Network” (using intelligence to optimize network operation) and “Network for AI” (enhancing the system to efficiently support AI/ML applications and data flows).

Building on foundational frameworks and studies from Release 18, Rel-19 delivers normative specifications for air-interface use cases, core-network analytics enhancements, model lifecycle management, federated learning support, and application enablement. This creates more intelligent, automated, and efficient 5G-Advanced systems while laying groundwork for AI-native 6G designs. Work spans RAN (air interface and NG-RAN), SA (architecture, requirements, management), and CT groups. Specs form part of the frozen Rel-19 package (implementable by end-2025).

Two complementary directions define the effort: optimizing network performance via AI/ML models, and providing robust transport, exposure, training, and inference support for AI workloads (including third-party applications).

AI for the Network – Air Interface and RAN

AI/ML for NR Air Interface (NR_AIML_air)

Rel-19 standardizes a normative framework for one-sided AI/ML models (located at the gNB or UE) with a generic lifecycle-management structure covering model identification, data collection, transfer/delivery, capability reporting, and inference.

Primary use cases include:

  • Beam management — Prediction of downlink transmit beams in time and spatial domains to reduce overhead, latency, and improve selection accuracy (building on Rel-18 study results for both UE-side and network-side models).
  • Positioning accuracy enhancements — ML-based location derivation to improve precision beyond traditional measurement-based methods.
  • CSI feedback enhancement / prediction — Device-side models that predict or compress channel state information, allowing the network scheduler to operate on forecasts rather than instantaneous snapshots. Capability signaling includes AI processing budgets (e.g., CPU pools per component carrier or shared across carriers).

Two-sided CSI compression and certain UE training-data collection aspects were deferred to Rel-20. Evaluations in the underlying study (TR 38.843) assessed performance, complexity, generalization, and overhead trade-offs. Protocol aspects cover signaling for model lifecycle, meta-information, and testability/interoperability considerations.

Enhancements for AI/ML for NG-RAN

Extends Rel-18 work with new use cases and refined procedures:

  • Network slicing (AI/ML-based resource allocation, parameter prediction, dynamic deployment, and automated management).
  • Coverage and capacity optimization (CCO) — prediction of issues, strategy exchange between nodes, and AI-driven cell shaping.
  • Improved data collection, prediction, and coordination across split architectures, NR dual connectivity mobility, and continuous Minimization of Drive Tests (MDT) for end-to-end intelligent RAN optimization.

Related RAN2 studies address AI/ML for mobility (e.g., measurement prediction based on historical data to reduce overhead and power consumption).

Core Network Enhanced Support for AI/ML

NWDAF (Network Data Analytics Function) receives significant enhancements, incorporating Model Training Logical Function (MTLF) capabilities:

  • AI/ML-based positioning — Location Management Function (LMF) derives UE position using ML models (trained by LMF or NWDAF) rather than solely on traditional signaling measurements.
  • Vertical Federated Learning (VFL) — Enables collaborative model training across entities with different feature spaces for the same samples, without sharing raw data. Each participant retains model ownership. New service APIs on NWDAF, NEF, and AF support model preparation, training, and inference.
  • NWDAF-assisted policy control and QoS management — Provides analytics and predictions (e.g., QoS parameter forecasts) to the Policy Control Function (PCF) for finer, more intelligent policy and QoS tuning per UE or group.
  • Support for detecting and mitigating abnormal or excessive signaling (signaling storm prevention).
  • Broader analytics exposure and collaboration with third-party AI/ML applications.

These capabilities improve automation, positioning accuracy, privacy-preserving training, and network performance optimization.

Network for AI – Model Transfer, Enablement, and Management

AI/ML Model Transfer Phase 2

Adds requirements for operator-controlled UE participation in AI/ML tasks, including:

  • Sidelink and network-assisted operation.
  • Service continuity under poor coverage.
  • QoS assurance for model-related traffic.
  • UE selection criteria and charging support.

This facilitates efficient distribution, updating, and execution of models between network and devices (or among devices).

Application Enablement and Media Aspects

Architecture and procedures (e.g., AIMLE / SEAL-related services) expose AI/ML capabilities uniformly to verticals and application service providers. Support covers split inference pipelines, model composition, data channel integration (including IMS), and media-related AI/ML use cases such as object recognition, video quality enhancement, and real-time processing.

AI/ML Management Phase 2

Enhances lifecycle management across the 5GS (management system, 5GC, NG-RAN). Focus areas include federated and vertical federated learning, reinforcement learning, distributed training, knowledge transfer, pre-training/fine-tuning, monitoring of training data statistics, model confidence/explainability, and energy-efficient training practices. Builds on Rel-18 foundational management services.

Additional elements include protocols for AI data collection from the User Plane Function (UPF) and consistency alignment studies across 3GPP domains.

Practical Implications and Outlook

Rel-19 AI/ML features enable more proactive, predictive, and automated network behavior—reducing overhead (CSI, beam management), improving accuracy (positioning, QoS), and supporting privacy-preserving multi-party training. Device-side models introduce a “compute contract” via capability signaling, allowing the network to configure inference within UE resource budgets.

Challenges remain around model generalization across vendors and environments, data quality/collection, testability, energy cost of training/inference, and ensuring interoperability. Many aspects (especially two-sided models and further mobility intelligence) continue evolving into Rel-20.

These capabilities position 5G-Advanced as an intelligent platform ready for AI-driven services and form a direct foundation for AI-native 6G architectures. Primary references include the AI/ML sections of TR 21.919, the NR_AIML_air and related work items, TR 38.843 (air-interface study), NWDAF/TS 23.288 enhancements, management specifications (e.g., TS 28.105 lineage), and associated stage-2/3 documents for model transfer and enablement.

In operation, operators can deploy AI/ML for smarter resource use, better user experience, and efficient support of emerging AI applications, while vendors implement the standardized frameworks for multi-vendor interoperability.

4) MIMO and Radio Enhancements

MIMO and Radio Enhancements in 3GPP Release 19 continue the evolution of massive MIMO, beam management, mobility, duplex operation, and multi-carrier techniques to deliver higher spectral efficiency, lower latency, better coverage, and improved uplink performance in commercial 5G-Advanced deployments.

These refinements address limitations observed in earlier releases (Rel-15 through Rel-18) regarding signaling overhead, latency, support for larger antenna arrays, multi-TRP operation under non-ideal conditions, and device capabilities. The work focuses on practical gains for dense urban, high-capacity, and heterogeneous network scenarios while maintaining backward compatibility. Core items include NR MIMO Phase 5, mobility enhancements (Phase 4), sub-band full duplex (SBFD) evolution, multi-carrier improvements, high-power UE support, and related RF/RRM/demodulation refinements. Specs form part of the frozen Rel-19 package.

