PaperlayerView exact revision on arXiv

Source-linked AI summary

Towards Enabling Critical mMTC: A Review of URLLC within mMTC

Shiva Raj Pokhrel, Jie Ding, Jihong Park, Ok-Sun Park, Jinho Choi

arXiv:2007.08793v1cs.NIeess.SP

TL;DR

Critical mMTC must combine mMTC scalability with URLLC reliability and latency for resource-constrained devices, a convergence that existing approaches do not fully address. The paper surveys separate mMTC and URLLC technologies, identifies their conflicting requirements, and reviews potential solutions across layers. It characterizes critical mMTC requirements and discusses enabling directions including access, allocation, networking, and higher-layer techniques.

  • Problem

    Critical mMTC must support delay-sensitive and delay-tolerant devices together while meeting scalability, reliability, and latency requirements.

  • Method

    The paper reviews mMTC and URLLC state-of-the-art technologies, identifies their conflicting requirements, and examines potential solutions across different layers.

  • Results

    The review characterizes critical mMTC requirements and identifies enabling approaches spanning random access, resource allocation, congestion management, networking, and higher layers.

  • Takeaways & Limitations

    Critical mMTC requires coordinated treatment of heterogeneous devices and requirements rather than relying solely on separate mMTC and URLLC approaches.

Abstract

from arXiv · show

Massive machine-type communication (mMTC) and ultra-reliable and low-latency communication (URLLC) are two key service types in the fifth-generation (5G) communication systems, pursuing scalability and reliability with low-latency, respectively. These two extreme services are envisaged to agglomerate together into \emph{critical mMTC} shortly with emerging use cases (e.g., wide-area disaster monitoring, wireless factory automation), creating new challenges to designing wireless systems beyond 5G. While conventional network slicing is effective in supporting a simple mixture of mMTC and URLLC, it is difficult to simultaneously guarantee the reliability, latency, and scalability requirements of critical mMTC (e.g., < 4ms latency, $10^6$ devices/km$^2$ for factory automation) with limited radio resources. Furthermore, recently proposed solutions to scalable URLLC (e.g., machine learning aided URLLC for driverless vehicles) are ill-suited to critical mMTC whose machine type users have minimal energy budget and computing capability that should be (tightly) optimized for given tasks. To this end, our paper aims to characterize promising use cases of critical mMTC and search for their possible solutions. To this end, we first review the state-of-the-art (SOTA) technologies for separate mMTC and URLLC services and then identify key challenges from conflicting SOTA requirements, followed by potential approaches to prospective critical mMTC solutions at different layers.

I. INTRODUCTION

5G’s mMTC and URLLC services pursue scalability and reliable low latency, but their convergence into critical mMTC creates difficult requirements for emerging applications. The paper reviews this convergence, its use cases, and potential solutions across network layers.

  • Critical mMTC combines scalability with stringent reliability and latency requirements for applications such as smart manufacturing and autonomous vehicles.These applications include industrial automation and large-scale connected vehicles in smart cities.
  • Wireless industrial automation requires jointly integrating URLLC and mMTC because standardized wireless technologies and 5G NR may not meet its low-delay and ultra-high-reliability requirements.
  • Conventional URLLC–mMTC mixtures separate services through physical resources, whereas scalable URLLC addresses increased MTC connections with varied URLLC requirements.The passage distinguishes separable sliced mixtures from non-separable scalable URLLC traffic.
  • Critical mMTC enhances URLLC requirements for only a fraction of massively connected devices, making resource-configuration overhead significant when device roles vary.The paper also notes that MTC devices generally have limited computing capability and energy resources.

B. Review of Related Survey

Existing surveys separately cover URLLC, mMTC, access schemes, machine learning, and dense IoT, but the convergence of URLLC and mMTC remains insufficiently reviewed. This paper organizes the related landscape, identifies critical mMTC challenges, and discusses enabling directions across the cellular IoT stack.

  • Earlier surveys review URLLC and mMTC enabling approaches, challenges, and applications in different IoT contexts, while critical mMTC remains poorly reviewed.
  • Related literature separately surveys random-access schemes, machine-learning and data-analytics approaches, and highly dense IoT relevant to cellular mMTC.
  • The paper addresses the missing systematic study of large device populations carrying mixed mMTC and URLLC services in cellular IoT networks.It highlights heterogeneous QoS provisioning, URLLC within mMTC, transmission scheduling, and RAN congestion avoidance as key concerns.
  • C. Contributions: The contributions classify major challenges, characterize traffic and vertical applications, and identify future ideas for supporting URLLC within large mMTC populations.
  • C. Contributions: The paper reviews higher-layer approaches for separate URLLC and mMTC operation and discusses concepts for jointly handling them under critical mMTC specifications.
  • D. Structure of the Paper: The paper structure covers mMTC state of the art, URLLC functionality, network-level enhancements, and subsequent application and solution topics.

