Mobile Edge Computing (MEC) in 5G network Presenter: Ching-Chun Chou Advisor: Hung-Yu Wei 2018/12/03
Outline Mobile Edge Computing (MEC) Wireless communication MEC in 5G Definition & elements MEC platforms Wireless communication Generation of wireless communication standard Towards 5G Between MEC and 5G MEC in 5G ETSI specification: framework and scenarios
Mobile edge computing
What is MEC? Mobile Computing + Edge Computing Mobile computing Computing to be transported during usage Mobile communication + hardware + software Especially mobile devices Edge computing Computing to be hosted close to the UE’s access
Edge Computing Service to be hosted close to the UE’s access point of attachment Achieve an efficient service delivery Reduced end-to-end latency and load on the transport network Centralized Cloud Model Edge-centric computing Pedro Garcia Lopez, et al., “Edge-centric Computing: Vision and Challenges,” SIGCOMM Computer Communication. Rev. 45, 5, 37-42, September 2015
Open EDGE Computing Jointly initiated by Intel, Nokia, Deutsche Telecom, Vodafone, NTT, Carnegie Mellon Univ., Crown Castle http://openedgecomputing.org/
Vodafone’s Demo on MWC 2018 Improving video streaming user experience Time to start/number of stall/waiting time Saguna edge computing platform https://www.saguna.net/blog/improving-video-streaming-customer-experience-with-multi-access-edge-computing-research-results-from-vodafone-and-saguna/
Multi-Access Edge Computing from Nokia Real-time Interactive Secure Analytical Distributed https://networks.nokia.com/solutions/multi-access-edge-computing
Nokia Edge Cloud Scalable data center for work balance between cloud and edge computing https://networks.nokia.com/solutions/edge-cloud
Edge Computing Platform from NTT Locating edge servers in the vicinity of users/devices Passing some of the load of the center servers and/or user devices http://www.ntt.co.jp/news2014/1401e/140123a.html
IoT Edge Computing Massive distributed computing system https://www.postscapes.com/iot-edge-computing-software/
Elements in Edge Computing To be performed in the edge Proximity Physical and logical Intelligence Autonomous decision making Trust Personal and social sensitive data Control Management and coordination Humans Human-centered design for applications Pedro Garcia Lopez, et al., “Edge-centric Computing: Vision and Challenges,” SIGCOMM Computer Communication. Rev. 45, 5, 37-42, September 2015
Wireless communication With 5G Networks Wireless communication
IMT-2020 Specified by ITU-R International Telecommunication Union Radiocommunication Sector The requirement for the next generation mobile communication Enhanced Mobile Broadband (eMBB) Ultra-Reliable and Low Latency Communication (URLLC) Massive Machine Type Communication (mMTC) ITU-R M.1645
IMT-2020 Vision 5G usage scenarios ITU-R M.1645
5G Capability IMT-Advanced (4G) VS. IMT-2020 (5G) ITU-R M.1645
What is Proposed? Let’s take 4G for example IMT-Advanced: the so-called 4G WiMAX standard Developed by the WiMAX forum Technology specification from IEEE 802.16 series LTE standard Developed by the 3GPP Technology specification from the 3GPP Releases 5G is the 3GPP Release-15 & Release-16
3rd Generation Partnership Project The 3GPP is composed of 7 organizations ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC With market presenters
3GPP Standard Releases Note that 3G is IMT-2000
From the LTE System to 5G 5G is forward compatibility Connectivity between 5G and LTE
3GPP Timeline for 5G By the end of 2019 To be submitted to ITU-R 2017 2018 2019 2020 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Rel-15 NSA (option-3) freeze Rel-15 NSA (option-3) ASN.1 Rel-15 SA (option-2) freeze Rel-15 SA (option-2) ASN.1 Rel-15 late drop freeze Rel-15 late drop ASN.1 Rel-16 SI Rel-16 SI/WI phase Rel-16 freeze Rel-16 ASN.1
Service and Requirements Everything connection More applications Service extension Capability extended
Between MEC and 5G Mobile communication + Edge Computing Can the requirement of MEC be handled by 5G? Let’s look at AR & VR VR/AR average 5.6Mbps 7.1% of global communication above 25Mbps Mushroom networks, “Bandwidth requiremnet for VR and AR, infographic 2017”
VR and AR Requirement Qualcomm whitepaper, “VR and AR pushing connectivity limits,” October 2018
The framework and usage scenarios MEC in 5G
ETSI ISG – Mobile Edge Computing Initiated in Oct. 2014 Huawei, IBM, Intel, Nokia, NTT DoCoMo, Vodafone Defining Use cases Deployment scenarios Infrastructure APIs in June, 2017 Whitepaper in June, 2018
3GPP Key Enablers for Edge Computing Local routing and traffic steering A PDU session may have multiple N6 interfaces Application function to influence UPF (re)selection By Policy Control Function Indirectly by Network Exposure Function Session and Service Continuity (SSC) Mobile scenarios Support of Local Area Data Network Serving PLMN providing services to UEs by Tracking Area 3GPP TS 23.501 v15, 2018/03
5G Architecture and MEC Service based architecture (SBA) Interaction between the network function of 5G core networks and MEC entities
MEC Deployment Scenario LTE eNB RNC Radio Network Controller Multi-technology Cell aggregation Indoor/outdoor Local delivery
Physical Deployment Examples of MEC Physical computing resource may be chosen by the network operator Site facilities, supported applications and their requirements, measured/estimated user load Other technical or business factors
MEC Use Cases Active device location tracking Augmented reality content delivery Video analytics RAN-aware content optimization Distributed content and DNS caching Application-aware performance optimization Mobile-Edge Computing – Introductory Technical White Paper, http://www.etsi.org
Use Case 1: Active Device Location Tracking Enable real-time, network measurement based tracking of active terminal equipment Mobile advertising, Smart City, campus management, etc.
