PaperlayerView exact revision on arXiv

Source-linked AI summary

A Survey on 5G: The Next Generation of Mobile Communication

Nisha Panwar, Shantanu Sharma, Awadhesh Kumar Singh

arXiv:1511.01643v1cs.ITcs.DCcs.NI

TL;DR

The paper examines how 5G can address growing device numbers, data volumes, and rate demands beyond 4G. It surveys architectures, technologies, challenges, and demonstrations, concluding that energy consumption and unresolved implementation issues remain central concerns.

  • Problem

    Growing numbers of devices, data usage, and required transfer rates exceed current cellular solutions, motivating the question of what 5G will provide and how.

  • Method

    The paper reviews 5G visions, applications, architectures, technologies, implementation issues, real demonstrations, and testbeds.

  • Results

    The survey identifies 5G features and reviews architectures using small-cells, cognitive radio, device-to-device communication, and cloud-based radio access networks.

  • Takeaways & Limitations

    Energy consumption by infrastructure is a major concern, while open issues across the reviewed architectures may drive future research.

  • Takeaways & Limitations

    The survey identifies unresolved limitations and challenges, including bursty-traffic support, inefficient base-station processing utilization, and security and privacy concerns.

Abstract

from arXiv · show

The rapidly increasing number of mobile devices, voluminous data, and higher data rate are pushing to rethink the current generation of the cellular mobile communication. The next or fifth generation (5G) cellular networks are expected to meet high-end requirements. The 5G networks are broadly characterized by three unique features: ubiquitous connectivity, extremely low latency, and very high-speed data transfer. The 5G networks would provide novel architectures and technologies beyond state-of-the-art architectures and technologies. In this paper, our intent is to find an answer to the question: "what will be done by 5G and how?" We investigate and discuss serious limitations of the fourth generation (4G) cellular networks and corresponding new features of 5G networks. We identify challenges in 5G networks, new technologies for 5G networks, and present a comparative study of the proposed architectures that can be categorized on the basis of energy-efficiency, network hierarchy, and network types. Interestingly, the implementation issues, e.g., interference, QoS, handoff, security-privacy, channel access, and load balancing, hugely effect the realization of 5G networks. Furthermore, our illustrations highlight the feasibility of these models through an evaluation of existing real-experiments and testbeds.

1 Introduction

The paper motivates 5G by limitations in conventional cellular systems and reviews its requirements, architectures, technologies, applications, implementation issues, and demonstrations. It distinguishes its review through broad architectural and implementation coverage while intentionally avoiding an mmWave-oriented discussion.

  • Motivation: Growing device numbers, data usage, and required data rates expose limitations in current cellular systems and motivate 5G development.The paper describes this growth as the increase of device, data, and data-transfer-rate demands.
  • 5G vision: 5G targets user, service-provider, and network-operator needs through connectivity, uninterrupted services, intelligent systems, and secure, scalable infrastructure.The three perspectives include 24×7 connectivity, mission-critical monitoring, energy efficiency, programmability, and security.
  • Paper scope: The paper reviews 5G visions, advantages, applications, architectures, implementation issues, methodologies, technologies, demonstrations, and testbeds.Its architecture coverage includes multi-tier, cognitive-radio, cloud-based, device-proximity, and energy-efficient approaches.
  • Positioning: The review differs from cited prior surveys by covering varied architectures, implementation affairs, technologies, applications, and real-testbed demonstrations while avoiding mmWave-oriented discussion.The authors explicitly contrast their perspective with three earlier review works.
  • 4G limitations: Conventional cellular networks face bursty-traffic overhead, inefficient base-station processing use, co-channel interference, limited heterogeneous-network support, indoor-outdoor inefficiency, and access latency.The reported latency for accessing the best candidate base station is several hundreds of milliseconds.

2 Desideratum of 5G Networks

The desiderata for 5G arise from continued growth in connected devices and data traffic. The paper emphasizes massive scalability, high-rate data transfer, efficient spectrum use, ubiquitous connectivity, and delay-bound services with suitable QoS and QoE.

