Introduction

As the telecommunications industry shifts its focus toward the sixth generation (6G), it is becoming clear that mere enhancements in capacity and throughput will no longer be sufficient. 6G must address more fundamental requirements including the native integration of vertical services into the physical layer, as conceptually illustrated in Figure 1. These services include “Ubiquitous connectivity – Non-Terrestrial Networks (NTN)”, “Massive Communication – Internet of Things (IoT)”, “Integrated Sensing and Communication (ISAC)”, and “Hyper Reliable and Low-Latency Communication & AI and Communication – AI-powered Robotics”. In 4G and 5G, vertical use cases were often considered add-ons—introduced after the initial network deployment through additional specifications, hardware, or software upgrades. This retrofit approach led to increased complexity, suboptimal performance, and higher costs for operators and decreased corresponding adoption rates.

To avoid repeating those mistakes, 6G must adopt a "vertical-integration" philosophy in its air interface design. This includes revisiting lessons learned from 5G vertical technologies such as Ultra-Reliable Low-Latency Communication (URLLC), sidelink communication (V2X), and unlicensed spectrum operation, which despite their potential, faced numerous implementation challenges and were not widely deployed.

Retrospective: Vertical Technologies in 5G and Their Challenges

URLLC: Ambitious Design, Constrained Realization

Ultra-Reliable Low-Latency Communication (URLLC) was introduced in 5G to meet the demands of mission-critical applications such as autonomous vehicles, remote robotic surgery, industrial control systems and extended reality (XR). It was formally defined in 3GPP Release 15 and enhanced in Releases 16 and 17, incorporating techniques such as mini-slot scheduling, flexible Transmission Time Intervals (TTIs), enhancements to configured grant-based uplink access, and pre-emption indication to ensure collision avoidance for coexistence with eMBB traffic. These enhancements were designed to deliver sub-millisecond latency with 99.999% reliability, a cornerstone for time-sensitive services.

URLLC deployment remains limited despite its technical promise due to configuration complexity and ecosystem immaturity. Implementing URLLC requires intricate scheduling, resource allocation, and signaling, increased overhead in multi-user scenarios, and some enhanced UE capabilities. While PHY/MAC layer guarantees exist, inconsistent E2E performance arises when transport/application layers fail to ensure concurrent latency and reliability. Industries hesitate to adopt URLLC owing to fragmented standardization that spanned multiple releases and included a large number of features, scarce chipsets, and unproven field stability, hindering its transition from theoretical potential to widespread commercial use.

Sidelink (V2X): Technically Promising, Practically Constrained by Automotive Realities

Sidelink communication was introduced in LTE-V2X (3GPP Release 14) and later extended in 5G NR (3GPP Releases 16 and 17) to enable direct vehicle-to-vehicle (V2V), vehicle-to-infrastructure (V2I), and vehicle-to-pedestrian (V2P) communication. This device-to-device (D2D) communication paradigm offers theoretically low-latency, low-overhead links ideal for dynamic traffic environments, particularly when cellular infrastructure is unavailable. Two operational modes exist: Mode 1 with centralized scheduling, and Mode 2 where vehicles autonomously select radio resources.

Sidelink commercialization remains limited due to standardization fragmentation and ecosystem immaturity. LTE-NR sidelink incompatibility at physical/protocol levels isolates vehicles across technology generations, which is particularly problematic in automotive where backward compatibility is critical. The industry’s long vehicle lifecycles (10-20 years) delay NR sidelink adoption, requiring over a decade for meaningful road penetration and discouraging near-term OEM investment. V2X success further hinges on immature coordination among infrastructure, regulators, and insurers. For sidelink to thrive, future standards must prioritize backward compatibility, long-term support, and automotive-tailored migration paths.

NR-U: Competing with Wi-Fi in an Already Saturated Market

New Radio in Unlicensed spectrum (NR-U) was introduced as part of 5G to extend cellular capabilities into unlicensed frequency bands, such as the 5 GHz, 6 GHz, and above 52.6GHz bands. Building on the legacy of LTE-LAA, NR-U incorporates mechanisms like Listen-Before-Talk (LBT), Dynamic Frequency Selection (DFS), and adaptive coexistence techniques with Wi-Fi. It was designed to enhance spectrum efficiency and offer flexible deployment options for private networks and enterprise environments.

NR-U adoption has underperformed due to market dynamics rather than technical limitations. By its introduction, Wi-Fi already dominated unlicensed spectrum with a mature ecosystem, offering cost-effectiveness and interoperability. NR-U’s cellular integration and deterministic performance failed to justify its complexity and cost, especially when Wi-Fi met most indoor/enterprise needs at lower investment.