NR MIMO Phase 5

This is the flagship MIMO work item, introducing targeted enhancements where prior releases showed constraints in latency, overhead, spectral efficiency, and coverage.

UE-initiated, event-driven beam reporting

UEs, which often have more timely awareness of beam quality changes, can now trigger reports rather than relying solely on network-scheduled periodic or aperiodic reporting.

  • Two modes: DCI-scheduled PUSCH after a one-bit UE-initiated indication, or configured-grant PUSCH triggered by the same indication.
  • Multiple event types support rapid degradation detection, timely beam switching, and accelerated Transmission Configuration Indicator (TCI) state updates.
  • Cross-carrier reporting is supported to aid multi-carrier deployments. This reduces overhead and latency compared with always-on or network-driven reporting.

CSI support extended to 128 ports

Prior PMI-based CSI reporting was limited to 32 CSI-RS ports. Rel-19 extends this to 128 ports for both fully digital and hybrid MIMO architectures.

  • Aggregation of multiple non-zero-power (NZP) CSI-RS resources.
  • Enhancements to Type-I and Type-II codebooks.
  • Extensions to CSI Resource Indicator (CRI)-based reporting allowing multiple CSI hypotheses.
  • SRS port grouping to reduce UE complexity for reciprocity-based CSI acquisition in TDD higher-layer operation.

These changes better match state-of-the-art massive MIMO antenna configurations and improve multi-user MIMO capacity and accuracy.

Support for coherent joint transmission (CJT) under non-ideal conditions

Aperiodic CSI reports assist network calibration across multiple Transmission Reception Points (TRPs), covering inter-TRP delay, frequency, and DL/UL phase offsets. This enables more robust multi-TRP operation when backhaul is not ideal.

Uplink enhancements for 3TX UEs

Support for UEs with three transmit antennas:

  • Codebook-based and non-codebook-based 3TX PUSCH transmission.
  • Corresponding updates to SRS, DMRS, and PTRS.
  • New UE capability indicating maximum PUSCH rank of three.
  • Optimized DCI signaling.

This improves uplink spectral efficiency and throughput for devices with more advanced antenna configurations.

Additional MIMO-related aspects include asymmetric DL single-TRP / UL multi-TRP configurations with enhanced power control and path-loss adjustments, suitable for heterogeneous networks (e.g., macro DL with micro TRP UL).

Mobility Enhancements (Phase 4)

Building on Rel-18 L1/L2-Triggered Mobility (LTM), which reduced interruption times within the same gNB Central Unit (CU):

  • Extension of LTM to inter-CU handovers (cells associated with different CUs).
  • Support for dual connectivity scenarios during inter-CU LTM (maintaining connection with the other cell group).
  • Event-triggered L1 measurement reporting (e.g., LTM3/LTM4 events analogous to legacy A3/A4) to reduce overhead versus periodic reporting.
  • Extension of L1 measurements to include CSI-RS (in addition to SSB) for better target-cell throughput immediately after cell switch.
  • Conditional LTM support.

These changes make lower-layer mobility more ubiquitous, further reducing interruption times and signaling load across the network.

Evolution of NR Duplex Operation: Sub-Band Full Duplex (SBFD)

SBFD allows the gNB to perform simultaneous downlink transmission and uplink reception on non-overlapping frequency sub-bands within a conventional TDD carrier.

  • Indication of time and frequency locations of SBFD sub-bands to UEs.
  • Associated transmission, reception, measurement, and random-access procedures (primarily 4-step RACH, with support for various triggers).
  • Cross-link interference management considerations.

Benefits include extended effective uplink duration within TDD frames, improved uplink coverage and capacity, and reduced latency, while remaining compatible with legacy TDD operation.

Multi-Carrier and Related Radio Enhancements

  • Multi-carrier enhancements Phase 3 (further refinements to CA operation, including low-band aggregation via switching).
  • Support for additional channel bandwidths, CA/DC combinations (including more flexible x DL / y UL configurations and SUL), and simultaneous Rx/Tx band combinations.
  • NR RRM Phase 5 and demodulation performance Phase 5 (updated requirements for measurements, mobility, and link performance).
  • UE RF enhancements Phase 4 for FR1/FR2 and EN-DC (power handling, spatial aspects, high-power UE operation in CA/DC scenarios, including Power Class 1.5/2 refinements).
  • FR1 TRP/TRS and MIMO OTA testing enhancements.
  • Support for intra-band non-collocated EN-DC/NR-CA with new receiver types.

High-power UE (HPUE) capabilities are expanded for various CA and DC band combinations to improve uplink coverage, particularly in FWA or vehicle-mounted scenarios.

Practical Impact and Outlook

These enhancements deliver tangible gains in spectral efficiency (via larger CSI port support and better MU-MIMO), reduced beam and mobility latency/overhead, improved uplink performance (3TX and SBFD), and more robust multi-TRP/CJT operation. Operators can achieve higher capacity in dense deployments and better user experience under mobility, while device vendors gain clearer requirements for advanced antenna and power configurations.

Challenges include managing complexity of higher-port CSI, ensuring calibration accuracy for non-ideal multi-TRP, and balancing SBFD gains against interference. Many elements (especially AI-assisted beam management and further mobility intelligence) interact with the Rel-19 AI/ML work.

Primary references include the NR MIMO Phase 5 summary in TR 21.919, related RAN1/RAN2/RAN4 work items (e.g., NR_MIMO_Ph5, NR_Mob_Ph4, NR_duplex_evo), TS 38.214/38.331/38.321 updates, and supporting RF/RRM specifications.

Overall, Rel-19 MIMO and radio enhancements refine the physical-layer foundation of 5G-Advanced for higher performance and prepare key techniques (larger arrays, advanced duplex, lower-layer mobility) that will continue evolving toward 6G.

5) XR / Immersive Services and Edge Computing

XR / Immersive Services and Edge Computing in 3GPP Release 19 significantly advance support for Extended Reality (XR), Augmented Reality (AR), Virtual Reality (VR), Metaverse applications, and related edge offloading, targeting the strict latency, capacity, power, and reliability demands of immersive experiences.

These features build on Rel-17/18 foundations (XR awareness, PDU Set QoS, L4S, initial power-saving and split-rendering concepts) by refining RAN handling of XR traffic, enabling richer media capabilities (avatars, split rendering over IMS), and strengthening edge computing architecture for low-latency processing close to the user. The goal is higher XR capacity, reduced device power consumption, better quality of experience (QoE), and practical support for glass-type AR/MR devices, industrial/enterprise use cases, and multi-user immersive services. Specs form part of the frozen Rel-19 package.

