1 of 23

6G Vision : Requirements, Spectrum and Architectural View

Abhay Karandikar

Secretary to Government of India

Ministry of Science & Technology, Department of Science & Technology

(On leave from Professor, IIT Bombay)

dstsec@nic.in

(Based on research done by my group at IIT Bombay and IIT Kanpur)

​

2 of 23

Agenda

  • IMT 2030 (6G) System Requirements
  • Spectrum for 6G
  • Architecture
    • Limitations of 5G System Architecture
    • Points to Ponder
    • 6G System Architecture - Some Thoughts

3 of 23

6G Requirements

4 of 23

Requirements for 6G System

  • Huge Data Volume
    • Mobile Networks – Primary vehicle for Connectivity
  • Massive Connectivity
    • Massive signalling load
  • Service (Use Case) Diversity
    • “Very High Throughput” to “Very Low Throughput” Applications
    • “Latency Tolerant” to “Stringent Low Latency” Applications
  • Diverse Set of Users
    • Stationary Users, Mobile Users, Users moving at very high speeds
    • Connectivity for everything/everywhere
  • A Variety of Access Technologies
    • Cellular Access, WLANs, Satellite Access…
    • Small Cells, Large Cells
    • Unicast, Broadcast
  • Efficient & Cost-effective Service Delivery
    • Sustainability, Energy Efficiency

IMT 2030 System

5 of 23

Mobile Data Volume in 2029 – Estimates for India

  • Mobile Data Traffic Estimates for India - 2029
    • Most households likely to have mobile broadband access
    • Conservative Estimates
      • ~62 Exabytes/month
    • Realistic Estimates
      • ~92 Exabytes/month
  • Even a conservative estimate indicates
    • Huge data volume by 2029
    • ~62 Exabytes monthly
    • ~744 Exabytes annually

The huge data traffic coupled with Massive and Ubiquitous Connectivity scenarios would generate a sizeable signalling load and put significant pressure on control plane functions

6 of 23

6G Spectrum

7 of 23

Recommendations for 6G Bands

Frequency

Spectrum

Bandwidth

THz Bands

Low and mid

Bands

Expand and position a larger mid-band

To expand and position a larger mid‐band (7‐24 GHz) to meet requirements of 6G technologies

Open up few lower, mid and mmWave bands

To enable maximization of the spectrum and use and socioeconomic benefits, open up a few bands to generate demand (for example 450-470 MHz, 526-612 MHz, 31-31.3 GHz, etc.)

Delicensed or license-exempt bands

To enable innovation and gigabit public Wi-Fi by exploiting technology innovation for example Wi-Fi 6E or WiGig etc.,

The lower part of 6 GHz band and at least 4.32 GHz in V band can be delicensed

Tera Hz Bands (spectrum from 90 GHz to 3000 GHz)

To bring focus on 6G driven active antenna systems and Intelligent Reflector Surfaces (IRS)

To encourage research, make some of the THz bands’ licenses exempt for some periods both for commercial deployment and R&D

Coverage

8 of 23

6G Architecture

9 of 23

5G Network Architecture – Some Limitations

3GPP 5G Architecture

Converged Core - Multi-RAT Unification in Core

  • But No Unification at RAN Level
  • Fragmented Decision Making in RAN

1

Tight and proprietary coupling between Radio and CN protocol stacks

    • Loss of Flexibility – Can you connect 5G RAN to 4G Core or directly to Internet w/o Core?

2

3

Service/User Agnostic Handling

    • Fixed Route/Path for Control & Data flows
    • Usage of Core Network in every Scenario
      • Usage of Tunnels for all data flows
    • No use case specific variation in Protocol Behaviour

10 of 23

Architecture for 6G - Points to Ponder (1/3)

  • Scalable Architecture
    • High signalling load – How to avoid making control plane a bottleneck?
      • Further Disaggregation of Control Plane
      • Decoupling of Signalling Handling and User Plane Control
      • Decoupling of Signalling and Data
    • Further Disaggregation of User Plane
  • Unified Multi-access RAN
    • Multi-Connectivity & Multi-access Convergence
    • Unified Treatment of Dual Connectivity …
  • Usage of SDN Paradigm

11 of 23

Architecture for 6G - Points to Ponder (2/3)

  • Need for core in cellular network
    • Mobility - Anchored in Core
    • Also Authentication, Access Control…
  • A large % of mobile network users not “mobile”
    • Rural Broadband Connectivity, IoT …
    • Can we bypass core for such users?
    • Direct Connectivity to Internet from RAN
  • Should we decouple RAN from Core?
    • Interworking of any RAN with any Core
      • Non Standalone Architecture requires 5G RAN to interwork with 4G Core
        • Achieved with the help of 4G-RAN
        • Not possible w/o 4G RAN
    • Connect future 6G RAN to 5G Core