6G Verticals

While 5G introduced several vertical enablers, their limited adoption highlights the gap between technical innovation and real-world deployment. The lessons from URLLC, sidelink, and NR-U demonstrate that ecosystem maturity, long-term compatibility, and economic incentives are just as critical as the technical merits of the air interface itself. As we look toward 6G, a new design philosophy is needed — one that proactively embeds vertical use cases into the physical layer from the beginning. This includes a reimagined approach for integrating IoT, ISAC, NTN, and Robotics in a unified, scalable, and future-proof manner.

6G IoT: Day 1 Support for Scalable and Cost-Efficient IoT

Historically, IoT support in cellular systems has been treated as an afterthought. 4G/LTE introduced NB-IoT and LTE-M through separate standardization efforts for low capability IoT devices. 5G/NR followed a similar path with Reduced Capability (RedCap), eRedCap, and ambient IoT devices introduced only from Release 17 onward, while relying on LTE for intermediate IoT devices. Among all, LTE Cat.1bis—which operates within the existing LTE framework without requiring base station hardware changes—has seen significantly broader adoption despite having a higher device cost than NB-IoT or LTE-M. This underscores a critical insight: leveraging existing RATs for post-eMBB IoT deployment enables wider ecosystem expansion by minimizing infrastructure cost and ensuring seamless integration with current networks.

While NB-IoT and LTE-M offer deep coverage and energy efficiency, their need for coexistence with eMBB devices, scheduling adaptation, and baseband upgrades presents operational hurdles. Similarly, although 5G RedCap aims to reduce complexity via lower bandwidth and capability profiles, reducing RF/baseband bandwidth shows nonlinear cost-benefit behavior. Reducing baseband bandwidth from 100 MHz to 20 MHz significantly lowers complexity and power consumption. A further reduction to 5 MHz offers additional savings, but shrinking further below 5 MHz offers diminishing returns that may not justify the trade-offs in performance and ecosystem support. Further, as RedCap was introduced in later NR releases, namely Rel-17 and Rel-18, it had to overcome design challenges for coexistence with eMBB, particularly for initial access. To overcome these limitations and establish a sustainable path for massive IoT, 6G must adopt a common RAT design that supports both eMBB and IoT services natively, rather than maintaining separate access technologies. A unified air interface operating over a minimum bandwidth between 5–20 MHz provides an optimal balance among device cost, coverage, and power efficiency.

In line with this design philosophy, 6G does not target legacy NB-IoT or LTE-M devices as baseline IoT endpoints. Instead, entry-level support is expected to start from LTE Cat-1 bis or 5G eRedCap-class capabilities, which offer a more practical balance of performance, complexity, and ecosystem maturity. Figure 2 illustrates this envisioned device evolution, showing how 6G aims to streamline its IoT portfolio by focusing on scalable device classes while deferring legacy IoT compatibility to future considerations.

Additionally, the 6G uplink design should support low-power, low-signaling operation, such as grant-free access and sparsely scheduled transmission opportunities, to enable energy-efficient device behavior and extended battery life. At the network level, service-aware resource partitioning is essential to prevent low-rate IoT traffic from interfering with latency-sensitive or high-throughput eMBB sessions. By unifying access under a single RAT and avoiding the fragmentation that plagued prior generations, 6G can deliver true day-one readiness for nationwide IoT—scalable, affordable, and deeply integrated with the broader mobile ecosystem.

6G ISAC: Enabling Intelligent Networks through Embedded Sensing

Integrated Sensing and Communication (ISAC) is a foundational concept in 6G that aims to unify environmental sensing and wireless data transmission into a single, coordinated network function. By leveraging the communication hardware without any modifications, networks can gain real-time awareness of user presence, spatial dynamics, and environmental context. This sensing capability enables intelligent and adaptive resource allocation, leading to self-optimizing networks that are more efficient and responsive. A key communication-specific application of ISAC lies in infrastructure optimization—such as dynamic beam shaping, interference avoidance, and energy-efficient scheduling—ultimately reducing operational complexity for network operators.

Beyond these communication-centric use cases, ISAC also enables broader, network-driven environmental monitoring. This functionality introduces new monetization opportunities for operators and supports government-specific applications in areas such as public safety, disaster response, and urban planning. For example, a 6G network equipped with RF sensing can function as a distributed sensor array, continuously gathering contextual information. Base stations can use this to adapt transmission strategies in real time—adjusting beam directions based on building layouts or local interference patterns. In addition, such sensing capabilities support automated site planning, antenna reconfiguration, and applications like crowdedness detection, structural health monitoring, and drone navigation, contributing to national infrastructure and urban intelligence.

