5 g network white paper
0 likes219 views
5G networks will require architectural changes to support new capabilities and use cases. Key changes include adopting a cloud-native architecture with network softwarization using NFV, SDN, and network slicing. This will allow the network to be controlled by software and separated into multiple virtual networks. The 5G radio access network architecture will also change with the introduction of cloud-RAN to replace distributed base stations and reduce small cell deployment costs. Network slicing will enable logical isolation of network resources to provide different services on the same physical network, such as enterprise, OTT, and MVNO services.
5 g network white paper
- 1. 5G Networks Architecture and Operations Ravi Sharma Dec, 2018 5G NETWORKS !1 5G Network for future
- 2. 1.Wireless Networks Evolution 1.Wireless Networks Evolution 2 2. 5G Network - use cases: Mobility to Industry 4 3. 5G Networks KPI’s - Industrial grade reliability 6 4. 5G Architectural Changes & challenges 7 A. Cloud native Architecture 8 B. Network slicing - 5G Services Architecture 9 C. RAN Architecture Changes 10 D. Mobile edge computing 11 5. New horizons - AI/ML lead Operations 12 A. Zero Touch Services Orchestration 13 B. Cognitive Network Operations 16 C. Closed Loop Automation 18 6. Conclusion 19 5G NETWORKS !2
- 3. Wireless Network has evolved over the years from only voice(2G) to voice over IP (4G LTE) networks. The traditional telcos are transforming towards CSP (communication service providers), moving from traditional business of voice and data, to OTT Apps ecosystem for mobile payments and infotainment . In the latter part of 2000’s Internet has become the only solution for all day to activity with advent of Smartphones on 3G/4G connectivity from mere mail solution on 1990’s. Today NLP of Alexa and Google are moving the day to day life of average individual tightly bound to mobile only ecosystem. Video traffic on mobile network are increasing with the 3G, 4G becoming the de-facto mobile connectivity around the world. According to ericsson mobility report 2018, 4G technology will become most dominant wireless access by 2023. The wireless standard for new 5G NR is also getting finalised and it will be available for mass deployment by end of 2019, the limited trails are taking place globally in 2018 with early basestation and MIMO designs. The video traffic is going to be dominant with advent of more streaming &. broadcasting services, augmented reality videos, virtual reality headsets, increasing use of video calling and video surveillance solutions etc. The 5G technology is bound to increase trend. The terms exabyte is coined for growth of mobile internet traffic. (Source: Ericsson mobility report 2018 ) 5G NETWORKS !3 Source : Ericsson Mobility Report, 2018
- 4. 2. 5G Network - Use Cases: Mobility To Industry In the new era of Digital Service Providers (DSPs), new requirements challenges service providers on legacy networks in terms of technologies and business models. Thus arising a need for more mature wireless access to support people's demand for digital lifestyle, and focuses on services which needs higher bandwidth, such as high definition (4K) videos, virtual reality (VR), and augmented reality (AR). Considering the new advancement in media and platform services, International Telecommunication Union (ITU) has classified 5G NR mobile network services into three categories: 1. Enhanced Mobile Broadband (eMBB)-allowing seamless customer experience from a broadband speed point of view and the rise of new applications on-the-go such as UHD Augmented Reality; 2. Ultra-reliable and Low-latency Communications (uRLLC) - essentially enabling the smart cities of the future by using internet of things (IOT) and Industrial IOT; 5G NETWORKS !4 Source: Ericsson Mobility Report, 2017
- 5. 3. Massive Machine Type Communications (mMTC)- a must for mass deployment of autonomous car types of applications. 5G goals are aligned with the Operators vision globally where they want to transform them self into a digital services provider by focusing on 1) Communication services, 2) Media, lifestyle apps & Content based services, 3) IOT services and 4) Cloud Based services • Communication Services: The exponential growth in connected devices and bandwidth will still remain a major driving factor while consumers expect to have a significantly enhanced differentiated user experience. 5G NETWORKS !5 Source: ITU - 5G White Paper
- 6. • Media & Life style Apps : will be at the core of future digital services. These services include all rich content oriented new services and business models, such as, 4K/8K video, online gaming, Virtual Reality, etc. • IOT Services: IOT Services will bring innovations within various industry segments it serves with pre-packaged vertical solutions. • Cloud Based Services: The 5G architecture will enable transformation in network and introduce Network slicing to enable new active sharing among Operators. It will also push uCPE and SDWAN as whole infrastructure will be transform on telco cloud. 5G standards are evolving and the changes it brought to the networking ecosystems are still evolving. ITU’s plan for 5G standard finalisation are show below. It is expected to have mass deployment from early 2021-22. 3. 5G Networks KPI’S - Industrial Grade Reliability The ITU has also introduced the key aspirations for 5G technology over and above LTE 4G technology. Many of them will rely on enhanced and more 5G NETWORKS !6