12 of 23

Architecture for 6G - Points to Ponder (3/3)

  • Flexible Architecture
    • Flexible Protocol Structure
      • Not rigidly layered
      • Tunnelling protocols not required for all users
    • Virtualization of Network Resources
      • Better support for Network Slicing,…
  • Energy Efficient Networks
  • Intelligence-driven Network
    • Optimization of Services/Applications
      • AI/ML Model/Data Distribution
      • Federated Learning
    • AI-powered Network Design & Optimization
      • AI-powered Optimization
      • AI-powered Protocol Stacks
      • Learning-oriented Network Design

Service/user dependent path for a flow

Service/user based choice of protocols, and its configuration

  • Flexibility
  • Improved Performance
  • Reduced Cost
  • Higher (Energy) Efficiency

​

13 of 23

6G System Architecture – A few proposals

14 of 23

Scalable Architecture - RAN User(Data) Plane Disaggregation

  • RAN User (Data) Plane of most RATS perform similar functions in 5G
    • Radio Tx/Rx
    • PHY & MAC
    • Link Adaptation
    • Security (Encryption)
    • Optimization - Header Compression …
    • Interworking with Core
  • Can we Disaggregate RAN along these simpler functions?
  • Does it help in unified treatment of RATs?
  • Does it help in Load Management, Dual Connectivity?

15 of 23

Unification & Virtualization of Disaggregated Multi-RAT RAN

  • Virtualization Layer (SDN Middleware)
    • A Layer between Control & Data Plane
    • Abstract Information Model of Multi-RAT RAN Data plane
      • Virtualize Underlying Data Plane
      • Modularized Information Model
    • Unify Control & Management of Multi-RAT RAN
  • Unified Control Plane
    • Usage of SDN Technology
    • Controls RAN Data Plane Functions of all RATS
    • SDN Middleware Abstraction helps in Unified Control
  • Improved handling of
    • Load balancing, Dual Connectivity, Network Slicing

Courtesy : IEEE 1930.1-2022

16 of 23

5G-Flow - Core Bypass using SDN

Macro BS Radio IF

WLAN Radio Interface

SDN Switch

SDN Controller

Unified Core

Network Interface

Data Network Interface

Multi-RAT Network Switch

UE

UE

SDN Switch

Macro BS IF

WLAN IF

UE

Data

Other Radio Interfaces

CN Comm.

(Signaling)

DHCP Server

Content Server

Edge Storage and Compute

Auth Server

Introduction of SDN switches and controller in access network

    • Replaces proprietary interfaces
    • Enables unified interworking

Introduction of SDN switches at UE

    • Replaces proprietary interfaces

SDN based RAT agnostic controller

    • Unified RAN level control & management

Decouples UE’s communication with CN from its communication with RAN

    • UE - CN communication - Overlay over UE -RAN communication

17 of 23

5G-Serv : UE Signalling as Payload

  • 5G System
    • Separate Control and User Plane Functions
  • User Plane in 5G System
    • Responsible for Data Forwarding
  • Control Plane in 5G System
    • Performs two types of tasks
    • Task #1
      • Controls User Plane - “Resource Control”
    • Task #2
      • Exchanges Signalling Messages with UE
      • UE Control & State Management
      • Provide services such as Mobility, Authentication…
  • Let us separate Task #1 and #2
    • Separation of User Plane Control and UE Signaling Exchange functionalities?
    • Leads - Further Disaggregation of Control Plane

Reference: “5G-Serv: Decoupling User Control and Network Control in the 3GPP 5G Network”; Meghna Khaturia, Akshatha M Nayak, Pranav Jha, Abhay Karandikar, ICIN 2021

​

Recommendation ITU-T Y.2325: “Architectural evolution for NGN control plane by applying SDN technology”

18 of 23

5G-Serv : Impact on RAN+CN Control Plane

  • UE Signalling Exchange functionality separated from Control Plane Functions
    • Signalling Service Functions – NAS Server, RRC Server, Authentication Server, …
    • Service Function/Orchestrator – Mobility/PDU Conn. Service Orchestrator, …
  • Control Plane : User Plane Control (Resource Control)
  • UE Signalling (RRC/NAS) Messages
    • A form of Data (Payload) flowing through 5G network