Despite its potential, technical feasibility for ISAC remains a critical consideration. While non-OFDM waveforms such as FMCW or chirp have been studied for enhanced sensing accuracy, these approaches would require significant hardware changes—particularly at the base station (gNB) side. Moving away from OFDM implies a redesign of RF chains, baseband processing, and synchronization mechanisms, leading to high CAPEX and operational complexity. Further, such signaling would represent additional overhead for a network as it cannot be used for communication, therefore questioning the “I” for “integrated” in ISAC and creating a disincentive for adoption. Therefore, 6G ISAC should retain OFDM compatibility, using DFT-s-OFDM or similar variants to allow reuse of existing infrastructure. The sensing function can be embedded directly into standard transmissions—avoiding dedicated sensing symbols or waveform switching.

Use cases for ISAC are broad and increasingly practical. In indoor environments, sensing can identify when areas are unoccupied, enabling proactive energy saving by turning off idle transmission ports or access points. In underground or low-visibility locations, base stations can be selectively activated based on user detection, reducing idle power use. For mobility prediction, trajectory awareness can improve scheduling and handover. On a city-wide scale, ISAC can support environmental monitoring, detecting crowded areas, parking spaces, drone navigation, or structural changes for real-time alerts and disaster response. With proper architectural boundaries and design focus, ISAC will become a key enabler of 6G’s vision: a wireless network that not only communicates, but perceives, adapts, and optimizes its operation intelligently—without relying on disruptive hardware overhauls.

6G NTN: Global Reach without Terrestrial Compromise

NTN, which include LEO/MEO/GEO satellite systems, High Altitude Platforms (HAPS), and UAV-based access points, are expected to play a vital role in 6G by providing ubiquitous global coverage — bridging remote, underserved, and maritime regions. However, the critical design constraint remains unchanged: NTN integration must not compromise the performance, simplicity, or resource efficiency of terrestrial networks (TN), which will continue to serve as the foundation of mobile broadband.

In 5G, NTN capabilities have been incrementally introduced across 3GPP Releases 17, 18, and 19. Release 17 focused on the basic functionality for time/frequency compensation as well as signaling timing adjustments to accommodate the long propagation delay, while Release 18 extended support for uplink coverage enhancement and network verified positioning. Release 19 further explores enhancements such as uplink cell throughput enhancement and downlink coverage enhancement. While each release was purpose-driven — with Release 17 focusing on basic operation and Release 18 extending coverage and positioning — the absence of architectural convergence has led to increased implementation complexity and delayed ecosystem readiness. This fragmentation increases implementation burdens and slows down real-world deployment.

Moreover, as 5G NTN features continue to evolve, it becomes increasingly important to avoid diverging too far from the foundational design principles of TN. If 6G NTN is optimized in an overly specialized direction—centered on satellite-specific mechanisms without alignment to TN architecture—it will inevitably introduce serious challenges. For instance, user equipment may require high-performance RF components, advanced Doppler compensation, and multi-band/multi-timing synchronization, all of which significantly increase hardware complexity, power consumption, and manufacturing cost. These burdens are particularly problematic for consumer-grade devices, which operate under strict constraints on size, cost, and battery life.More critically, such architectural divergence can hinder commercial viability—not only by fragmenting the standards but also by reducing the incentive to adopt 6G NTN over widely deployed proprietary satellite solutions. In scenarios where 6G NTN becomes just another alternative—more costly than TN and possibly even more complex than proprietary NTN offerings—its potential advantages in performance may not justify the adoption overhead. On the other hand, a harmonized TN-NTN design can minimize incremental cost by enabling reuse of infrastructure, waveform, and protocol elements, thereby enhancing commercial attractiveness. That said, this argument can be a double-edged sword: too much emphasis on unification could also stifle optimization opportunities that are essential for satellite-specific constraints.

In this context, NTN should not be treated as an isolated overlay but rather as a native extension of the terrestrial system. This calls for a unified and streamlined framework that consolidates essential NTN features introduced in 5G—across Releases 17, 18, and 19—under a single, coherent architecture. Key design enablers include shared initial access procedures (e.g., SSB acquisition and random access procedure), harmonized numerology that accounts for long round-trip times and Doppler effects, and centralized coordination through TN-based NTN gateways. Such a unified strategy can reduce UE complexity, simplify deployment, and ensure that NTN becomes a scalable, seamless, and economically viable component of the global 6G ecosystem.

