5G-Advanced, 6G, Governance and Sustainable Wireless Futures

Introduction and Context

Wireless communication is entering a strategic transition. Fifth-generation mobile systems have moved beyond the initial promise of higher data rates toward programmable, cloud-native, multiservice infrastructures. 5G-Advanced extends this foundation through improved radio intelligence, positioning, energy efficiency, automation, and support for integrated sensing and communication. Research on 6G explores still more ambitious capabilities, including native artificial intelligence, ubiquitous connectivity, communication at sub-terahertz and terahertz frequencies, highly accurate sensing, and closer integration with satellites, aircraft, vehicles, and industrial systems.

These developments cannot be evaluated only by peak throughput. A wireless system is also shaped by reliability, latency, coverage, energy consumption, resilience, privacy, accessibility, affordability, regulation, and social legitimacy. Consequently, the engineering problem is increasingly a systems-governance problem: technical architecture, spectrum policy, standards, business models, environmental constraints, and ethical safeguards must be designed together.

Learning Objectives

After studying this chapter, students should be able to:

1. Explain the role of massive MIMO, ultra-reliable low-latency communication, integrated sensing and communication, and reconfigurable intelligent surfaces.
2. Analyse the opportunities and limitations of terahertz and other 6G candidate technologies.
3. Describe digital twins and their use in wireless planning, operation, and industrial control.
4. Evaluate spectrum governance, standards, security, privacy, accessibility, and ethical responsibilities.
5. Assess the sustainability of future wireless networks using technical, environmental, and social criteria.
6. Formulate responsible research and deployment priorities for 5G-Advanced and 6G.

From 5G to 5G-Advanced and 6G

5G introduced three broad service families: enhanced mobile broadband (eMBB), ultra-reliable and low-latency communication (URLLC), and massive machine-type communication (mMTC). These categories are useful, but real deployments often combine them. A smart factory, for example, may require high-rate video, deterministic control, machine sensing, positioning, and cybersecurity simultaneously.

5G-Advanced is best understood as an evolution rather than a completely separate generation. Its likely emphasis includes:

• improved uplink and mobility performance;
• enhanced positioning and timing;
• reduced energy consumption;
• AI-assisted network optimization;
• extended reality and industrial wireless applications;
• non-terrestrial network integration;
• improved multicast, broadcast, and sidelink operation;
• support for sensing functions alongside communication.

6G remains a research and standardization direction rather than a universally finalized commercial specification. Proposed characteristics include native AI, joint communication and sensing, sub-THz or THz operation, extreme device density, ubiquitous coverage, and tighter cyber-physical integration. Claims about exact 6G data rates or latency should therefore be treated as targets or scenarios, not guaranteed specifications.

Massive MIMO and Distributed Intelligence

Multiple-input multiple-output (MIMO) uses several antennas at the transmitter and receiver to exploit spatial dimensions. Massive MIMO extends this idea by equipping a base station with a large antenna array serving multiple users on the same time-frequency resources. Spatial processing separates users and concentrates energy toward intended terminals.

In a simplified narrowband uplink model,

where 𝐲 is the received signal vector, 𝐇 is the channel matrix, 𝐱 is the transmitted symbol vector, ρ is a power-scaling factor, and 𝐧 is receiver noise. The model assumes synchronized symbols, a linear channel over the signaling interval, and suitable knowledge of channel state information.

A linear detector may estimate the transmitted vector using zero-forcing processing:

provided that 𝐇ᴴ𝐇 is nonsingular and the number and quality of spatial observations are sufficient. Here, 𝐇ᴴ denotes the conjugate transpose. Zero forcing suppresses inter-user interference but can amplify noise when the channel is ill-conditioned. Minimum mean-square error processing usually provides a better compromise between interference suppression and noise enhancement.

With many antennas, independently varying small-scale fading can average out, a phenomenon often called channel hardening. However, practical systems do not enjoy unlimited gains. Performance is constrained by pilot contamination, correlated antenna channels, hardware impairments, finite-resolution converters, mutual coupling, pilot overhead, blockage, and power consumption. Distributed or cell-free massive MIMO places cooperating access points over an area and can reduce cell-edge disparity, but it increases fronthaul, synchronization, coordination, and security requirements.

Ultra-Reliable and Low-Latency Communication

URLLC supports applications in which a late or lost packet may be more harmful than a modest reduction in throughput. Examples include closed-loop industrial control, cooperative transport, remote operation, and some healthcare systems. Reliability must be defined carefully: it may refer to the probability that a packet is delivered within a specified deadline, not merely the average packet error rate.