Use Case 2: Augmented Reality Content Delivery Minimize round-trip time and maximize throughput for quality of experience Consumer propositions, local object tracking, local AR content caching
Use Case 3: Video Analytics Process the video data to detect/notify specific configured events Safety, public security, smart cities Video data → low-bandwidth metadata
Use Case 4: RAN-aware Content Optimization Improving QoE Dynamic configuration Enable new service and revenue opportunities Promoted content delivery and subscriber throughput boosting
Use Case 5: Distributed Content and DNS Caching Central cache and local cache Reduce backhaul transport Improve QoE Content caching: up to 35% DNS caching: 20%
Use Case 6: Application-aware Performance Optimization Increase browsing throughput and reduce video stalling Latency may also be reduced Provide independent metrics on application performance Video stalls, browsing throughput, latency, etc.
MEC Application Mobility Detection of UE mobility MEC subscribes to event notification through NEF May also include RNIS Validation o of application mobility User context transfer and/or application instance relocation Post-processing of application relocation
MEC Server Architecture Hosting Managing VM IaaS PaaS
Scope of ISG MEC Platform services and APIs Virtual Machine (VM) SLA Communication Service Service registry RNIS TOF Virtual Machine (VM) SLA Description and negotiation Supervision and trouble shooting MEC application platform and management interface
MEC Application Platform Service Infrastructure service Communication service Service registry Radio Network Information Service (RNIS) Traffic Offload Function (TOF)
Infrastructure Service Service-Oriented-Architecture (SOA) Communication service Application on a MEC server communicate with application platform service and each other Through service-specific APIs Service registry Visibility of service on the MEC server Authentication and authorization control
Radio Network Information Service (RNIS) Provide authorized applications with low-level radio and network information Deliver information from the radio network relating to users and cells Provide indications relating to the activation of UE Parameters of UE context and E-RAB Indicate modification Measurement and statistics information
Traffic Offload Function (TOF) Prioritize traffic Route selected, policy-based, user-data stream To and from applications that are authorized to receive the data. Application is supplied using Pass through mode End-point mode Set filters at the E-RAB and packet level E-RAB: SPID, QCI, ARP Packet: (UE IP, Network IP, IP protocol)
3GPP Standardization Status TS 23.501 System Architecture for the 5G System TS 23.502 Procedures for the 5G System Select a UPF close to the UE and executes the traffic steering from the UPF to the local network Via N6 interface Based on UE’s subscription data, location, info from AF
Functionality Supporting for Edge Computing User plane (re)selection (re)selects UPF to route the user traffic to the local Data Network Local Routing and Traffic Steering selects the traffic to be routed to the applications in the local Data Network Session and service continuity enable UE and application mobility Application Function influence UPF (re)selection and traffic routing via PCF or NEF Network capability exposure provide information to each other via NEF QoS and Charging PCF provides rules for QoS Control and Charging for the traffic routed to the local Data Network Support of Local Area Data Network support to connect to the LADN in a certain area where the applications are deployed
User Plane (re)Selection Performed by the SMF by considering UPF deployment scenario Centralized/distributed UPF close to/at the access network Enable different capability Home routed roaming
Local Routing and Traffic Steering Traffic offloading PDU sessions simultaneously correspond to multiple N6 interface SMF control the data path
Session and Service Continuity Multi-homed PDU sessions A PDU session associated with multiple IPv6 prefixes Branching Point provides forwarding
Application Function Influence on traffic routing For non-roaming and to LBO deployments Third-party application which the VPLMN has an agreement AF may send request to influence SMF routing decisions for traffic of a PDU session Allow routing user traffic to a local access to a data network May use NEF to interact with the core
Network capability exposure AMF service operations information flow UDM service operations information flow NEF service operations information flow
Support of Local Area Data Network LADN service area is a set of Tracking Areas Access to a DN via PDU session in LADN AMF provides to the UE the LADN information AMF tracks and informs SMF
MEC in Local DN Capability exposure MEC orchestrator as a 5G AF Interact with NEF and relevant NF Monitoring, provisioning, policy and charging MEC host at the edge of 5G RAN Need the RAN information
Conclusion Edge computing development status ETSI ISG MEC 3GPP standardization activities Potential applications and use cases for MEC
Q & A Thanks for your listening