  • Device scalability: 5G must support a dramatic increase in connected smartphones, appliances, vehicles, robots, sensors, wearables, and other devices.The paper expects this growth to continue exponentially and characterizes 5G as supporting massively connected devices.
  • Data rate and streaming: Data trading is expected to become 100-times greater than in 2014, requiring new architectures, methods, technologies, and distribution capabilities.The passage links device growth with increased video, audio, browsing, gaming, real-time, bursty, and multimedia traffic.
  • Spectrum utilization: 5G requires improved spectrum utilization because separate uplink and downlink channels are redundant and allocated spectrum is partly under-utilized.The paper calls for access control improvements and spectrum broadening.
  • Connectivity: Ubiquitous connectivity requires support for diverse radios, radio-access technologies, frequency bands, and duplex options across heterogeneous networks.The paper envisions seamless UE connectivity over HetNets despite non-identical operating bands and market-specific duplexing.
  • Latency and service quality: 5G is expected to support real-time and delay-bound applications, including the tactile Internet, with optimized bandwidth, latency, jitter, packet loss, packet delay, QoS, and QoE.The stated target is suitable service quality for applications with varying QoS and QoE requirements.

3 Challenges in the Development of 5G Networks

5G development must address capacity, energy, latency, reliability, scalability, mobility, QoS, interference, and economic constraints through coordinated architectural and technological solutions.

  • Core challenges: 5G must substantially increase network capacity and data rates while keeping energy consumption and deployment costs under strict control.More base stations, higher frequencies, and improved links can expand capacity but create economic and energy burdens.
  • Core challenges: 5G must support scalable demand across heterogeneous networks while maintaining performance across data rate, coverage, QoS, latency, reliability, fairness, and energy-efficiency.These parameters involve tradeoffs that motivate joint optimization.
  • Core challenges: Interference will intensify with HetNets, CRNs, full duplex, and D2D communication, requiring reliable management for allocation, power, association, and load balancing.Potential interference sources include multiple macrocell base stations, UEs, and small-cell base stations.
  • Core challenges: Energy efficiency is essential because radio access networks consume 70%-80% of total power, motivating more efficient communication systems, hardware, and technologies.Proposed approaches include Cloud-RAN, VLC, mmWave, multi-tier architectures, D2D, massive MIMO, and full duplex radios.
  • Core challenges: Extremely low latency and reliable delivery are difficult for large-scale real-time applications without increasing infrastructure cost.Relevant applications include patient monitoring, life-safety systems, gaming, nuclear reactors, drones, and connected transportation.
  • Core challenges: High mobility across cells and radio-access technologies requires handoff mechanisms that preserve active service connections, including inter-tier, intra-tier, multi-RAT, and secure handoffs.Rapidly changing user locations and concentrated movement patterns complicate spectrum use and service continuity.

4 Architectures for the Future/5G Mobile Cellular Networks

The paper reviews proposed 5G architectures and evaluates them through their advantages, disadvantages, and unresolved challenges.

  • Architecture scope: The paper examines multi-tier, CRN-based, D2D-based, and cloud-based architectures in light of their benefits, drawbacks, and open challenges.These architectures are presented as existing proposals for future 5G networks.

4.1 Two-tier Architectures

Two-tier architectures place small-cells beneath macrocells to extend coverage and services, while densification, mobile cells, relaying, and self-healing address capacity and mobility needs.