XR for NR Phase 3 (RAN Enhancements)

The core RAN work item (NR_XR_Ph3) focuses on improved handling of XR traffic characteristics while reducing device power consumption. Key enhancements include:

  • Capacity and QoS-oriented improvements — Better scheduling using packet delay information, optimized rate adaptation on uplink and downlink, and more efficient resource allocation for periodic, latency-sensitive XR flows.
  • Cancellable measurement gaps — Network can dynamically cancel measurement gaps via DCI so that data transmission continues in otherwise unused gaps, minimizing throughput impact on continuous XR streams.
  • Optimized RLC/PDCP handling — Refined protocol behavior tailored to XR traffic patterns (e.g., PDU Sets, data bursts).
  • Enriched network interfaces and tighter 5GC coordination — Improved signaling of XR application information (periodicity, QoS needs) from core to RAN, enabling more precise awareness and control.
  • Power-saving refinements — Building on Rel-18 DRX and related mechanisms, with further adaptations for non-integer XR traffic periodicity and reduced impact of retransmissions or monitoring.

These changes increase the number of simultaneous XR users a cell can support and help maintain high data rates and low latency under mobility or measurement conditions.

Immersive Media, Avatars, and Split Rendering

Media capabilities and device categorization

TS 26.119 defines media capabilities and an XR baseline terminal architecture for AR-capable devices, consolidating categories (e.g., thin AR glasses, fuller AR glasses, XR phones) to improve interoperability among applications, content creators, and manufacturers. It covers decoding, rendering pipelines, pose formats, sensors, and system functions.

Avatar Communications in AR Calls

Extends IMS-based AR real-time communication (TS 26.264) with avatar call functionality:

  • Support for 2D/3D avatars.
  • Data-channel–based animation streams.
  • SDP negotiation and flexible rendering activation (receiver-driven, sender-driven, or network-driven scenarios).
  • Base avatar management via a Base Avatar Repository service.
  • Interoperability with XR runtimes (e.g., OpenXR).

Focus starts with 1-to-1 communications, enabling more natural and immersive AR calls.

Split Rendering over IMS

A major Rel-19 advance (TS 26.565 / related IMS work):

  • Architecture, codecs, metadata, metrics, and end-to-end procedures for distributing XR rendering between the UE (e.g., AR glasses) and network (edge/cloud).
  • Pose correction (e.g., Asynchronous Time Warping) remains on-device to preserve motion-to-photon latency, while heavy graphics workload is offloaded.
  • Session establishment, modification, adaptation, and QoE monitoring (uplink and downlink).
  • Use cases span industrial (monitoring, maintenance, teleoperation), enterprise, education, cloud gaming, and shared collaborative XR experiences.
  • Utilizes IMS data channel with “3gpp-sr” sub-protocol and standardized pose formats.

This reduces computational and power burden on constrained devices while meeting strict latency targets (pose-to-render-to-photon often targeted in the 10–50 ms range depending on configuration).

Other media/immersive aspects

  • Extended Reality and Media Phase 2 refinements (more adaptive PDU Set–based QoS, support for encrypted/multiplexed flows, dynamic traffic control).
  • Localized Mobile Metaverse Services (spatial anchors/mapping, precise localization, digital assets, multi-user communication, security/charging).
  • Studies and support for spatial computing, glass-type AR/MR devices, and immersive voice/audio (IVAS) enhancements.

Edge Computing Enhancements (Phase 3)

Edge computing is essential for XR/immersive services because it places compute resources close to the user, enabling low-latency split rendering, spatial computing, and local processing. Rel-19 Phase 3 advances include:

  • Enhancement of support for Edge Computing in 5G Core (eEDGE_5GC_Ph3) — Lightweight local offloading, reduced control-plane impact, and latency-aware selection of local User Plane Functions (UPFs) and edge application resources (e.g., via N6 delay mechanisms).
  • Enabling Edge Applications Phase 3 (EDGEAPP_Ph3) — Improved Edge Enabler Layer (EEL) for better discovery, selection, and service continuity of common/bundled Edge Application Servers (EAS), including federation, roaming, and Edge Network Services scenarios.
  • Edge Computing for Industrial Scenarios — Clarifications for mobile robot privacy/security and digital-twin quality inspection with local edge processing.

These enable more reliable offloading of XR workloads (rendering, perception, spatial mapping) while supporting service continuity and multi-access edge scenarios.

Practical Implications and Outlook

Rel-19 XR and edge features make high-quality immersive experiences more feasible on power- and form-factor-constrained devices (especially AR glasses) by combining RAN traffic optimization, intelligent offloading, and standardized media pipelines. Operators gain higher XR capacity and better QoE tools; device and application developers benefit from clearer capability frameworks, split-rendering APIs, and avatar support.

Challenges remain around end-to-end latency under mobility, multi-user synchronization, security of spatial data, and balancing offload gains against network load. Many elements interact with Rel-19 AI/ML (for predictive rendering or traffic management) and energy-saving features.

Further evolution is expected in Rel-20 and beyond as Metaverse and spatial computing mature. Primary references include the XR/Metaverse/Edge sections of TR 21.919, NR_XR_Ph3 work item, TS 26.119 (AR media capabilities), TS 26.264 (IMS AR), TS 26.565 (split rendering), edge-related SA2/SA6 specifications, and supporting RAN protocol updates.

In summary, Release 19 delivers a more complete toolkit for delivering responsive, high-fidelity XR and immersive services over 5G-Advanced networks by tightly coupling radio optimizations, media standardization, and edge compute capabilities.

6) IoT, RedCap, and Ambient IoT

IoT, RedCap, and Ambient IoT in 3GPP Release 19 expand the cellular IoT ecosystem toward lower complexity, lower power, and new ultra-constrained device classes, enabling broader industrial, logistics, inventory, and sensing applications.

Release 19 builds on earlier IoT foundations (NB-IoT, eMTC/LTE-M, RedCap from Rel-17, and eRedCap from Rel-18) by refining RedCap capabilities and introducing Ambient power-enabled IoT (A-IoT or Ambient IoT) as a new ultra-low-power category. These advances target massive connectivity, maintenance-free operation, and use cases that existing LPWA technologies cannot fully address due to power, cost, or size constraints. Related work also covers NTN support for RedCap devices and management aspects. Specs form part of the frozen Rel-19 package.

RedCap and Enhanced RedCap Refinements

RedCap (Reduced Capability) devices, introduced in Rel-17, provide a mid-tier IoT option with reduced bandwidth, fewer antennas, half-duplex operation in some cases, and peak rates generally below 250 Mbps—positioned between full NR devices and classic LPWA (NB-IoT/eMTC). eRedCap (Rel-18) further caps peak rates at 10 Mbps for even simpler, lower-cost devices.