A useful conceptual requirement is

where D is end-to-end delay, D_max is the deadline, and ε is the permitted deadline-violation probability. This is a service-level condition, not a universal physical law. It must specify traffic type, packet size, channel conditions, mobility, geographic area, and observation period.

Latency includes radio transmission, processing, queuing, transport, and application delays. Reducing radio transmission time alone may not produce low end-to-end latency. Techniques include short transmission time intervals, grant-free access, edge computing, packet duplication, redundant paths, deterministic scheduling, network slicing, and local breakout. Reliability can be improved through diversity in frequency, time, space, nodes, and routes.

There is an unavoidable design tension among reliability, latency, spectral efficiency, and energy. Aggressive retransmission can improve reliability but increase delay and energy use. Redundancy can improve resilience but consume spectrum. A responsible design therefore specifies the consequence of failure and selects reliability targets proportionately.

Integrated Sensing and Communication

Integrated sensing and communication (ISAC) uses wireless signals both to transfer information and to infer properties of the environment. A base station may estimate range, velocity, angle, occupancy, or motion from reflected signals while simultaneously serving communication users.

For a simplified monostatic radar interpretation, range is related to round-trip delay τ by

where R is target range in metres, c is propagation speed in metres per second, and τ is measured round-trip delay. The factor 2 applies because the signal travels to the target and back. Real systems must account for waveform bandwidth, sampling, multipath, synchronization, antenna geometry, and estimation error.

ISAC can support traffic monitoring, robot localization, gesture recognition, industrial inspection, and assisted living. Nevertheless, sensing changes the privacy model of a network. A system may infer presence, movement, behaviour, or occupancy without a user transmitting an explicit application message. Governance should therefore address informed notice, data minimization, retention, access control, purpose limitation, and the possibility of inference from supposedly anonymous measurements.

Terahertz and Sub-Terahertz Wireless Links

Terahertz research commonly concerns frequencies approximately between 0.1 and 10 THz, although boundaries vary by field. Higher frequency provides wider potential bandwidth and enables very small antennas, but propagation is difficult. Free-space path loss increases with frequency and distance. In a simplified isotropic model,

where L_FS is the linear free-space power-loss ratio, d is distance, f is frequency, and c is propagation speed. In decibels, the corresponding loss is

The equations assume unobstructed free-space propagation and isotropic antennas; practical antenna gains, atmospheric absorption, rain, blockage, and reflection must be added separately. Molecular absorption can create frequency windows with comparatively lower attenuation. THz links are therefore most promising for short-range, line-of-sight, high-capacity applications such as chip-to-chip communication, indoor backhaul, data-centre interconnection, and specialized sensing.

Major obstacles include weak received power, difficult beam alignment, hardware nonlinearity, expensive converters, blockage by people or objects, and limited range. THz should not be presented as a universal replacement for lower-frequency coverage layers. A heterogeneous architecture combining low-band coverage, mid-band capacity, millimetre-wave hotspots, optical or wired links, and possibly THz short-range links is more realistic.

Reconfigurable Intelligent Surfaces

A reconfigurable intelligent surface (RIS) is an engineered surface containing many controllable elements that modify the phase, amplitude, or polarization of incident electromagnetic waves. It can redirect energy around obstacles or improve a propagation path without using a conventional active radio chain at every element.

An RIS-assisted channel is often represented conceptually as

where 𝐇_direct is the direct channel, 𝐇₁ is the transmitter-to-surface channel, 𝐇₂ is the surface-to-receiver channel, and 𝐐 is a controllable reflection matrix. This model is useful for analysis but assumes an appropriate narrowband representation and accurate channel knowledge. It does not imply that an RIS creates energy or eliminates path loss.

RIS deployment requires control signaling, calibration, surface-observer coordination, and optimization under discrete phase states and hardware limitations. Passive surfaces may have low operating power, but large-scale control infrastructure and manufacturing impacts remain. RIS is most valuable where geometry, blockage, and user concentration make a carefully positioned surface useful; random deployment does not guarantee improvement.

Digital Twins and AI-Native Networks

A digital twin is a continuously updated computational representation of a physical system. In wireless networking, it may represent buildings, antenna sites, users, traffic, channels, energy flows, faults, and policies. Data from measurements update the twin, while simulations and optimization tools use it to test changes before applying them to the physical network.