  • Architecture: Two-tier architectures use a macrocell top tier and supervised small-cells beneath it, with both tiers sharing an identical frequency band.Small-cells may include femtocells, picocells, and microcells.
  • Mobility: Mobile small-cells inside trains, buses, and buildings can aggregate users so the macrocell serves each group as a single unit.External antenna arrays communicate with the macrocell while internal users connect through the mobile small-cell or WiFi.
  • Densification: Network densification combines more antennas and base stations with spectral aggregation above 3 GHz to expand coverage and capacity.Cell-range expansion biases handoff toward nearby small-cells, while indoor-to-outdoor service extends small-cell utility.
  • Advantages: Small-cells improve data rate, spectrum efficiency, energy use, cost, on-demand capacity, macrocell congestion, and handoff overhead.Shorter communication ranges reduce signaling overhead, and mobile small-cells can perform handoff on behalf of associated users.
  • Open issues: Two-tier deployments face inter-tier and intra-tier interference, difficult backhaul transfer, implementation costs, authentication demands, and frequent topology updates.Backhaul options include wired optical fiber, wireless point-to-multipoint, and wireless point-to-point links.
  • Self-healing: Self-healing architectures detect failed cells through centralized or distributed approaches, trading global coordination against communication and computational overhead.Failure detection can use abnormal user behavior, received signal strengths, and handoff patterns.

4.2 Cognitive Radio Network based Architectures

CRN-based architectures opportunistically use heterogeneous spectrum to build multi-tier 5G networks, reduce interference, and improve spectrum utilization, while retaining energy and sensing tradeoffs.

  • CRN foundations: A cognitive radio network uses secondary users with learning and adaptive capabilities to scan and opportunistically operate across heterogeneous channels when primary users are absent.The architecture distinguishes licensed primary users from secondary users exploiting available spectrum.
  • CRN role: CRNs support 5G multi-tier architectures, interference reduction, and lower network energy consumption.Their use spans cognitive small-cells and other heterogeneous network configurations.
  • CRN architectures: Non-cooperative CRNs form multi-RAT systems with separate licensed and cognitive-channel interfaces, while secondary users operate only on cognitive channels.The CRN overlays the existing licensed cellular network and can integrate with it at upper layers.
  • Cooperative CRNs: Cooperative CRNs let a small-cell scan macrocell activity and use temporarily unoccupied spectrum holes while minimally disrupting macrocell operations.The small-cell acts as a secondary user serving its UEs on bands unused by primary users.
  • Advantages: Cognitive small-cells can minimize interference by avoiding channels used by neighboring small-cells and increase capacity by exploiting spectrum holes.Spectrum-hole use can enhance bandwidth utilization and support higher data transfer rates.
  • Open issues: CRN deployment must balance energy efficiency against the tradeoff between spatial frequency reuse and outage probability through effective spectrum sensing.Cellular networks also consume energy in circuits, cooling systems, and radio transmission.

4.3 Device-to-Device Communication Architectures

The paper surveys D2D communication architectures, types, resource-allocation methods, benefits, and implementation challenges within 5G networks.

  • D2D communication lets nearby UEs communicate over licensed cellular bandwidth without, or with controlled involvement of, an MBS.
  • Challenges: Key open issues include interference management, secure and privacy-preserving relay communication, network coding, and simultaneous D2D and BS communication modes.
  • D2D communication includes operator- or device-controlled direct communication and relaying, including two-way and multiple-access relay channels.
  • Architectures and resource allocation: Social-aware D2D architecture uses ties, communities, centrality, and bridges to guide resource allocation and improve discovery, congestion, or throughput.
  • Architectures and resource allocation: Proposed resource-allocation methods address fairness, service requirements, throughput, and delay by adapting subcarrier, power, and channel assignments.
  • Advantages: D2D communication can improve link reliability, data rate, instant communication, spectral efficiency, and power consumption while offloading traffic from an MBS.

4.4 Cloud-based Architectures

Cloud-based architectures move selected mobile-network functions into cloud layers to provide scalable, flexible, and software-oriented 5G services, while introducing fronthaul and real-time constraints.

  • C-RANs execute most MBS functions in the cloud by dividing functionality into control and data layers, enabling dynamic service allocation without costly network-device installation.
  • BBUs commonly reside in the cloud while RRHs remain at MBSs, producing an easily scalable and flexible architecture.
  • Advantages and challenges: C-RAN deployment requires efficient RRH-to-BBU fronthaul and sufficiently fast cloud processing to maintain real-time performance.
  • C-RAN architectures: Architectures include full or partial centralization, nearby or larger clouds for latency-sensitive applications, and hierarchical controllers for MBS, RAN, and network functions.
  • C-RAN architectures: Basic C-RAN data layers perform signal processing, while control layers manage baseband processing, resources, QoS, mobility, security, and power control.
  • C-RAN architectures: CONCERT adds a software-defined service layer above the control layer to address raw-baseband exchange and processing-delay disadvantages of full centralization.
  • Advantages and challenges: C-RANs support software-based, power-efficient, flexible, and scalable services, including dynamic interference, traffic, load, and mobility management.