Rel-19 focuses on practical refinements and broader deployment support:

  • Power Class 2 RedCap UE in FR1 — Extends higher uplink power capabilities for RedCap devices in Frequency Range 1, improving coverage in challenging environments.
  • Management aspects of RedCap features — Enhanced OAM and network management for RedCap/eRedCap devices, including configuration, monitoring, and optimization.
  • NAS layer overhead reduction for data transfer using CP CIoT — Further efficiency gains for control-plane optimized small data transfers.
  • NTN support — RedCap and enhanced RedCap devices gain better support for operation over Non-Terrestrial Networks (FR1 NTN bands), including adaptations for timing advance, half-duplex constraints, and coverage, enabling mid-rate IoT services in remote or satellite-covered areas.

These changes make RedCap more versatile for wearables, industrial sensors, smart meters, surveillance cameras, and similar use cases that need more capability than NB-IoT but less complexity and cost than full 5G smartphones.

Ambient Power-Enabled IoT (Ambient IoT / A-IoT)

This is a major new feature area in Rel-19, targeting devices that are battery-less or have only limited energy storage (e.g., capacitors) and harvest ambient energy from radio waves, light, motion, heat, or other sources. Ambient IoT devices offer complexity and power consumption orders of magnitude lower than existing 3GPP LPWA technologies (peak power around ~1 µW for the baseline “Device 1”), enabling maintenance-free, long-lifespan (e.g., >10 years), small-form-factor deployments that were previously impractical.

Key characteristics of Rel-19 Ambient IoT Device 1

  • Ultra-low peak power (~1 µW).
  • Energy storage (or battery-less).
  • Simple RF envelope detector receiver (non-coherent, low complexity).
  • High initial sampling frequency offset tolerance (up to 10⁵ ppm).
  • No active amplification for Reader-to-Device (R2D) or Device-to-Reader (D2R) links in the baseline design.
  • D2R transmissions typically use passive backscatter modulation of an external RF carrier (no active transmitter power amplifier).
  • Very low cost, wide manufacturing tolerances, and small size.

Representative use cases (from TR 38.848 and related studies)

Focus on indoor scenarios in Rel-19:

  • Indoor inventory (e.g., asset tracking, stock counting).
  • Indoor command (simple actuation or control).

Broader potential includes sensors, tracking, and actuators in logistics, warehousing, smart agriculture, and industrial monitoring.

Radio and architecture support

  • RAN aspects — Simplified PHY/MAC interface integrated into the 5G system. Communication occurs between A-IoT devices and an A-IoT reader (typically a gNB-reader). Supports inventory and command procedures. Devices monitor R2D messages when energy is available.
  • System architecture (TS 23.369 and related) — Defined primarily for isolated private networks (e.g., SNPN) in Rel-19, without interaction with public networks. Functions include:
    • A-IoT device identification and discovery/inventory in a given area.
    • Exchange of application data despite intermittent, energy-constrained operation.
    • Device disablement.
    • Exposure, charging, security, and privacy considerations.
  • Service requirements (TS 22.369) cover communication, positioning, management, and performance KPIs for inventory, sensing, tracking, and actuation.

Ambient IoT addresses the extreme low end of the IoT market that NB-IoT, eMTC, and RedCap cannot efficiently serve due to residual power and complexity requirements.

Broader IoT Context and Related Enhancements

  • Continued evolution of classic LPWA (NB-IoT/eMTC) features, including IoT-NTN Phase 3 aspects (Store-and-Forward, uplink capacity, TDD mode, public warning).
  • Integration with other Rel-19 themes such as energy efficiency (LP-WUS benefits low-power devices), NTN coverage expansion, and management/OAM for massive IoT.
  • Security, charging, and privacy adaptations tailored to constrained and energy-harvesting devices.

Practical Implications and Outlook

Rel-19 strengthens the cellular IoT portfolio across a wider capability spectrum:

  • RedCap/eRedCap for mid-tier, cost- and power-optimized devices with moderate data rates and better coverage options (including NTN).
  • Ambient IoT for ultra-constrained, battery-free or near-battery-free massive deployments.

Operators and verticals can deploy denser, lower-maintenance IoT networks for logistics, inventory, industrial sensing, and remote monitoring. Device manufacturers gain standardized ultra-low-power designs that reduce cost and enable new form factors. Challenges include energy availability variability, limited range/coverage of backscatter links, security of extremely simple devices, and ensuring reliable intermittent connectivity.

Further enhancements (additional device types, outdoor scenarios, public network integration, richer positioning) are expected in subsequent releases. Primary references include the IoT/RedCap/Ambient IoT sections of TR 21.919, TS 22.369 (service requirements), TS 23.369 (architecture), TS 38.300 (RAN support), related RAN1 work items on Ambient IoT solutions, RedCap RF/management specifications, and supporting studies (e.g., TR 38.848).

In summary, Release 19 makes cellular connectivity practical for an even broader range of IoT devices—from simplified mid-tier RedCap terminals to truly ambient, energy-harvesting sensors—supporting the continued expansion of massive, sustainable IoT deployments within the 5G-Advanced framework.

7) Sidelink enhancements in 3GPP Release 19

Sidelink enhancements in 3GPP Release 19 focus on multi-hop relaying (both UE-to-Network and UE-to-UE), Proximity-based Services (ProSe) Phase 3, and ranging/sidelink positioning, significantly extending coverage, resilience, and location capabilities for public safety, V2X, industrial, and commercial use cases.

These features build on single-hop sidelink relays introduced in Rel-17/18 and earlier ProSe foundations. They enable devices to communicate or access the network via intermediate relay UEs when direct links or network coverage are unavailable, while also advancing peer-to-peer ranging and positioning over the PC5 interface. Specs form part of the frozen Rel-19 package and support both in-coverage and out-of-coverage operation.

Multi-Hop UE-to-Network (U2N) and UE-to-UE (U2U) Relays

NR Sidelink Multi-Hop Relay (NR_SL_relay_multihop)

Rel-17 and Rel-18 supported single-hop Layer-2 (L2) and Layer-3 (L3) U2N and U2U relays. Rel-19 extends this to multi-hop operation, introducing up to two additional relay hops (practical limits configurable by the network, often based on service requirements).

  • U2N multi-hop — A remote UE connects to the network via one or more intermediate relay UEs. Only the final (or first, depending on direction) relay needs to be in network coverage and establish a Uu connection; intermediate relays can be in- or out-of-coverage.
  • U2U multi-hop — End-to-end communication between two remote UEs via one or more intermediate relays, useful for extending sidelink coverage in off-network scenarios.

Key architectural and procedural aspects

  • Support for both L2 (AS-layer aware, with SRAP-like forwarding) and L3 (higher-layer transparent) architectures.
  • Discovery (Model A and Model B) adapted for multi-hop: Model A allows each UE to know only its next hop; Model B can provide end-to-end path knowledge.
  • Relay selection/reselection based on PC5 measurements (SL-RSRP / SD-RSRP) plus higher-layer criteria.
  • Connection establishment, path switching (direct ↔ indirect, single-hop ↔ multi-hop), and service continuity procedures.
  • QoS-aware operation, security (key management adapted for multi-hop), and charging support.
  • Hop limits can be configured per Relay Service Code (RSC) or service to balance latency versus coverage.