A. Features of NB-IoT

NB-IoT supports wide-area massive connectivity through narrowband resource configurations, extended coverage, and power-saving mechanisms. Its 3GPP evolution further targets latency, spectral efficiency, energy efficiency, and operational improvements.

  • Deployment: NB-IoT uses 180 kHz carriers and supports in-band, stand-alone, and guard-band deployment modes.These modes reuse LTE PRBs, GSM carriers, or LTE guard-band spectrum.
  • Massive Connectivity: Up to 48 devices can be simultaneously allocated using 3.75 kHz subcarrier spacing, supporting power-constrained wide-area devices.The 3.75 kHz mode uses single-tone resource units lasting 32 ms.
  • Wide Coverage: 164 dB MCL extends NB-IoT coverage by 20 dB compared with GSM and general packet radio service.Transmission repetitions improve decoding when received signal power is below the noise power.
  • Low Power Consumption: Wake-up signals, power saving mode, and extended discontinuous reception reduce device energy consumption.Wake-up signals avoid regular paging checks, while PSM and eDRX provide additional LTE-inherited power-saving functions.
  • Evolution: 3GPP Releases 14–16 enhanced NB-IoT data rates, latency, coverage, spectral efficiency, energy efficiency, and access procedures.Release 16 included device-group wake-up signals, improved idle-mode uplink transmission, and a common two-step PRACH.
  • Research Directions: Research on NB-IoT has addressed random-access evaluation, coverage, energy efficiency, and co-channel interference.

C. RAN & Network Enhancements for Massive Connectivity

5G enhancements combine cellular IoT, vertical-domain support, service-based architecture, slicing, automation, and time-sensitive networking with NR mechanisms for URLLC. These mechanisms target massive connectivity alongside strict latency and reliability requirements.

  • 5G Vertical Support: 5G expands cellular IoT support through 5G CIoT, NR IIoT, vertical LAN, advanced V2X architecture, and network automation.
  • Service-Based Architecture: Service-based architecture enables flexible implementation by selectively deploying defined network functions and services.
  • Network Slicing: Network slicing reallocates access, mobility, session-management, and control functions while supporting slice-specific authentication and authorization.
  • Automation and Orchestration: Automation and orchestration use data collection and network analytics for slice-specific load balancing, performance evaluation, mobility, and sustainable QoS.
  • Time-Based Networking: Time-based networking provides packet-delivery synchronization across hops for time-sensitive services such as industrial automation.
  • URLLC Requirements: URLLC targets mission-critical applications with >99.999% BLER and <1 ms user-plane latency.
  • Low-Latency Mechanisms: A 60 kHz, 14-symbol TTI is 0.25 ms, while a 2-symbol mini-slot reaches 35.7 µs to facilitate low-latency transmission.

B. Enhancements for URLLC

Later 3GPP releases evolved the NR URLLC foundation toward stricter requirements and new verticals, while research proposed analytical and advanced grant-free access schemes. These efforts focus on reducing access failures and outage under tight latency constraints.

  • 3GPP Evolution: Release 16 expanded URLLC to factory automation, transport, and electrical power distribution, while Releases 16 and 17 enhanced existing capabilities.
  • Grant-Free Access: Analytical work compared reactive, K-repetition, and proactive grant-free random access schemes for URLLC.The proactive scheme produced the lowest latent access failure probability under shorter latency constraints.
  • Beyond-5G Research: Beyond-5G proposals use NOMA-based grant-free access and repetition to address shortcomings of existing URLLC access schemes and reduce outage.

IV. NETWORK LEVEL URLLC AND MMTC ENHANCEMENTS

Network-level enhancements address the limits of conventional cellular architectures through slicing, SDN/NFV, and edge computing. These approaches can adapt resources and processing to heterogeneous URLLC and mMTC demands, but dynamic sharing and delay–power trade-offs remain open challenges.

  • Network Architecture: Conventional cellular architecture requires broader redesign because its mobile-broadband fundamentals cannot support diversified services.
  • Network Slicing: Network slicing creates multiple virtual domains over shared infrastructure, reserving resources and limiting interruptions among eMBB, mMTC, and URLLC services.
  • Dynamic Slicing: Dynamic slicing can plug and play slices as critical-mMTC traffic shifts between uplink and downlink.
  • Dynamic Slicing: Constrained dynamic optimization allocates shared resources under loss and delay constraints to satisfy application-dependent URLLC service-level agreements.
  • Open Challenges: Dynamic resource isolation and sharing remain challenging because traffic fluctuations and channel variation complicate slice management.
  • SDN and NFV: SDN separates control and forwarding planes for dynamic flow management, while NFV provides programmable and flexible network functions.
  • Edge Computing: Mobile edge and end-node computing can reduce latency and save computation and transmit power through optimized offloading and resource allocation.
  • Open Challenges: The delay–computing-power trade-off and optimal offloading and scheduling at edge and end devices require further investigation.