- 7. efficient management of the network, in addition to improvements in underlying technologies and Radio Access bandwidth. • 1000 times higher mobile data volume per geographical area. Approx 10-25 Gb/Site • 10 to 100 times more connected devices. 1 Million/Km2 • 10 times to 100 times higher typical user data rate. • 10 times lower energy consumption. • End-to-End latency of < 5ms, in air latency <1 ms • Ubiquitous 5G access including in low density areas. • Reliability of 99.999% • Services Deployment time to be reduced by 1000 in comparison with the current 4G system apron 90 minutes The 5G KPIs ensures that it can be used for not only connectivity consumers but can easily be suitable for Industry users. The availability, reliability and latency are inline with industrial use cases of autonomous cars, Industrial IOT driving big factories etc. At the same time higher throughput and 3X spectrum efficiency ensure seamless coverage for each and every users. 4. 5G Architectural Changes & Challenges 5G NETWORKS !7
- 8. The 5G KPIs and use case need the architecture changes from current 4G RAN and Backhaul Network. 5G will completely revolutionised mobile networks for accommodating the ever increasing data demands of users, services and application. 5G require much complex management requirements based on the softwarization of network resources. The ITU has defined a cloud native architecture as a base line for 5G network. This has been achieved by control plane user plane CUPS architecture. The control and user plane separation divides complex control logic functions for convergence into control planes, which reduces the costs of distributed edge gateways(routers/switches/BBU etc.) deployment, interface load, and number of alternative signaling routes. In addition, the control plane and user plane separation supports scaling of the forwarding and control planes, which further improves network architecture flexibility, facilitates centralized control logic functions, and ensures easy network slicing for diversified industry applications. ITU IMT Vision for 5G are shown in below picture which explains the three main drivers 1)Softwarization 2)Flexibility & 3)Costomization To achieve above following are the changes discussed in details. A. Cloud native Architecture End-to-end flexibility will be one of the defining features of 5G networks.This flexibility will result in large part from the introduction of network softwarization where the core network hardware and the software functions are separated. Network softwarization – through network functional virtualization (NFV), software defined networking (SDN), network 5G NETWORKS !8
- 9. slicing and Cloud-RAN (C-RAN) – aims to increase both the pace of innovation and the pace at which mobile networks can be transformed. • NFV – replaces network functions on dedicated appliances – such as routers, load balancers, and firewalls, with virtualized instances running on commercial off-the-shelf hardware, reducing the cost of network changes and upgrades. • SDN – allows the dynamic reconfiguration of network elements in real-time, enabling 5G networks to be controlled by software rather than hardware, improving network resilience, performance and quality of service. • Network slicing – permits a physical network to be separated into multiple virtual networks (logical segments) that can support different RANs or several types of services for certain customer segments, greatly reducing network construction costs by using communication channels more efficiently. • C-RAN – is presented as a key disruptive technology, vital to the realization of 5G networks. It is a cloud-based radio network architecture that uses virtualization techniques combined with centralized processing units, replacing the distributed signal processing units at mobile base stations and reducing the cost of deploying dense mobile networks based on small cells. B. Network slicing - 5G Services Architecture The network slicing is emerging as a future-proof framework for accommodating to the technological and business needs of different industries and respective services. 5G network slicing enables logical isolation of programmable infrastructure resources (i.e., physical and/or logical resources) to configure functions and services. The network slicing will require the upgrade of existing 4G network to cloud based infrastructure based on SDN & NFV. For achieving the service orchestration in cloud based network ETSI, has been developing MANO architecture which is detailed in next section. 5G NETWORKS !9
- 10. ! A A network slice can provide the functionality of a complete network, including radio access network functions and core network functions (e.g., potentially from different vendors). One network can support one or several network slices. Same is explained in picture above where the network is divided into three slices of Enterprise , OTT and MVNO and providing QoS for each service seperatly on same network. C. RAN Architecture Changes The 5G RAN architecture has several changes due to the KPIs required from RAN, i.e higher throughput, lower latency, lower energy consumption, high device density and ubiquitous coverage. Small Cell The deployment of small cells is one way of boosting the capacity and quality of existing 4G networks while laying the foundation for commercial 5G networks and early eMBB services. Small cells are already being used by some wireless operators to boost the capacity and coverage of their existing 4G networks particularly in a dense urban setting Massive MIMO & Beam forming Massive MIMO provides independent narrow beams targeted at multiple users and transmits data through a user-spe- cific space isolation system. This helps increase system throughput by 5G NETWORKS !10