These capabilities improve coverage extension, resilience in challenging environments (e.g., disaster scenarios, dense urban canyons, or remote areas), and energy efficiency by allowing lower-power links over shorter hops.

UE-to-UE Multi-Hop Relay Requirements

Stage-1 requirements emphasize off-network device-to-device communications using multiple relays to extend coverage, supporting commercial, public-safety, and V2X needs.

Proximity-based Services in 5GS Phase 3 (5G_ProSe_Ph3)

ProSe Phase 3 integrates and extends the multi-hop relay capabilities into the 5G System architecture:

  • Secure, QoS-aware multi-hop U2N and U2U relaying.
  • Defined discovery procedures, security frameworks (including key management for multi-hop), and charging support.
  • Support for intermediate U2N relays.
  • Enhancements for Non-Public Networks (NPN) and related vertical scenarios.
  • Alignment with NR sidelink multi-hop mechanisms for consistent end-to-end behavior.

ProSe continues to support direct discovery and communication (unicast, groupcast, broadcast) both in- and out-of-coverage, now with robust multi-hop options.

Ranging and Sidelink Positioning

Rel-19 advances ranging-based services and sidelink (SL) positioning over PC5, enabling absolute location, relative position, or range/direction estimation between UEs.

Operational modes

  • Network-assisted.
  • Network-based (involving LMF and 5GC).
  • UE-only (no network involvement).

Key capabilities

  • Use of Located UEs (anchors with known positions) to determine the location of a Target UE via range and/or direction measurements.
  • Ranging/SL Positioning Protocol (RSPP) transported over PC5 (reusing V2X or 5G ProSe procedures).
  • Support for group and unicast sessions.
  • Discovery of suitable reference/located UEs.
  • Integration with multi-hop relays (e.g., when direct PC5 links are unavailable).
  • Charging aspects for ranging and SL positioning services (including 5GC-assisted and exposure scenarios).

Use cases and performance Public safety, industrial IoT, indoor commercial applications, V2X, and scenarios requiring high-accuracy relative positioning. Rel-19 expands positioning service levels with tighter accuracy and latency targets compared with prior releases. Multi-hop SL positioning can improve Line-of-Sight probability and energy efficiency by using intermediate anchors.

Additional Related Enhancements

  • NR Sidelink Intra-band CA in ITS band — RF and RRM requirements for contiguous and non-contiguous carrier aggregation in Intelligent Transport Systems spectrum, including Power Class 2/3 support to improve bandwidth, coverage, and capacity while meeting regional regulations.
  • Security adaptations for multi-hop and ranging (key management, authorization).
  • Interactions with other Rel-19 features such as energy efficiency, AI/ML (for smarter relay selection or positioning), and mission-critical services.

Practical Implications and Outlook

Multi-hop sidelink relays dramatically expand the reach of direct device communication and network access without requiring dense infrastructure, making them especially valuable for public safety, disaster response, industrial campuses, and vehicular networks. Ranging and SL positioning complement cellular positioning by providing high-accuracy relative and absolute location in coverage-challenged or peer-centric scenarios.

Challenges include managing end-to-end latency and QoS across multiple hops, efficient relay selection and discovery under mobility, security in multi-hop chains, and power consumption of intermediate relays.

These Rel-19 capabilities provide a mature foundation for commercial and public-safety sidelink deployments and will continue evolving (e.g., further multi-hop optimization and AI-assisted positioning) in subsequent releases. Primary references include the Sidelink/Proximity sections of TR 21.919, work items NR_SL_relay_multihop and 5G_ProSe_Ph3, TS 38.300 (sidelink and relay architecture), TS 23.304 (ProSe Stage 2), TS 23.586 (ranging/SL positioning architecture), and related security and charging specifications.

In summary, Release 19 transforms sidelink from primarily single-hop direct communication into a flexible multi-hop fabric that extends coverage, improves resilience, and delivers advanced proximity and positioning services within the 5G-Advanced system.

8) Mission-critical / public-safety / FRMCS enhancements

Mission-critical / public-safety / FRMCS enhancements in 3GPP Release 19 mature the 5G-based framework for reliable, prioritized voice, data, and video communications tailored to public safety agencies, railways, utilities, and other critical sectors.

These features build on the Mission Critical Services (MCS / MCX) suite—MCPTT (Push-to-Talk), MCData, and MCVideo—introduced and refined across earlier releases, while specifically advancing the Future Railway Mobile Communication System (FRMCS) as the standardized 5G successor to GSM-R. Rel-19 completes key stages of FRMCS Phase 5 (FRMCS_Ph5), enabling the first wave of commercial and operational deployments. Specs form part of the frozen Rel-19 package and integrate with sidelink/ProSe multi-hop, NTN, and other Rel-19 capabilities for resilience in challenging environments.

Mission Critical Services (MCX) Core Enhancements

The MCX framework provides a common architecture (TS 23.280 and related) for on-network and off-network operation, supporting group and private calls, prioritization, pre-emption, floor control, and multimedia services. Rel-19 refinements focus on operational robustness for public safety and critical users:

  • Improved support for temporary disable/re-enable of MC UEs (useful for operational control by supervisors).
  • Logging and recording enhancements.
  • Ambient listening pause/resume capabilities.
  • Stronger integration with ProSe multi-hop relays for extended coverage in off-network or partial-coverage scenarios (public safety often relies on sidelink when infrastructure is unavailable).
  • Security, identity, and interconnection enhancements between MC systems.
  • QoS and priority mechanisms that align with 5QI values optimized for mission-critical delay-sensitive signaling and media (e.g., 5QI 69 for signaling).

These ensure secure, prioritized communications even under congestion or emergency conditions, with support for both commercial and dedicated public-safety deployments.

FRMCS-Specific Enhancements (FRMCS Phase 5)

FRMCS is the UIC-driven, 5G-based evolution of GSM-R, designed for higher data rates, lower latency, multimedia support, and digitalization of rail operations (train control, safety, shunting, trackside maintenance, passenger services). It reuses and extends the 3GPP MCX framework while adding railway-specific requirements (TS 22.289 and related).