D. Self-organizing Networks

Self-organizing networks are proposed to offset the management complexity introduced by integrating slicing, SDN, NFV, and other techniques for diverse critical mMTC services. The reviewed approaches span autonomous service-aware management and cross-layer optimization of delay, reliability, and energy.

  • D. Self-organizing Networks: Integrating network techniques can improve scalability and flexibility but may compromise quality of service and user experience through management complexity.Self-organizing management is proposed to provide optimization, distributed management, intelligence, and automation.
  • D. Self-organizing Networks: Cross-layer design can improve end-to-end delay and reliability by jointly optimizing interdependent physical, link, and network-layer delay components.The discussion connects packet transmission, buffer queuing, and routing delays to resource utilization under end-to-end constraints.
  • D. Self-organizing Networks: Cross-layer scheduling has been studied using queueing and channel conditions to minimize average power under an average-delay limit.Related work also applies cross-layer methods to wireless sensor-network latency and energy efficiency.
  • D. Self-organizing Networks: A wireless-sensor-network cross-layer approach improved energy efficiency while managing data-networking latency.The cited work addresses latency through cross-layer coordination rather than an isolated protocol layer.
  • D. Self-organizing Networks: Cross-layer techniques have also been applied to URLLC radio communication, offloading, and NOMA systems.These examples extend cross-layer design beyond sensor-network latency and power management.

V. CRITICAL MMTC USE CASES AND EXISTING SOLUTIONS

The paper frames critical mMTC as coexistence between critical and non-critical machine traffic with distinct service requirements. It reviews use cases and existing solutions based on RAN slicing, edge computing, and self-organizing networks.

  • V. CRITICAL MMTC USE CASES AND EXISTING SOLUTIONS: Critical mMTC use cases are studied where massive non-critical and critical MTC devices coexist with heterogeneous traffic.The section first summarizes mMTC and URLLC service-level agreements before discussing use cases and solutions.
  • A. Critical mMTC Use Cases: Smart-city critical mMTC can involve environmental and infrastructure monitoring, smart grids, and industrial automation triggered by events.After a triggering event, subsets of devices sense critical information and are grouped by functionality or geographic location.
  • A. Critical mMTC Use Cases: Industrial-automation traffic includes regular control transmissions with stringent latency requirements and separate alarm traffic classes.Control messages must arrive within a specified control-loop period.
  • A. Critical mMTC Use Cases: Reserving preambles can guarantee URLLC access, but limited resources may increase non-URLLC collision probability.Exploiting sporadic URLLC arrivals is discussed as a way to improve preamble utilization.
  • B. Existing Solutions: RAN slicing allocates resources across service slices, while edge and end-node computing move connectivity, storage, processing, and decision-making closer to devices.RAN slicing increasingly addresses lower-layer radio allocation, whereas edge computing shifts processing from centralized servers toward end nodes.
  • B. Existing Solutions: Self-organizing networks complement slicing and edge computing because integrating multiple techniques can complicate configuration management and compromise perceived service quality.The proposed role is adaptive management of the resulting heterogeneous network.

VI. CHALLENGES TOWARDS ENABLING CRITICAL MMTC

Critical mMTC must reconcile mMTC’s massive, low-power connectivity with URLLC’s stringent latency and reliability requirements. The paper identifies mixed-numerology interference and grant-free access-resource trade-offs among the central challenges.

  • VI. CHALLENGES TOWARDS ENABLING CRITICAL MMTC: Existing 5G challenges for critical mMTC arise from conflicting service requirements and are summarized in the paper’s challenge framework.The section introduces these challenges before discussing physical-layer and access-related constraints.
  • A. Mixed-Numerology Interference: mMTC uses narrowband transmission, small subcarrier spacing, and low sampling rates, whereas URLLC uses larger subcarrier spacing and higher sampling rates.These configuration disparities support different connectivity, coverage, power, and latency objectives.
  • A. Mixed-Numerology Interference: Heterogeneous baseband and RF configurations inevitably create significant interference in critical mMTC.The interference follows from combining the contrasting physical-layer requirements of mMTC and URLLC.
  • B. RA Preamble Resource Utilization: Grant-free access removes request-before-transmission signaling to reduce control overhead and access latency.The approach is considered for both URLLC and mMTC, but scarce preambles must serve their different access requirements.
  • B. RA Preamble Resource Utilization: Reserving preambles protects URLLC priority but can waste resources and reduce availability for massive sporadic mMTC traffic.Contention-based access uses preambles more flexibly but can cause simultaneous-device collisions that jeopardize reliability and latency.

