In our previous post, we explored the evolution from 4G to 5G and how the telecom industry is now gearing up for the next leap: 6G.

And at the recent 3GPP 6G Workshop in Korea (Incheon, March 10-11), key industry players gathered to share their early visions of what 6G might become.

Below, I’ve pulled together some highlights from their perspectives.

6G Perspectives

In this first post we’ll cover the visions of some of the main players in the 6G ecosystem: the MNOs, the infra vendors and the chipset makers. The perspectives from other stakeholders, along with a summary of the 3GPP workshop, will be shared in a second post.

Network operators

• AT&T:
  • Focus: Reducing Total Cost of Ownership (TCO) and improving customer experience; exploring new services and revenue opportunities; enhancing existing services.
  • ATT 6G Priorities: software driven network, scalable design, new paradigm for testing and certification, integrated native AI/ML, sensing-as-a-service (ISAC).
• TMO US:
  • Strategic Drivers for 6G: traffic growth, AI, customer experience, 5G to 6G transition, network monetization, sustainability.
  • Suggested Priorities for 6G: voice and regulatory services from day 1, DSS in 7 GHz band and improved unlicensed 6G in 6 GHz, FWA optmizations for reduced payload and better SE, SA based on 5G, intent based programmable architecture.
• Verizon:
  • Goals: increased NW performance, simplification of NW operations with automation, reduce NW slice complexity.
  • Security: quantum-safe cryptography.
  • Data collection framework: 6G system will be rich in data / stats, so minimum data set should be identified and collected in standardized way.
• DTAG:
  • Aims for > 10% spectrum and energy efficiency gain, on an apples-to-apples comparison.
  • Too early to exclude any architecture (still considering NSA 6G).
  • Do not need interworking with: 2G, 3G, and 4G.
  • No need for MBMS, IAB, mobile relay, sidelink with little to no commercial implementations. No need for Home-NodeB/CSG and RAN level aggregation of non-3GPP technologies à la LWA/LWIP.
  • Unrealistic technology drivers should be removed (10x lower latency, 10x higher peak data rate, 10x more accurate positioning, etc.)
  • One solution for one problem: no migration fof each and every legacy function.
• Vodafone:
  • No revolution: Significant changes to the system must be justified from the global business point of view.
• China Mobile:
  • New services: immersive communication for people, LLM as new applications, AI agent as new users, sensing as new capability.
  • Semantic communications will bring a new paradigm shift in wireless communications.
• Orange:
  • IMT-2020 created false expectations on 5G; IMT-2030 should remain realistic.
  • No intention to deploy brand new core network.
• Bouygues:
  • Small data transmission is considerable part of real traffic patterns and is not correctly handled in NR; should be treated by 6G.
  • Network slicing has rigidities that are unacceptable (such as TAC-list of slices link), 6G should make this flexible.
  • TDD frame structure is very rigid and forces operators to synchronize with each other. Issues at country borders. 6G needs to address that.
  • URLLC family target reliability does not make any sense: how to have 99.999% reliability if the availability of the radio system is not even close to that? For example: HW used in base stations have random accidental failures and restarts.

Infra Vendors

• Nokia:
  • Spectrum considered for 6G: Upper 6 GHz; 7.125-8.4 GHz; and 3.1-3.45 GHz in the US. 14 GHz not seen as very attractive.
  • Values of 6G will be: single architecture, no multiple optionsm first release as baseline and end user visibility.
  • 6G specification modernization: GIT as version control tool, and no more Word (is this the best contribution so far?)
• Ericsson:
  • “Minimize Complexity, Maximize Performance”
  • Initial releases for mass-market products; later releases for use cases that gained market traction.
  • Called the 6G Base Station “xNB” in one slide (might be the chosen name for the eNB/gNB equivalent).
• Huawei:
  • Mentions a “NetGPT” as an LLM in an Agentic AI-Framework for 6G core network.
  • Claims 6G will potentially bring 2x spectrum efficiency gain.
• ZTE:
  • Consumer market is still key for 6G.
  • Rising devices: IC/XR, Wearables, CPE/FWA, service robots
  • Discusses NW Architecture for AI (as most of the others do)

Chipset Providers

• Qualcomm:
  • User experience improvement by: better cell-edge data rates, fast access to broadband, optimized UE power consumption.
  • Potential new services: integrated sensing and communication (ISAC)m native high-accuracy positioning, immersive communications.
  • Lean specs: reduce number of options, granularity of UE features and configurability.
  • 4G/5G lessons learnt: separate networks for wide area IoT lead to high costs, for 6G single network for eMBB and IoT is better; long life-cycles of IoT was not considered, for 6G virtualized cores to support legacy systems.
  • Admits most MNOs prefer to reuse 5GC with minimal changes.
• Mediatek:
  • 6G design ought to be realistically ambitious, with practical improvements for users, and with meaningful technology.
  • Critical drivers: Energy efficiency, Robust UL coverage, Ubiquitous AI, Create new business/service opportunities
  • Convergence: communication and computing, terrestrial and non-terrestrial networks, communication and sensing.
• Apple / Apple Silicon:
  • (Yes, they’re making modems now.)
  • 6G is needed to bring: clean and lean design, deprioritizing features not commercially deployed; native support of important use cases without the restriction of backwards compatibility; native support of opportunities of new spectrum.
  • Important to have 5G-6G MRSS (Multi-RAT Spectrum Sharing) from day one.
  • Unified designed with NTN is expected.
  • PHY aspects:
    • Scalable numerology consistent with 5G (SCS = 15*2^n kHz)
    • No need for mixed numerology in the same carrier.
    • 7 GHz support with max channel BW at 200 MHz, and mMIMO.
    • 4k QAM modulation introduction to be studied.
    • Larger SSB periodicity may be considered.
  • User Consent Framework: Privacy and user consent highlighted for 6G.
• Samsung / LSI:
  • (Yes, they’re still making Exynos modems)
  • 5G in the eyes of consumers: enhancement of peak performance alone is not enough anymore.
  • AI-native design from 1st release.
  • No support of 5G-6G dual connectivity and no direct interface between 5G and 6G RAN.
  • Radio: Green Network (G-Net), AI-RAN, Network energy-efficient X-MIMO (NeX-MIMO), enhanced subband full duplex (eSBFD), N-carrier single-cell (NSC), enhanced basic elements (eBE), new verticals (NV).

Hold on – There’s Still more from the Workshop

We’ll continue with a summary of the 3GPP workshop and with the views from device manufacturers, industry alliances and other non-telco members in a second post. Meet you there!