AI-Powered Robotics: Enabling Physical Autonomy

The convergence of artificial intelligence and robotics is driving demand for highly reliable and responsive wireless networks. While 5G URLLC established an initial foundation, its real-world implementation remains limited due to complexity and the absence of end-to-end integration. As 6G emerges, a new generation of autonomous machines — commonly referred to as "Physical AI (i.e., AI-powered robots interacting with the physical world)" — will require seamless, real-time interaction with dynamic environments. These systems, including industrial robots, autonomous delivery units, and mobile medical devices, will rely on wireless connectivity not just for data transport but for precise coordination and control.

To support these intelligent agents, 6G networks must go beyond traditional reliability benchmarks and embed deeper network intelligence. This includes predictive scheduling enabled by AI, allowing the network to anticipate robot movements and proactively allocate resources. Anchorless mobility, achieved through simultaneous multi-TRP connectivity, will ensure continuous sessions without handover delays. Equally important is support for multi-level latency requirements, as robotic systems perform various control tasks that demand different responsiveness levels. Additionally, energy-aware state control mechanisms will allow robots to optimize power consumption by adapting their activity states based on task relevance and urgency.

These capabilities will enable transformative applications across industrial and service domains. In smart manufacturing, robots will take over repetitive, precision-driven tasks on assembly lines—rapidly adjusting to product variations or interacting safely alongside human workers. In autonomous logistics, fleets of delivery robots and automated guided vehicles (AGVs) will navigate complex, dynamic environments such as warehouses, ports, and city streets, coordinating in real time without centralized control. Ultimately, such systems have the potential to replace a broad spectrum of manual, routine, and repetitive labor traditionally performed by humans—from sorting packages and stocking shelves to transporting goods—leading to greater efficiency, reduced labor costs, and increased operational uptime.

Robotics will serve as a defining use case for evaluating the real-time and deterministic performance of 6G. Meeting these demands will require tight co-design between the 6G air interface and the robotics/AI application stack. Only by natively embedding robotic requirements into network architecture can 6G enable the next wave of intelligent automation—machines that perceive, decide, and act autonomously and reliably in the physical world.

Figure 4 illustrates a range of robotics use cases enabled by 6G connectivity. These include autonomous systems such as robot taxis, drones, and delivery robots, humanoid robots designed for indoor manufacturing and deployment scenarios spanning human-assisted factories, fully virtualized production lines, and general-purpose outdoor robots. Such applications highlight the need for low-latency, high-reliability, and intelligent scheduling capabilities as foundational enablers of physical AI.

A Unified Framework: Convergence without Complexity

As demonstrated across diverse verticals—ranging from global NTN coverage and cost-sensitive IoT deployments to context-aware ISAC and real-time autonomous robotics—6G must be designed with the needs of vastly different devices and services in mind. A fragmented or overly specialized architecture risks repeating the shortcomings of 5G, where vertical features were added piecemeal and often failed to scale. Instead, 6G must embrace a unified physical layer design that can seamlessly accommodate the full spectrum of vertical use cases from day one, without compromising efficiency or scalability.

This unified framework should be modular, supporting heterogeneous traffic types through flexible numerology, waveform abstraction, and scalable scheduling mechanisms. It must also be power-efficient, particularly for battery-limited and intermittently active IoT devices. Furthermore, the design must be inherently cross-layer aware, embedding hooks for AI-based resource scheduling, adaptive security contexts, and application-driven control feedback. Through such convergence—with minimal complexity—6G can achieve both technical versatility and operational simplicity, ensuring it is not only powerful, but truly usable across industries.

While accounting for vertical-specific requirements is crucial in 6G design, optimizing the architecture for all verticals simultaneously is neither feasible nor practical. Instead, the focus should be on establishing a foundational eMBB-first framework that inherently supports forward compatibility like vertical-integration. This approach ensures that future vertical services—which may emerge at different timelines than 6G’s initial deployment—can be seamlessly integrated as a unified RAT. By prioritizing modularity and scalability in the core design, 6G can maintain efficiency while enabling tailored adaptations for diverse vertical use cases post-deployment.

Conclusion

The 5G experience makes one lesson abundantly clear: vertical services cannot be treated as afterthoughts. When support for use cases like IoT, V2X, or industrial automation is introduced too late or in fragmented form, it results in deployment hesitation, ecosystem fragmentation, and increased cost for both vendors and operators. These challenges highlight the need for a more deliberate and unified approach in 6G—one that anticipates vertical demands and embeds them into the physical layer from the very beginning.

By enabling harmonized NTN integration, embedding IoT support within a common RAT, designing ISAC features around existing OFDM infrastructure, and supporting robotics through deterministic, real-time scheduling, 6G redefines itself not just as a communications system but as a foundational infrastructure for intelligent industries and a framework for future applications. A unified, forward-compatible physical layer will allow 6G to scale with vertical demands, paving the way for truly ubiquitous, adaptive, and intelligent connectivity.