4.5 Energy-Efficient Architectures for 5G Networks

The surveyed energy-efficient 5G architectures jointly consider energy, spectrum, channel conditions, traffic load, and network association.

  • Energy-efficient 5G infrastructure is pursued through joint optimization of energy efficiency and spectral efficiency.
  • User-centric networks can let UEs select uplink and downlink channels from different BSs according to load, channel conditions, services, and application requirements.
  • An energy-efficient C-RAN balances UE service across RRHs and manages interference to reduce SBS and MBS power consumption.
  • UE association should consider both uplink and downlink channels rather than relying entirely on either channel.

5 Implementation Issues in 5G Networks

The paper reviews 5G implementation issues and proposed techniques for interference, handoff, QoS, and security, while identifying unresolved demands such as tactile-Internet latency.

  • Interference management: Interference management spans advanced receivers, joint scheduling, distributed cell access, power control, and resource-aware association in multi-tier networks.
  • Interference management: Self-interference cancellation enables small-cells to provide services and backhaul simultaneously, achieving almost the performance of wired optical-fiber connections while reducing cost.
  • Interference management: Full-duplex interference mitigation is more complex in multi-cell networks because uplink and downlink interference can occur across cells on identical time and frequency.
  • Handoff management: 5G handoff is complicated by network densification, high mobility, zero-latency goals, and access to multiple radio-access technologies.
  • Handoff management: Proposed handoff methods use predicted measurements, thresholds, adjacent-cell parameters, and authentication modules to select cells and prepare secure transitions.
  • QoS management: QoS approaches include heterogeneous statistical delay-bounded provisioning, cloud-based quality monitoring, and link-specific routing for multi-hop D2D communication.
  • QoS management: Current proposed architectures do not efficiently support tactile-Internet services requiring latency around 1 millisecond for remote sensory interaction.
  • Security and privacy management: 5G security and privacy must address fast authentication, secure handoff, multi-RAT protection, cloud risks, authorization, availability, confidentiality, and integrity.

6 Methodologies and Technologies for 5G Networks

The survey reviews enabling technologies and methodologies for 5G, including radio, network, deployment, caching, and software-based approaches. It also notes that several reviewed methods remain subjects of ongoing research.

  • Architectures and implementation: The survey covers cognitive radio, D2D communication, dense or multi-tier deployment, C-RANs, green communication, and techniques for interference, QoS, handoff, channel access, and load balancing.These methodologies and technologies are distributed across the paper's architecture and implementation discussions.
  • Radio and network techniques: Full-duplex radios require self-interference cancellation, while downlink and uplink decoupling lets users select different base stations using link quality, load, and backhaul capacity.SIC supports high-throughput and low-latency applications; DUD contributes to a user-centric architecture.
  • Cloud and software-based networking: SDN separates programmable network control from data forwarding, while NFV implements network functions in software and C-RANs combine both approaches.These approaches offload functionality to software running on commodity servers and target scalability and flexibility.
  • High-speed wireless technologies: Millimeter-wave communication uses 30–300 GHz bands to pursue high-speed data transfer, but it faces challenges across physical, MAC, and network layers.Research focuses on bands including 28 GHz, 38 GHz, 60 GHz, and the E-band.
  • Device and service technologies: M2M communication connects network devices without human intervention, supporting applications such as transport, healthcare, infrastructure monitoring, retail, and security.The paper identifies further development challenges for M2M communication.
  • High-speed wireless technologies: Massive MIMO serves many user equipments simultaneously through hundreds of antennas, relying on spatial multiplexing and channel knowledge at the macro base station.The approach requires expensive equipment mounted at the macro base station.