- 11. dozens of times. Leading operators around the globe have already begun deploying commercial Massive MIMO. It support of beamforming, essential for efficient power transmission. Massive MIMO increases spectral efficiency and in conjunction with dense small cell deployment, will help operators to meet the challenging capacity requirement of 5G. Fronthaul In current 4G wireless network, the fronthaul link exists between radio frequency (RF) func- tion and the remaining layer 1, 2 and 3 (L1/L2/L3) functions. Recommendation ITU-T Y. 3100 defines fronthaul as “a network path between centralized radio controllers and remote radio units (RRU) of a base station function”. This architecture allows for most stringent fronthaul latency and bandwidth requirements. The increase in data rates in 5G makes it impractical to continue with the conventional Common Public Radio Interface (CPRI) fronthaul implementation. Thus following technologies are getting popular as frontal options 1) XGSPON 2) CWDM 3)25 GE/50GE ethernet 4) OTN transport. The concept of “Anyhaul” is also progressing which talks about 5G Antenna active antenna unit (AAU) getting connected to 5G DU on Fronthal and DU to CloudRAN on Midhaul. The conventional Backahul will be used to connect the CloudRAN to MEC and Core Cloud. D. Mobile edge computing In-network content caching provided by the operator, a 3rd party or both, can improve user experience, reduce backhaul resource usage and utilize radio resource efficiently. 5G NETWORKS !11
- 12. The operation of in-network caching includes flexible management of the location of the content cache within the network and efficient delivery of content to and from the appropriate content caching application. Fox example : CCTV camera live feed is monitored and processed at the MEC server near the LTE Base Station. Only video clips, meta data, triggers etc. which are low bandwidth are sent to the core network. This improves the response time and reduces the load on the backhaul.One application could be to public safety, Smart cities etc. MEC benefits can be summed up but not limited to : 1) (Ultra-)low latency: disruptive improvement of customer experience, 2) Reduction of backhaul/core network traffic: cloud services (e.g., big data) near to user & 3) In-network data processing 5. New Horizons - AI/ML Lead Operations In the 5G era and on the journey towards 5G, there is a need for a true and unified network management offering that can manage complex hybrid networks including 5G, all physical and virtual network elements, as well as VNF lifecycle management. To understand the requirements better, here are five key essentials required to win in 5G network management. AI is making the leap from use cases that mimic human behaviors to large complex systems that leverage human capabilities. Within the field of AI, there has been rapid progress in machine intelligence, a discipline which augments the structuring and modeling of machine learning with reasoning and planning techniques. • Managing the increasing complexity of networks • Machine learning plus reasoning and planning is machine intelligence • Enabling machine intelligence through structured knowledge The operators are focusing on real time operations with respect to operation support system OSS migration for 5G. The basic • Distributed real-time orchestration of diverse complex services across all network domains, including vCore and intelligent edge (e.g. MEC) • End-to-end management of hybrid virtual and physical cross-generation access, aggregation, and core infrastructure 5G NETWORKS !12
- 13. • Real-time QoS management with closed-loop and analytics-driven service assurance tools • Dedicated M2M/ IoT Platform for vendor- agnostic device and data management A. Zero Touch Services Orchestration Zero touch network and Service Management is conceived as a next-generation management system that leverages the principles of Network Functions Virtualization (NFV) and Software Defined Networking (SDN). It will be designed for the new, cloud-based network infrastructures and functions, and based on cloud-native principles to address zero- touch (fully automated) management and operation. Open network automation platform ONAP is a linux foundation initiative which is working on a NFV, SDN automation architecture using open standards and taken a lead by combining operators and vendors ecosystem. They have ready to deploy VNFs to create and manage the underlying vEPC and vIMS services by interworking with vendor-specific components, including VNFMs, EMSs, VIMs and SDN controllers, across Edge Data Centers and a Core Date Center. 5G Network is programmable, software-driven, service-based and holistically-managed infrastructures, utilising enablers and catalysts, such as NFV, SDN and Edge Computing. Metro ethernet forum (MEF) has defined the standard for zero touch orchestration and customer journey for the services such as IoT, MVNO etc. Several other standard are also coming into such as ZOOM from TM forum. The MEF has defined its zero touch requirement in the life cycle service Orchestration (LSO) specifications, following journeys are covered with architecture references in LSO : 5G NETWORKS !13
- 14. 1- Agile methodology for service provision life cycle 2- Order fulfilment and Services Control 3- Security Management 4- Analytics Managment 5- Billing & usages management 5- Standards APIs for customer, service provider and partner domains. 5G NETWORKS !14
- 15. Oracle, Juniper and Infovista has demonstrated LSO based architecture for vCPE deployment over SDN based infrastructure. Similarly other services can be used and tightly coupled into ETSI MANO architecture ETSI MANO Open Source Mano is an ETSI-hosted initiative to develop an Open Source NFV Management and Orchestration (MANO) software stack aligned with ETSI NFV. The MANO architecture provides VNF definition, use case and network slicing orchestration architecture. MANO also details about virtual network function VNF management. “VNF” refers to the implementation of a network function using a virtual machine that lies on top of the underlying hardware.Effective lifecycle management of these VNFs is essential for comprehensive network management of a complex hybrid network that may include 4G, Enterprise MEN & 5G. MANO Functional Blocks • NFV Orchestrator: • on-boarding of new Network Service (NS), VNF-FG and VNF Packages 5G NETWORKS !15