Key Rel-19 advances (completion of Stage 3 for FRMCS_Ph5)

  • Advanced gateway capabilities for interworking and migration.
  • More responsive ad-hoc group communications and combining of ad-hoc group calls with emergency alerts.
  • Unified location protocols to improve coordination across networks and borders.
  • Enhanced interworking with GSM-R (group calls, point-to-point calls, floor control mapping, late entry, etc.) to support smooth migration.
  • Functional addressing / functional aliases refined for role-based (rather than device-based) communications.
  • Call forwarding and related private-call features carried forward and improved from prior phases.
  • Support for multipath connectivity (leveraging multiple UEs or networks for resilience and capacity).
  • QoS frameworks tailored to railway performance requirements (very low latency and ultra-high reliability for critical data/video at speeds up to 500 km/h).

Architecture principles

FRMCS is structured in strata:

  • Transport Stratum — 5G connectivity (Uu, possibly NTN or shared access), mobility, QoS, policy, authentication.
  • Service Stratum — Centered on 3GPP MCX (MCPTT, MCData, MCVideo) with tight- or loose-coupled modes for applications.
  • On-board and trackside gateways, reference points (e.g., OBAPP, TSAPP), and interfaces for applications.

Performance targets (from TS 22.289) include end-to-end latency as low as ≤10–100 ms, reliability up to 99.9999%, and support for high-speed scenarios along rail tracks.

Public Safety and Broader Critical Communications

  • Public emergency call and warning broadcast integration (including location-based targeting).
  • Assured voice communication and safety-related critical advisory messaging.
  • Support for interworking with Land Mobile Radio (LMR) systems.
  • Alignment with ProSe Phase 3 and multi-hop sidelink for off-network public-safety operations.
  • Potential leverage of Rel-19 NTN enhancements for coverage in remote or disaster areas.

These features serve police, fire, emergency medical, utilities, maritime, and other sectors that require guaranteed priority and reliability.

Practical Implications and Outlook

Rel-19 delivers a mature, fully standardized FRMCS feature set, shifting industry focus from specification to deployment and interoperability testing (e.g., ETSI FRMCS Plugtests). Rail operators gain a clear migration path from GSM-R, with support for coexistence, multipath resilience, and higher-capacity applications (video, sensors, automation). Public-safety agencies benefit from more robust MCX services that integrate seamlessly with 5G networks, sidelink relays, and edge capabilities.

Challenges include spectrum coordination (e.g., dedicated 1900 MHz railway bands vs. shared spectrum), cross-border interoperability, security in multi-hop and multipath scenarios, and ensuring ultra-reliable performance under high mobility. Further enhancements (e.g., deeper NTN integration, additional FRMCS phases) continue in Rel-20 and beyond.

Primary references include the relevant sections of TR 21.919, FRMCS-related work items (FRMCS_Ph5 lineage), TS 22.289 (railway requirements), TS 23.280 / TS 23.379 / TS 23.281 / TS 23.282 (MCX architecture), TS 23.289 (MCS over 5GS), UIC FRMCS specifications, and ETSI RT documents defining transport/service strata and on-board interfaces.

In summary, Release 19 solidifies 5G-Advanced as a production-ready platform for mission-critical and public-safety communications, with FRMCS providing the standardized foundation for the next generation of railway operations.

9) Network slicing, Service-Based Architecture (SBA)/protocol improvements, QoS/policy enhancements, and multi-access (ATSSS Phase 4)

Network slicing, Service-Based Architecture (SBA)/protocol improvements, QoS/policy enhancements, and multi-access (ATSSS Phase 4) in 3GPP Release 19 refine the 5G core and system architecture for more flexible, efficient, secure, and multi-path connectivity.

These core-network focused features improve isolation and customization of services (slicing), streamline inter-NF communications and security (SBA), provide finer-grained traffic treatment (QoS/policy), and enable intelligent use of multiple access types simultaneously (ATSSS). They support diverse verticals, media services, edge computing, and resilient multi-access scenarios while maintaining backward compatibility. Specs form part of the frozen Rel-19 package.

Network Slicing Enhancements

Network slicing continues to evolve as a foundational 5G capability, allowing logical networks with tailored isolation, resources, and characteristics for different services or tenants.

Key Rel-19 aspects

  • Network Slice Capability Exposure for Application Layer Enablement (NSCE) — Procedures for registration, slice API configuration/translation, application-layer lifecycle management, optimization based on VAL (Vertical Application Layer) server policy, performance/analytics monitoring, predictive modification (including edge and inter-PLMN continuity), coordinated multi-slice resource optimization, diagnostics, fault management, and requirements verification.
  • Support for media services and 5G Media Streaming (5GMS) with network slicing (provisioning, dynamic policy, slice-specific Application Servers, traffic migration between slices).
  • Integration with AI/ML for slicing-related use cases (e.g., coverage/capacity optimization).
  • Management and orchestration refinements (concepts, use cases, NRM support for slice and slice subnet, GST/NEST attributes).
  • RAN awareness and enforcement (differentiated handling, resource isolation via RRM policies, Slice-MBR, QoS differentiation within a slice).
  • Charging, authentication/authorization (NSSAA), and wholesale/roaming aspects carried forward and refined.
  • Network-controlled slice selection enhancements.

These enable more dynamic, application-aware, and predictive slice management, better support for verticals (e.g., media, industrial), and improved exposure to external applications.

Service-Based Architecture (SBA) and Protocol Improvements

The Service-Based Architecture underpins 5GC interactions via service-based interfaces (SBIs). Rel-19 focuses on maturity, security, and efficiency:

  • UPF enhancement for Exposure and SBA Phase 2 — Improved user-plane function exposure and integration with SBA principles.
  • Automatic Certificate Management Environment (ACME) for SBA — Streamlined certificate handling for secure SBI communications.
  • Reducing Information Exposure over SBI — Privacy and security hardening by limiting unnecessary data exposure between network functions.
  • Service-Based Interface Protocol Improvements — General protocol refinements for reliability, efficiency, and interoperability of SBI signaling.
  • Related support in areas such as GBA (Generic Bootstrapping Architecture) for SBA.

These reduce operational complexity, strengthen security posture, and improve the robustness of NF interactions in large-scale and multi-vendor deployments.

QoS and Policy Enhancements

Policy and Charging Control (PCC) and QoS mechanisms receive targeted refinements for better end-to-end treatment:

  • Enhanced QoS parameter mapping (AF, PCF, SMF, MB-SMF) and signaling flows over Npcf, Nsmf, and related interfaces.
  • QoS monitoring enhancements.
  • Support for dynamic policy in conjunction with network slicing and media services.
  • Per-subscriber VLAN instructions and related control.
  • Improved handling for variable-bitrate/adaptive codecs and asymmetric sessions (studies and related protocol work).
  • Alignment with multi-access and edge scenarios for consistent policy enforcement.

These provide more precise resource allocation, better adaptation to application needs (e.g., XR, media), and tighter integration between application functions and the core.

Multi-Access: ATSSS Phase 4

Access Traffic Steering, Switching, and Splitting (ATSSS) enables a Multi-Access (MA) PDU Session that can simultaneously or selectively use 3GPP and non-3GPP accesses (or dual 3GPP in related work).