7 Applications of 5G Networks

The paper links 5G's expected low latency, high speed, and ubiquitous connectivity to applications spanning industry, homes, societies, energy, transport, healthcare, and logistics.

  • Application scope: 5G's zero latency, high-speed data transfer, and ubiquitous connectivity are presented as foundations for a wide range of applications and services.The section organizes these applications across industrial, personal, domestic, societal, infrastructure, and tracking contexts.
  • Personal usages: Personal 5G services would support scalable and heterogeneous user equipment while satisfying multimedia, voice, web, and QoS requirements.
  • Virtualized homes: C-RAN-based virtualized homes could use low-cost user equipment while moving higher-layer applications to the cloud for universal access and outsourced computation.
  • Smart societies: Smart societies would interconnect homes, offices, stores, and digital appliances so collaborative actions enhance user experience.
  • Smart grids: Smart grids would use frequent remote-sensor observations and analysis to adjust energy distribution and improve efficiency and economic benefits.
  • Mobility, healthcare, and tracking: Low-latency 5G would support automated vehicles, real-time vehicular communication, remote healthcare data transfer, and continuous location-based logistics tracking.The cited applications include vehicle coordination, urgent medical services, RFID-based tracking, and tactile Internet scenarios.
  • Industrial usages: Industrial 5G applications would provide robots, sensors, drones, devices, users, and data collectors with real-time data to manage industrial functions quickly while preserving energy.

8 Real Demonstrations and Testbeds for 5G Networks

The survey describes industry and academic demonstrations that evaluate 5G technologies through simulators, field experiments, testbeds, and large coordinated research projects.

  • Industry demonstrations: DOCOMO's simulator evaluated massive MIMO, small cells, and millimeter wave, showing 1000-times higher system capacity and 90% of users achieving 1Gbps.DOCOMO also conducted a 2012 experiment with 10Gbps uplink speed.
  • Industry demonstrations: Samsung's 28 GHz millimeter-wave experiment achieved 1.2Gbps on a vehicle traveling above 100km/h and up to 7.5Gbps near a stop.The peak rate was reported as more than 30-times faster than state-of-the-art 4G technology.
  • Research projects and testbeds: METIS developed more than 140 technical components spanning air interfaces, waveforms, access, antennas, interference, mobility, D2D, and resource management.Its testbeds covered D2D communication, massive machine communications, and waveform design.
  • Research projects and testbeds: European Commission projects including 5GNOW, TROPIC, MCN, COMBO, MOTO, and PHYLAWS contributed to ongoing 5G development.
  • Platforms and architectures: Existing testbeds and platforms covered massive MIMO, C-RANs, M2M communication, and full-duplex radio across industry, academia, and open experimentation projects.Examples include TitanMIMO, China Mobile's C-RAN work, iJoin, and OpenAirInterface.

9 Concluding Remarks

The survey concludes that 5G combines new architectures and techniques to pursue high-speed, ubiquitous connectivity, while implementation challenges and open issues remain. It emphasizes that 5G infrastructure design is still under progress, especially regarding security, privacy, and operational guarantees.

  • Contributions and findings: The survey reviews 5G architectures based on small cells, cognitive radio, D2D communication, and cloud-based radio access networks, including energy-efficient designs.
  • Implementation issues: 5G development must address interference removal, handoff management, QoS guarantees, channel access, and infrastructure load balancing alongside architectural design.
  • Reviewed techniques: The survey discusses full-duplex radios, dense deployment, SIC, DUD, millimeter wave, massive MIMO, and VLC as techniques relevant to 5G development.
  • Open issues: The design of 5G infrastructure is still under progress, and solutions to prominent issues are linked to early deployment and long-term network growth.
  • Open issues: Security and privacy for devices, infrastructure, communications, and data transfer remain unexplored, while future systems also require self-healing, self-configuring, and self-optimizing operation.The paper specifically discusses dynamic load balancing, QoS, traffic management, residual-resource pooling, virtualization, and real-time guarantees as unresolved concerns.
Loading 1511.01643v1…