- 16. • NS lifecycle management (including instantiation, scale-out/in, performance measurements, event correlation, termination) • global resource management, validation and authorization of NFVI resource requests • policy management for NS instances • VNF Manager: • lifecycle management of VNF instances • overall coordination and adaptation role for configuration and event reporting between NFVI and the E/NMS • Virtualised Infrastructure Manager (VIM): • controlling and managing the NFVI compute, storage and network resources, within one operator’s infrastructure sub-domain • collection and forwarding of performance measurements and events B. Cognitive Network Operations The Cognitive NOC has three main focus are according the challenges the operators are facing now • Real-time in measuring, detecting and resolving problems; • Proactive, predictive and prescriptive; • Automated in service provisioning and fulfillment; • Agile and adaptive to frequent configuration changes; and • Distributed in implementation to effectively realize the "recover first before trouble- shooting" approach The current challenges for a NOC solution are following • No single-pane-of-glass view or correlation across subsystems, leading to a lack of network-wide context • An excessive number of events, noise, and duplication, leading to too many non- actionable trouble tickets • A lack of support for collaboration, leading to domain-specific teams troubleshooting in silos • Reactive workflow – too much time spent on customer-reported faults, leaving insuf- ficient time for proactively preventing faults The only way that these difficult goals can be achieved is Future OSS with vastly increased automation in management and operation of 5G network 5G NETWORKS !16
- 17. Machine learning is the process by which a computer-driven machine can learn pattern recognition, make predictions, and provide actionable insights through sophisticated algorithms. It offers the opportunity to increase the level of automation in network management even further. The data already available in the network can be turned into insights on how to prioritize and fix network faults as quickly as possible. And this allow for significantly reduced investments in designing and maintaining automation solutions. The prototype NOC software enables automatic fault management by applying machine intelligence techniques. This enables it to: – Map composite conditions from historical information (performing intelligent grouping of cross-domain alarms for detection using pattern mining techniques) 5G NETWORKS !17
- 18. – Form rules from the composite conditions using machine learning – Detect incidents based on the rules – Identify root causes, and derive appropriate actions by mapping root causes to resolution procedures from system or solution documentation C. Closed Loop Automation A cornerstone in developing automated decision-making is knowledge collection and the subsequent organization of the knowledge into a graph of interlinked domains. Automation in activities such as generating actionable insights, network configuration, and ensuring network quality will be critical to ensure improved network performance and operational efficiency. Operators must prepare for the increased size and complexity of the network in the 5G and IoT era. The amount of opex spent on incident management in a network operations center (NOC) today will increase exponentially unless much higher efficiency is attained. Each NOC engineer will soon have to manage more than 10 times more complexity than today, and this challenge must be addressed with automation. Fault conditions and alarm grouping is made possible by • Embedding network information, such as alarms and events, into a telco knowledge graph which includes raw as well as insightful, derived information that forms the basis for enabling automated intelligent behaviors at the NOC • Automatically capturing the behaviors in the network data – alarms and events – in a data-driven manner into digitalized versions which we will refer to as machine learning (ML) generated rules. 5G NETWORKS !18
- 19. 6. Conclusion As operators plan for the deployment of new technologies such as 5G and NFV, the key criteria they are looking for from their service assurance solutions include: • Cross-domain (access edge, core, etc.) analysis with an end-to-end visualization of services across a hybrid of physical and virtualized network functions • Complete, real-time inventory/topology covering all functional elements • Combining bottom-up and top-down real-time inputs – infrastructure-level information such as SNMP traps can help determine the cause of a fault, while application-level information (gleaned from active monitoring) can help determine service health • Automated mapping of customer-reported issues to underlying infrastructure issues • Automated mapping of network events to their service impacts and prioritization of alarms according to customer impact level • Automated triggering of proactive, corrective actions (e.g., issuing requests to an orchestrator to resolve a performance degradation before it impacts the customer experience) Machine learning presents an interesting new toolset for service assurance. By automating the collection and analysis of monitoring telemetry (events, alerts, Syslog, SNMP traps, etc.), and using the power of machine learning, operators may be able to uncover new insights about their networks that enable them to reduce their cost of operations and improve the customer experience. The service assurance solution must also be closely integrated with fulfillment/ orchestration systems to deliver the "holy grail" of closed-loop automation. 5G NETWORKS !19