A practical twin lifecycle is:

1. define the system boundary and decision purpose;
2. collect and validate telemetry;
3. construct a model combining physics and data-driven components;
4. calibrate and quantify uncertainty;
5. run what-if analysis or optimization;
6. apply approved changes with safeguards;
7. compare predicted and observed outcomes.

AI can predict traffic, detect anomalies, select beams, allocate resources, and reduce energy use. Yet an AI model may fail under distribution shift, adversarial input, missing data, or unusual propagation conditions. Safety-critical decisions should use confidence estimates, human oversight, rollback procedures, audit logs, and conservative fallback modes. A digital twin is not automatically accurate merely because it is detailed; model validity must be measured against independent observations.

Spectrum Governance and Standards

Spectrum is a shared and finite public resource. Governance determines who may transmit, at what power, in which band, with what protection from interference, and under which licensing arrangement. Common approaches include exclusive licensing, general authorization, licence-exempt access, dynamic sharing, secondary markets, and local or private licences.

A sound spectrum decision considers propagation, incumbent users, geographic scale, coexistence, equipment ecosystems, emergency access, competition, and social value. Auctions can allocate rights but do not by themselves guarantee affordability, rural coverage, or innovation. Regulators may use coverage obligations, open-access provisions, local licensing, spectrum-sharing databases, and interference limits to pursue broader objectives.

Standards bodies develop interoperable specifications through technical contributions, testing, consensus, and intellectual-property processes. Standards reduce vendor lock-in and enable roaming, but they may also reflect the resources and priorities of powerful participants. Engineers should distinguish a research prototype, an industry specification, a national regulation, and a deployed product. These are related but not interchangeable.

Security, Ethics, Accessibility, and Human Rights

Future wireless networks enlarge the attack surface because they connect vehicles, factories, homes, sensors, cloud platforms, and satellites. Important controls include strong device identity, secure boot, key management, software-update mechanisms, network segmentation, zero-trust principles, intrusion detection, supply-chain assurance, and incident response.

Privacy requires more than encryption. Metadata, location, radio fingerprints, sensing outputs, and inferred behaviour can be sensitive. Ethical design should apply privacy by design, data minimization, explicit purpose specification, proportionality, transparency, and meaningful redress. Automated resource allocation should be tested for discriminatory effects, especially where coverage or service quality differs among communities.

Accessibility includes affordability, device usability, coverage for rural and remote populations, support for people with disabilities, multilingual interfaces, and service continuity during disasters. Universal design is preferable to treating accessibility as a later add-on. Public-interest networks, community networks, shared infrastructure, and open standards may help reduce the digital divide.

Sustainability and Life-Cycle Engineering

Wireless sustainability must be assessed across the entire life cycle: raw materials, manufacturing, transport, deployment, operation, maintenance, reuse, and end-of-life treatment. Network energy use depends on traffic, radio configuration, cooling, site power, backhaul, and embodied impacts.

A basic energy-efficiency metric is

where η_E is energy efficiency, B is delivered useful data in bits, and E is energy consumed in joules; the unit is bit per joule. The metric is meaningful only when the system boundary is stated. A site-level measure may exclude device energy, while a life-cycle measure may include manufacturing and disposal.

Sustainable practices include sleep modes, renewable-powered sites, efficient cooling, right-sized capacity, shared infrastructure, repairable equipment, software optimization, longer product lifetimes, and responsible recycling. However, improved energy per bit can produce a rebound effect if total traffic grows faster than efficiency improves. Sustainability must therefore consider absolute energy, material demand, emissions, water use, e-waste, and social effects—not efficiency alone.

Applied Case: Illustrative Smart-Port Wireless Transformation

Background

Consider an illustrative medium-sized container port planning a 5G-Advanced private network. It wants autonomous inspection vehicles, augmented-reality maintenance, environmental sensors, worker-location services, and camera-based safety monitoring.

Problem

The port currently uses separate Wi-Fi, proprietary radio, and wired systems. Coverage is inconsistent near metal containers, video traffic congests shared links, and the operator cannot demonstrate that worker-location data are used proportionately. The port also has a sustainability target to reduce network energy and equipment waste.

Analysis

A single service design is inappropriate. Vehicle control requires bounded latency and high reliability; cameras require uplink capacity; environmental sensors need low energy; and location services require privacy safeguards. Metal surfaces and moving cranes create rapidly changing channels. A digital twin can represent the yard, crane geometry, traffic patterns, radio measurements, and energy consumption.