Leads to a Generic Architecture

19 of 23

5G-Serv : A Service Driven Architecture for IMT-2030

*RAN and Core NFs would likely be separate

All Services treated Uniformly - External AF/AS/IMS based and Internal services (Mobility, PDU Connectivity…)

20 of 23

5G-Serv : Highlights

    • Disaggregated and Modular Control Plane
    • Possibility of Use case specific variants of UE Signalling Protocols
    • But Impact on UE Signalling Message not necessary
    • Flexible Signaling Handling function Placement and Chaining
    • Decoupling of Signalling and Data

Enhanced Modularity & Flexibility

    • Primarily controls User Plane as in SDN paradigm
    • Does not exchange signalling messages with UEs
    • Simpler message flow & protocols (simpler NGAP, F1AP as they do not carry UE signalling messages)
    • Reduced Load on Control Plane - as Signaling handling a part of Data Plane

Scalable Control Plane

    • UE Signalling as Payload (Data)
    • All Services treated Uniformly - External AF/AS/IMS based and Internal services (Mobility, PDU Connectivity…)
    • Improved Network Access Security

Change in Paradigm

21 of 23

Conclusion

Motivation

    • Massive & Ubiquitous connectivity
    • High speed data
    • Sensing, Intelligence
    • Sustainability

Spectrum

    • THz bands
    • More options in low and mid bands
    • Expanding the spectrum horizon with harmonization

Design principles

    • Service driven core
    • Multiple RATs
    • Energy awareness
    • AI/ML driven design
    • Core-free services

22 of 23

References

  • Standards Contribution
    • S1-241407, Dynamic service adjustment support based on energy information, 3GPP TSG-WG SA1 Meeting #106, 27-31 May 2024, Jeju, South Korea
    • S1-241306, Provisioning of energy aware security in the network, 3GPP TSG-WG SA1 Meeting #106, 27-31 May 2024, Jeju, South Korea
    • ITU-T Y.2325: ”Architectural evolution for the next generation network control plane by applying software-defined networking technology”, Dec 2023
    • IEEE 1930.1-2022, “IEEE Recommended Practice for Software Defined Networking (SDN) based Middleware for Control and Management of Wireless Networks”, Sep 2022
  • Journal & Conference Papers
    • An Architecture for Control Plane Slicing in Beyond 5G Networks, Rashmi Yadav, Rashmi Kamran, Pranav Jha and Abhay Karandikar, Journal of Computer Networks, Volume 249 (110511), July 2024
    • Green 6G: Energy Awareness in Design, Rashmi Kamran, Shwetha Kiran, Pranav Jha, Abhay Karandikar and Prasanna Chaporkar, Proceedings of Workshop on Standards-driven Research in 16th International Conference on Communication Systems and Networks (COMSNETS) 2024
    • 5G-Flow: A Unified Multi-RAT RAN Architecture for Beyond 5G Networks, Meghna Khaturia, Pranav Jha and Abhay Karandikar, Journal of Computer Networks, Volume 198 (108412), October 2021
    • 5G-Serv: Decoupling User Control and Network Control in the 3GPP 5G Network, Meghna Khaturia, Akshatha Nayak, Pranav Jha and Abhay Karandikar, Proceedings of Conference on Innovation in Clouds, Internet, and Networks (ICIN) 2021
    • Open5G: A Software-Defined Networking Protocol for 5G Multi-RAT Wireless Networks, Pradnya Kiri Taksande, Pranav Jha, Abhay Karandikar and Prasanna Chaporkar, Proceedings of Workshop on Open-RAN: Open Road to Next Generation Mobile Networks, IEEE WCNC 2020
    • Control and Management of Multiple RATs in Wireless Networks: An SDN Approach, Akshatha Nayak M., Arghyadip Roy, Pranav Jha, and Abhay Karandikar, Proceedings of IEEE 5G World Forum (WF-5G) 2019
  • Patents
    • A Network Architecture for Unified Handling of Services, P Jha, R Kamran, S Kiran, A Karandikar, P Chaoprkar, February 2024 (India) (202421014717) and July 2024 (USA) (18/773,102)
    • Methods and Systems for Radio Access Network Aggregation and Uniform Control of Multi-RAT Networks, Akshatha Nayak, Pranav Jha, Abhay Karandikar, Prasanna Chaporkar, US Patent (11,553,546) granted on January 2023
    • Methods and Systems for Controlling a SDN based Multi-RAT Communication Network, Abhay Karandikar, Pranav Jha, et.al., US Patent (10,187,928) granted on January 2019

23 of 23

THANK YOU