A Comprehensive Survey on 6G Networks: Applications, Core Services, Enabling Technologies, and Future Challenges

Amin Shahraki, Mahmoud Abbasi, Md. Jalil Piran, Mingzhe Chen, Shuguang CuiarXiv 2021 · OpenReview

Abstract

While 5G deployment continues globally, the research community is already defining the "Next G." This paper provides a comprehensive roadmap for 6G networks, moving beyond simple faster downloads to a world of "intelligence-as-a-service." 6G is characterized by the use of Terahertz frequencies, sub-millisecond latency, and ubiquitous 3D coverage spanning from the deep sea to outer space.

Introduction: Beyond the Limits of 5G

The authors argue that despite the revolutionary leap of 5G, current mobile networks will soon face significant bottlenecks. By 2023, projections suggest nearly 15 billion connected devices, many of which will demand performance levels that 5G cannot sustain.

Specifically, 5G faces challenges in:

  • Connection Density: Difficulty supporting the projected 10 million devices per square kilometer.
  • Extreme Latency: Emerging "Tactile Internet" applications (e.g., remote surgery) and holographic communications require latencies significantly lower than 1ms.
  • Fixed Architectures: Current networks are often too rigid for the dynamic nature of massive IoT.
  • Coverage Gaps: Terrestrial networks leave massive "dead zones" in rural areas, oceans, and the sky.

Methodology

As a comprehensive survey, this paper performs a systematic meta-analysis of the 6G landscape rather than proposing a single algorithm. The authors evaluate 14 other major 6G surveys to identify research gaps and create a detailed taxonomy of the field.

The structural taxonomy of the survey, illustrating the progression from requirements and research activities to evolutionary and revolutionary technologies.

Key Enabling Technologies

The paper categorizes 6G technologies into "Evolutionary" (upgrades to existing tech like MIMO and MEC) and "Revolutionary" (entirely new paradigms like Quantum Communications and THz bands).

A summary of the diverse enabling technologies required for 6G, ranging from AI and energy harvesting to non-terrestrial networks and holographic MIMO surfaces.

From 2D to 3D Coverage

6G envisions a "Space-Air-Ground-Sea" integrated network. This involves using UAVs (drones) as flying base stations and Low Earth Orbit (LEO) satellites to provide 100% global coverage, moving beyond traditional ground-based towers.

The "Self-X" and Cell-less Network

The authors highlight that 6G must be "Self-X"—self-learning, self-configuring, and self-healing. AI will act as the "operating system" of the network itself. Furthermore, 6G may move toward a "cell-less" architecture where a user-centric approach allows a device to be served by a cloud of cooperating access points simultaneously, eliminating handover delays.

The Tactile Internet

This represents a shift from the "Internet of Content" to the "Internet of Skills." By achieving sub-millisecond latency and centimeter-level positioning, 6G could allow a surgeon to operate on a patient remotely with real-time haptic (touch) feedback.

Performance Targets

The paper defines staggering performance targets for 6G compared to its predecessor:

  • Peak Data Rate: 1 Tbps (vs. 20 Gbps in 5G).
  • Latency: < 0.1 ms (vs. 1 ms in 5G).
  • Reliability: 99.99999% or "seven nines."
  • Energy Efficiency: 1 pJ/bit, aiming for "battery-less" IoT devices that utilize energy harvesting.
  • Spectrum: Moving into the Terahertz (THz) range (0.1–10 THz), offering massive bandwidth despite propagation challenges.

Future Challenges

Several massive hurdles remain for 6G implementation:

  • THz Fragility: High-frequency signals are easily blocked by physical obstacles. Developing "Large Intelligent Surfaces" (LIS) to reflect and steer these signals is a priority.
  • The AI Compute Paradox: Running complex Machine Learning models at the "edge" consumes significant power, potentially contradicting "Green 6G" goals.
  • Security: Ensuring privacy in a decentralized world (e.g., using Federated Learning) remains an unsolved challenge.
  • 3D Channel Modeling: There is a current lack of accurate mathematical models for signal behavior in complex 3D environments involving high-speed drones and satellites.

Summary updated

BibTeX
@misc{shahraki2021comprehensive,
  author = {Amin Shahraki and Mahmoud Abbasi and Md. Jalil Piran and Mingzhe Chen and Shuguang Cui},
  title = {{A Comprehensive Survey on 6G Networks: Applications, Core Services, Enabling Technologies, and Future Challenges}},
  year = {2021},
  url = {https://openreview.net/forum?id=e4kMC55B2e_},
}
Figure