D. Wide Area Coverage (WAC)

Wide-area critical mMTC must support delay-sensitive and delay-tolerant devices under constrained bandwidth, energy, and network-management capacity. The paper therefore highlights scalable networking, congestion control, distributed computation, and caching as enabling concerns.

  • D. Wide Area Coverage (WAC): Wide-area environments challenge critical mMTC latency and reliability because they are poorly suited to conventional URLLC solutions.The resulting constraints motivate new techniques and network-architecture principles.
  • D. Wide Area Coverage (WAC): Preserving massive preambles and increasing transmit power are infeasible for critical mMTC under limited bandwidth and energy.This boundary applies directly to wide-area deployment conditions.
  • D. Wide Area Coverage (WAC): Spatial diversity can complement time and frequency diversity, but accurate CSI acquisition introduces feedback overhead that must remain manageable.The trade-off is between exploiting spatial diversity and limiting signaling burden.
  • D. Wide Area Coverage (WAC): Critical mMTC requires flexible, scalable networking for device populations roughly ten to hundred times those of existing cellular networks.The network must preserve URLLC quality of service as mMTC density increases.
  • D. Wide Area Coverage (WAC): Converging critical mMTC with existing cellular networks can cause congestion across the RAN, core network, and signaling system.Careful monitoring and congestion management are consequently required.
  • D. Wide Area Coverage (WAC): Distributed computing and caching should be coordinated for context-aware, time-sensitive communication because centralized cellular management is sluggish.The paper identifies decentralized resource management as the needed direction.
  • D. Wide Area Coverage (WAC): Potential enablers target coexistence between delay-sensitive and delay-tolerant devices with different requirements.The section positions these enablers as mechanisms for critical mMTC deployment.

A. Cooperative transmissions

Critical mMTC requires resource-allocation and slicing approaches that balance delay-sensitive URLLC access against massive mMTC access under heterogeneous and time-varying conditions. The reviewed directions include predictive allocation, cooperative transmission, non-orthogonal sharing, and adaptive cross-slice resource management.

  • Cooperative transmissions: Cooperative transmission provides spatial diversity for single-antenna devices and can create virtual uplink MIMO using selected delay-tolerant devices.Battery life and geometric distance can guide cooperative-device selection.
  • Cooperative transmissions: Preamble allocation must balance URLLC’s high access priority against massive mMTC access because grant-free preamble resources are scarce.Predictive allocation can use pseudo-periodic device activity to reduce access delay without severe spectral-efficiency degradation.
  • Cooperative transmissions: Immediate retransmission can shorten access delay, but it increases system load and limits the number of delay-sensitive devices unless coded random access is adopted.A short or zero backoff delay enables immediate retransmission, creating a stability trade-off.
  • Cooperative transmissions: Non-orthogonal RAN sharing and heterogeneous NOMA can exploit reliability diversity across services to provide performance guarantees for critical mMTC.A context-aware, risk-sensitive formulation is suggested for allocating resources to delay-sensitive services.
  • Cooperative transmissions: Time-varying-channel RAN slicing must trade modulation-and-coding level against resource-block count, while cross-slice spectrum sharing addresses heterogeneous critical-mMTC requirements.Higher modulation and coding can improve spectral efficiency while requiring fewer resource blocks, but cross-slice sharing remains challenging.

E. Network and Higher Layers Approach

The paper identifies higher-layer mechanisms for supporting heterogeneous delay-sensitive and delay-tolerant devices in critical mMTC. Proposed directions span virtualization, distributed access control, cloud-edge cooperation, clustering, and coded multi-connectivity.

  • E. Network and Higher Layers Approach: NFV and SDN can divide a physical IoT network into virtual networks and allocate resources according to URLLC and mMTC requirements.A central controller provides a programmable, comprehensive view for updating resources across virtual networks.
  • E. Network and Higher Layers Approach: Distributed parallel queuing can eliminate backoff periods and data-packet collisions while stabilizing traffic and reducing control-signaling requirements.The approach is presented as an alternative to ALOHA-based RACH methods, which suffer from inefficiency, uncertainty, and instability.
  • E. Network and Higher Layers Approach: A tightly coupled cloud-edge architecture is proposed to combine the cloud’s capacity with the edge’s suitability for low-latency URLLC applications.Cloud round-trip delay limits direct applicability to critical mMTC, whereas edge computing has lower computing capacity and memory.
  • E. Network and Higher Layers Approach: Clustering devices by service and geolocation, followed by gateway packet segregation, can ameliorate RAN congestion and support energy-efficient critical-device operation.The approach is also linked to critical NB-IoT and delay-sensitive devices.
  • E. Network and Higher Layers Approach: Packet cloning, multiple connectivity, and network coding can compensate for lost packets across links without requiring long detection and retransmission waits.Conventional block codes such as Reed-Solomon codes are identified as a way to reduce storage costs relative to caching and cloning.
Loading 2007.08793v1…