Approach

The port deploys a private 5G-Advanced system with edge computing near the control centre. Massive MIMO provides spatial separation, while redundant paths and packet duplication support selected control flows. RIS is tested only at known blockage locations. ISAC is evaluated for vehicle and crane detection, but raw sensing data are retained briefly and access is role-controlled. The digital twin compares beam plans and sleep policies before deployment. A governance board includes engineers, operators, workers, safety specialists, and privacy officers.

Performance is evaluated using service-level indicators: deadline-violation probability for control traffic, availability by operating zone, uplink capacity, energy per delivered bit, false-alarm rate, accessibility, and incident-recovery time. A fallback wired or low-rate control mode is retained for safety-critical operations.

Outcome

The design improves operational visibility without assuming that every 6G technology is necessary. The port identifies that better site planning and edge placement produce more value than an indiscriminate THz deployment. It also discovers that energy savings are greatest when idle radios and computing resources are coordinated with actual work schedules.

Learning Points

Technology selection must follow service requirements. Sensing creates governance obligations. Digital twins are valuable only when calibrated. Reliability claims require explicit definitions. Sustainability and accessibility should be performance objectives from the beginning, not post-deployment corrections.

Research Directions and Best Practices

Important research questions include robust joint communication-sensing waveforms; low-complexity optimization for distributed arrays and RIS; channel models for dynamic and non-terrestrial environments; trustworthy AI under distribution shift; energy-aware orchestration across radio, edge, and cloud layers; privacy-preserving sensing; interoperable digital twins; and equitable spectrum-sharing mechanisms.

Best practice for engineers includes:

• define measurable service requirements before selecting technology;
• separate aspirational 6G targets from standardized capabilities;
• report assumptions, uncertainty, and system boundaries;
• test under mobility, blockage, interference, and failure conditions;
• design security, privacy, accessibility, and sustainability from the start;
• validate AI and digital-twin models with independent data;
• preserve safe fallback modes for critical functions;
• document spectrum, standards, licensing, and interoperability constraints;
• evaluate total cost and life-cycle impact rather than peak performance alone.

Chapter Summary

5G-Advanced and 6G represent an evolution toward intelligent, sensing-aware, heterogeneous, and deeply integrated wireless infrastructures. Massive MIMO exploits spatial dimensions but is limited by channel correlation, pilot overhead, hardware imperfections, and coordination cost. URLLC requires explicit probabilistic delay and reliability definitions, supported by diversity, edge processing, deterministic scheduling, and redundancy. ISAC combines data transfer with environmental inference, creating valuable applications but also significant privacy responsibilities.

Terahertz links offer potentially enormous bandwidth but face severe propagation, hardware, alignment, and blockage limitations. Reconfigurable intelligent surfaces can shape propagation in suitable geometries but do not remove path loss or eliminate the need for control. Digital twins support safer planning and optimization when their models are calibrated and uncertainty is managed.

Wireless futures depend equally on spectrum governance, open and accountable standards, cybersecurity, privacy, accessibility, and life-cycle sustainability. Responsible engineering evaluates not only speed and capacity but also social value, resilience, affordability, environmental impact, and human rights.

Key Terms

Massive MIMO: A multi-antenna system using large arrays to spatially multiplex and beamform signals.

URLLC: Wireless communication requiring very high reliability and bounded low latency.

ISAC: Integrated sensing and communication using shared signals and infrastructure.

Terahertz: A high-frequency region commonly associated with approximately 0.1–10 THz.

RIS: A controllable electromagnetic surface that modifies signal propagation.

Channel hardening: Reduction in relative channel fluctuations as antenna observations increase.

Pilot contamination: Interference caused by non-orthogonal or reused channel-estimation pilots.

Digital twin: A continuously updated computational representation of a physical system.

Network slicing: Logical partitioning of shared infrastructure into service-specific networks.

Edge computing: Processing data near the point of generation to reduce transport delay.

Spectrum sharing: Coordinated use of a frequency band by multiple users or systems.

Licence-exempt access: Radio use under general technical rules rather than individual licences.

Network slicing: Logical resource isolation for different service requirements.

Privacy by design: Incorporating privacy protections into architecture and operation from inception.

Energy efficiency: Useful delivered data or service per unit of consumed energy.

Life-cycle assessment: Evaluation of environmental effects across a product or system’s life.