ATSSS Phase 4 (and related multi-access) focuses on

  • Expanded steering functionalities: combinations of ATSSS-LL, MPTCP, and MPQUIC variants (UDP, IP, Ethernet).
  • Associated steering modes (active-standby, smallest delay, load balancing, priority-based, redundant).
  • Upper-layer traffic steering and switching over dual 3GPP access.
  • ATSSS rule provisioning via 3GPP access connected to EPC (for interworking scenarios).
  • Simplified or enhanced operation over non-3GPP access (including cases without traditional N3IWF/TNGF in study contexts).
  • Policy enhancements for dual-steer scenarios and better coordination with URSP (UE Route Selection Policy).
  • Measurement assistance, performance reporting, and traffic distribution procedures.

An MA PDU Session can use one access at a time or both simultaneously, supporting registration to the same or different PLMNs over the respective accesses. This improves resilience, throughput aggregation, and optimized path selection (e.g., Wi-Fi offload with seamless fallback, dual 3GPP for high reliability).

Practical Implications and Outlook

These Rel-19 enhancements make the 5G system more programmable, secure, and adaptable: operators can offer highly customized slices with dynamic management and exposure; SBA becomes more production-hardened; QoS/policy better matches modern application demands; and multi-access delivers seamless, high-performance connectivity across heterogeneous accesses.

They interact strongly with other Rel-19 areas such as XR/media, edge computing, AI/ML-driven optimization, energy efficiency, and verticals/NPN. Challenges include complexity of multi-slice and multi-access policy coordination, security of increased exposure points, and ensuring consistent performance across heterogeneous networks.

Further evolution (deeper AI integration, additional multi-access scenarios, slicing for new services) continues toward Rel-20/6G. Primary references include the Network Slicing / SBA / Multi-Access sections of TR 21.919, TS 23.435 (NSCE procedures), TS 23.501/23.502 (architecture and ATSSS), TS 24.193 (ATSSS Stage 3), TS 29.513 (PCC flows and QoS mapping), management specs (TS 28.530 series), and related SBA protocol work items.

In summary, Release 19 strengthens the core foundations of 5G-Advanced for flexible service delivery, intelligent multi-path connectivity, and robust policy-driven operation across diverse deployment scenarios.

10) Verticals / Non-Public Network (NPN) security and interconnect enhancements

Verticals / Non-Public Network (NPN) security and interconnect enhancements in 3GPP Release 19 strengthen the support for private and industry-specific 5G deployments, focusing on secure isolation, management, and connectivity between NPNs, PLMNs, and external identity providers.

Non-Public Networks (NPNs) are 5G deployments intended for non-public use (e.g., factories, enterprises, campuses, industrial sites). They can be deployed as Stand-alone Non-Public Networks (SNPNs)—operated independently without relying on a PLMN—or as Public Network Integrated NPNs (PNI-NPNs)—deployed with support from a PLMN (e.g., as a dedicated slice or hosted infrastructure). Rel-19 continues the evolution of NPN capabilities for vertical industries while placing particular emphasis on security considerations and interconnect options. Specs form part of the frozen Rel-19 package.

NPN Management and Vertical Support

Management of NPNs addresses the diverse SLA, coverage, and operational needs of verticals (manufacturing, energy, healthcare, logistics, etc.):

  • Concepts, use cases, requirements, and solutions for NPN management (TS 28.557).
  • Support for both SNPN and PNI-NPN deployment models.
  • Management modes such as MNO-Vertical Managed Mode, where a mobile operator handles primary management while the vertical retains partial control.
  • Assurance of diversified SLAs (e.g., ultra-reliable low-latency for industrial control, hybrid indoor/outdoor coverage).
  • Impact on overall 5G system management, including UE-related aspects.

These enable verticals to deploy and operate customized private networks with appropriate isolation and resource guarantees.

Security Enhancements for NPNs and Verticals

Security remains a core focus for NPNs because customer premises may have different trust levels and threat surfaces than public PLMNs.

Key Rel-19 security work

  • Security for PLMN hosting a NPN (PNI-NPN) — Addresses risks arising when a PLMN hosts customer-deployed or customer-managed network functions. Solutions include proxy-based controls and enhanced mechanisms to reduce exposure and prevent vulnerabilities in one domain from affecting the other. An informative annex provides security considerations for PLMNs hosting an NPN, emphasizing the need to clearly define security borders.
  • NPN security considerations — Broader study and refinements covering isolation, credential handling, and domain protection.
  • Continued support for flexible authentication in SNPNs (EAP framework allowing methods beyond 5G AKA/EAP-AKA’, including certificate-based options with or without external Credentials Holder).
  • Privacy protections (e.g., SUCI) and standard 5G security features applied in the NPN context.
  • Further related studies (e.g., protection of N9 interfaces or specific NFs such as NWDAF in PNI-NPN scenarios) point toward Rel-20.

Overall, NPN security builds on the high baseline of 5G PLMN security while adding flexibility for private deployments and stronger domain isolation.

Interconnect of Standalone NPNs (SNPN)

Interconnect of SNPN (ISN) enables scalable, secure connectivity between SNPNs and identity providers:

  • Support for SNPN grouping.
  • Dynamic cellular hotspots for vertical services.
  • Secure mechanisms for connectivity and identity federation.

This facilitates multi-site or multi-organization vertical deployments (e.g., interconnected factory campuses or enterprise sites) without requiring full public-network dependence.

ProSe Support in NPN

Rel-19 extends Proximity-based Services (ProSe) support to NPN environments (non-roaming scenarios):

  • Enables direct UE-to-UE communication within or associated with an NPN.
  • Complements multi-hop sidelink and public-safety/mission-critical use cases in private network settings.

Related Vertical and Exposure Features

  • Enhancements to northbound interfaces, CAPIF (Common API Framework) Phase 3, and SEAL data delivery for secure exposure of network capabilities to vertical applications.
  • OAM enhancements for management service exposure to external consumers.
  • Alignment with other Rel-19 areas such as network slicing (for PNI-NPN), edge computing, and mission-critical services.

Practical Implications and Outlook

These Rel-19 features make NPNs more practical and trustworthy for industrial and enterprise verticals by improving management flexibility, enforcing clearer security boundaries (especially for hosted PNI-NPN scenarios), and enabling controlled interconnect between private networks. Operators and verticals can offer or consume private 5G services with stronger isolation guarantees and better multi-site connectivity.

Challenges include consistent security policy enforcement across domains with differing trust levels, credential and identity management in hybrid SNPN/PNI-NPN setups, and operational complexity of multi-party management. Further refinements (deeper interface protection, additional interconnect scenarios) continue in subsequent releases.