Digital divide: Unequal access to connectivity, devices, skills, or beneficial services.

Distribution shift: Change between the data used to train a model and its operating environment.

Non-terrestrial network: A network using satellites or airborne platforms to provide connectivity.

Interoperability: The ability of independently produced systems to work together.

Review Questions

5 Short Answer Questions

1. What practical limitations prevent massive MIMO gains from increasing without bound?
2. Why must URLLC reliability be defined together with a deadline?
3. State two potential applications of ISAC and one associated privacy concern.
4. Why are THz systems generally more suitable for short-range links?
5. What is the purpose of calibrating a digital twin?

5 Long Answer Questions

1. Explain the architecture, benefits, and limitations of massive MIMO in a multi-user network.
2. Discuss how redundancy, edge computing, scheduling, and packet duplication can support URLLC.
3. Compare THz communication and RIS-assisted communication as candidate 6G technologies.
4. Analyse the relationship between spectrum policy, standards, competition, and public interest.
5. Explain why sustainability must include life-cycle impacts rather than operational energy alone.

3 Analytical / Application Questions

1. A factory requires a control service with a strict deadline, high reliability, and moderate data rate, while cameras require high uplink capacity. Design a service architecture and justify the placement of edge computing, redundancy, and radio resources.
2. A city proposes ISAC-enabled street infrastructure that detects pedestrian movement. Develop a governance plan addressing consent, data retention, access control, error handling, and public accountability.
3. Compare two proposed network upgrades using energy per delivered bit, coverage, resilience, accessibility, and life-cycle impact. Explain why peak data rate alone is insufficient for the decision.

Multiple Choice Questions with Answers

Question 1

Which feature most directly enables massive MIMO to serve several users on the same time-frequency resources?

A. Spatial separation using multiple antennas

B. Increasing packet size only

C. Removing channel estimation

D. Using a single omnidirectional antenna

Correct Answer: A. Spatial separation using multiple antennas

Question 2

In the expression P{D ≤ D_max} ≥ 1 − ε, ε represents the:

A. Carrier frequency

B. Permitted deadline-violation probability

C. Number of antennas

D. Propagation speed

Correct Answer: B. Permitted deadline-violation probability

Question 3

For a monostatic sensing system, why is range given by R = cτ/2 rather than R = cτ?

A. The signal travels at half the speed of light

B. The signal makes a round trip

C. The antenna has half power

D. The bandwidth is doubled

Correct Answer: B. The signal makes a round trip

Question 4

Which is a principal limitation of terahertz communication?

A. Insufficient theoretical bandwidth

B. Severe blockage and propagation loss

C. Inability to use directional antennas

D. Requirement for low-frequency oscillators

Correct Answer: B. Severe blockage and propagation loss

Question 5

What does an RIS primarily do?

A. Generate unlimited new radio energy

B. Modify the propagation of incident electromagnetic waves

C. Replace all baseband processing

D. Guarantee zero interference in every environment

Correct Answer: B. Modify the propagation of incident electromagnetic waves

Question 6

Which statement best describes a digital twin?

A. A static equipment inventory

B. A continuously updated computational representation of a physical system

C. A spectrum licence

D. A replacement for all physical measurements

Correct Answer: B. A continuously updated computational representation of a physical system

Question 7

Which spectrum arrangement allows operation under general technical rules without an individual exclusive licence?

A. Licence-exempt access

B. Exclusive national assignment

C. Satellite-only allocation

D. Emergency reservation only

Correct Answer: A. Licence-exempt access

Question 8

If a network delivers 2 × 10¹² bits while consuming 1000 J, its energy efficiency is:

A. 2 × 10⁶ bit/J

B. 2 × 10⁹ bit/J

C. 2 × 10¹² bit/J

D. 2 × 10¹⁵ bit/J

Correct Answer: B. 2 × 10⁹ bit/J

Question 9

Why can ISAC create additional privacy risks?

A. It prevents all encryption

B. It can infer presence or behaviour from environmental measurements

C. It cannot measure any physical property

D. It operates only in licensed spectrum

Correct Answer: B. It can infer presence or behaviour from environmental measurements

Question 10

Which practice best supports responsible AI-assisted wireless control?

A. Deploying without validation to maximize speed

B. Using confidence estimates, audit logs, and safe fallback modes

C. Discarding all human oversight

D. Training only on ideal laboratory data

Correct Answer: B. Using confidence estimates, audit logs, and safe fallback modes