Primary references include the Verticals and Non-Public Network sections of TR 21.919, TS 28.557 (NPN management), TS 33.501 (security architecture with NPN annexes and considerations), related work items on Security for PLMN hosting a NPN and Interconnect of SNPN, and ProSe extensions for NPN.

In summary, Release 19 advances Non-Public Networks toward more secure, manageable, and interconnected deployments, better serving the stringent isolation, performance, and multi-domain needs of industrial and enterprise verticals within the 5G-Advanced system.

11) Enhancements for User Equipment (UE) in 3GPP Release 19

Enhancements for User Equipment (UE) in 3GPP Release 19 span RF performance, power classes, reduced-capability and ultra-low-power IoT devices, energy efficiency, MIMO capabilities, mobility, XR optimization, and support for non-terrestrial networks.

These improvements address commercial device evolution (higher uplink power and better MIMO), mid- and low-tier IoT (RedCap refinements and Ambient IoT), power-constrained form factors (AR glasses, wearables), and specialized scenarios (NTN, FWA, vehicle-mounted). They build on prior releases while introducing new device categories and RF requirements. Specs form part of the frozen Rel-19 package.

High-Power UE (HPUE) and RF Enhancements

Rel-19 significantly expands high-power operation for improved uplink coverage and throughput, especially in CA, DC, FWA, and vehicle-mounted scenarios:

  • Power Class 1.5 and 2 for NR FR1 (TDD/FDD single band and intra-/inter-band CA/DC combinations, with/without SUL).
  • Support for higher maximum output power (e.g., up to 29 dBm in certain configurations) with refined Maximum Power Reduction (MPR) rules that are often carrier-power-based and PSD-agnostic. Separate treatments for handheld vs. Fixed Wireless Access (FWA) devices.
  • Power Class 1 operation for fixed-wireless/vehicle-mounted (FWVM) use cases in single NR or LTE bands.
  • UE RF enhancements for NR FR1/FR2 and EN-DC Phase 4: power handling, spatial aspects (MIMO), six-antenna reference sensitivity allowances, SRS antenna switching patterns, and downlink MIMO performance up to four layers (handheld) or six layers (FWA).
  • Additional band-specific introductions (e.g., Power Class 2 and 40 MHz channel bandwidth in n28) and support for intra-band non-collocated EN-DC/NR-CA with new receiver types.
  • Dynamic Tx switching and related downlink interruption handling in multi-band combinations.

These enable stronger uplink performance at cell edge, better FWA/vehicle coverage, and more realistic multi-antenna operation.

Reduced Capability (RedCap) and Enhanced RedCap

  • NR Power Class 2 RedCap UE in FR1 — Higher uplink power for mid-tier IoT devices.
  • Management aspects of RedCap features and NAS layer overhead reduction for control-plane CIoT data transfer.
  • Enhanced support for RedCap/eRedCap in NTN (FR1 bands), including timing-advance and half-duplex adaptations for paired spectrum.

RedCap continues to target cost- and complexity-reduced devices (bandwidth, antenna, and peak-rate limitations) suitable for wearables, industrial sensors, and smart meters, now with improved coverage options.

Ambient Power-Enabled IoT (Ambient IoT / A-IoT)

A major new ultra-constrained device category:

  • Baseline “Device 1”: ~1 µW peak power, energy storage (or battery-less), RF envelope detector receiver, high sampling-frequency-offset tolerance, passive backscatter for device-to-reader (D2R) transmissions, no active amplification in the baseline design.
  • Targets indoor inventory and command use cases with complexity and power orders of magnitude below existing LPWA (NB-IoT, eMTC, RedCap).
  • Associated RF requirements, coexistence studies, architecture (reader typically a gNB), protocol support, security for isolated private networks, and charging aspects.

This enables maintenance-free, long-lifespan, small-form-factor deployments previously impractical with cellular IoT.

Energy Efficiency and Low-Power Wake-Up

  • Low-Power Wake-Up Signal / Receiver (LP-WUS/WUR) — DFT-s-OFDM based design supporting OOK and sequence detection. Applicable in idle/inactive and connected modes. Enables deep/light sleep with significant power-saving gains (especially when combined with RRM measurement relaxation or offloading). Particularly beneficial for XR and infrequent traffic.
  • Support for on-demand system information and related energy-saving mechanisms that reduce UE monitoring load.
  • XR-specific power optimizations (cancellable measurement gaps, improved rate adaptation, protocol handling) to lower device consumption while maintaining capacity and QoS.

MIMO and Radio Capability Enhancements

  • NR MIMO Phase 5 — UE-initiated event-driven beam reporting for lower latency/overhead; CSI support extended to 128 ports; 3TX PUSCH (codebook and non-codebook) with corresponding SRS/DMRS/PTRS and capability signaling; enhancements for coherent joint transmission calibration.
  • Spatial aspects in RF work items (higher-layer MIMO performance, antenna switching).
  • Capability signaling expansions for new features (e.g., AI/ML beam prediction, simultaneous CSI reporting counts).

Mobility, XR, and Specialized Device Support

  • Mobility enhancements (LTM extensions to inter-CU, event-triggered reporting, CSI-RS measurements) improve UE experience under mobility with reduced interruption.
  • XR/immersive device optimizations (traffic handling, power saving, 2Rx XR device support in certain bands).
  • NTN UE enhancements for handset-type and RedCap devices (coverage repetitions, capacity multiplexing, GNSS-related adaptations).
  • Support for multi-hop sidelink relays and ProSe in various contexts, affecting remote/relay UE behaviors.

Practical Implications and Outlook

Rel-19 UE enhancements broaden the device ecosystem: higher-power and multi-antenna smartphones/FWA/vehicle devices gain better uplink and MIMO performance; RedCap advances mid-tier IoT; Ambient IoT opens a new ultra-low-power segment; and power-saving features (LP-WUS, XR optimizations) extend battery life for constrained form factors such as AR glasses.

Device vendors benefit from clearer RF requirements, capability frameworks, and test methodologies. Challenges include managing complexity of higher-port CSI and multi-antenna RF, ensuring coexistence for Ambient IoT, and balancing power-saving gains against latency/responsiveness.

Many UE features interact with network-side enhancements (energy savings, AI/ML, NTN, XR Phase 3). Further refinements continue in later releases. Primary references include the relevant sections of TR 21.919 (High Power UEs, IoT/RedCap, Ambient IoT, MIMO, energy efficiency, XR, NTN), TS 38.101 series (RF), TS 38.300 (architecture and capabilities), TS 38.306 (UE capabilities), and associated RAN4/RAN2 work items.

In summary, Release 19 delivers a richer set of UE capabilities that improve performance, expand IoT reach to ultra-constrained devices, and enhance energy efficiency across the 5G-Advanced device landscape.

Leave a Reply