Grazie
5G experimental activities in Flex5Gware project
focus on RAN virtualization
Dario Sabella (Telecom Italia)
IEEE 5G Berlin Summit
Fraunhofer-Forum Berlin, November 2, 2016
Grazie focus on RAN virtualization Dario Sabella (Telecom Italia) - - PowerPoint PPT Presentation
IEEE 5G Berlin Summit Fraunhofer-Forum Berlin, November 2, 2016 5G experimental activities in Flex5Gware project Grazie focus on RAN virtualization Dario Sabella (Telecom Italia) Agenda Telecom Italia presentation Experimental
focus on RAN virtualization
Dario Sabella (Telecom Italia)
IEEE 5G Berlin Summit
Fraunhofer-Forum Berlin, November 2, 2016
2 IEEE 5G Summit (Berlin, November 2, 2016)
3 IEEE 5G Summit (Berlin, November 2, 2016)
4 IEEE 5G Summit (Berlin, November 2, 2016)
Values referred to 1H2016 http://www.telecomitalia.com/tit/en/about-us/profile.html http://www.telecomitalia.com/tit/it/investors/financials/employees.html
5 IEEE 5G Summit (Berlin, November 2, 2016)
Italy Latin america: TIM Brasil Group International Wholesale services
TI sparkle is an important global service provider, with activites on Voice, IP and data transmission for fixed and mobile operators, ISP and multinational societies (MNCs)
6 IEEE 5G Summit (Berlin, November 2, 2016)
https://www.ict-earth.eu/ https://www.metis2020.com/ http://www.ict-ijoin.eu/
7 IEEE 5G Summit (Berlin, November 2, 2016)
8 IEEE 5G Summit (Berlin, November 2, 2016)
Project number: 671563 Project Coordinator: Intel Technical Management: CTTC Call / topic: H2020-ICT-2014-2 /
ICT-14-2014Objective 1.1
Duration: 24 months Begin: 01 July 2015
«Flexible and efficient hardware/software platforms for 5G network elements and devices»
1. Intel Mobile Comm. (DE) 2. Alcatel Lucent (DE) 3. Ericsson (SE) 4. NEC (UK) 5. Telecom Italia (IT)
7. CEA (FR) 8. CNIT (IT) 9. CTTC (ES)
9. Sequans (FR)
9. KU Leuven (BE)
https://www.flex5gware.eu
9 IEEE 5G Summit (Berlin, November 2, 2016)
The overall objective of Flex5Gware is to deliver highly reconfigurable HW platforms together with HW-agnostic SW platforms targeting both network elements and devices taking into account increased capacity, reduced energy footprint, as well as scalability and modularity, to enable a smooth transition from 4G mobile wireless systems to 5G.
Objectives:
1.
Increasing HW versatility and reconfigurability
2.
Designing and developing HW-agnostic, flexible and cost-effective SW platforms
3.
Increasing the
achievable capacity provided by 5G communication platforms
4.
Decreasing the overall energy consumed by 5G communication platforms
5.
Identifying and prototyping key HW and SW building blocks and providing a proof
concept for all developments in Flex5Gware
Legenda
10 IEEE 5G Summit (Berlin, November 2, 2016)
WP6 Focus: integration activities for the integrated PoCs and validation, testing and demonstration for all the PoCs (both standalone and integrated).
11 IEEE 5G Summit (Berlin, November 2, 2016)
12
Innovative proof of concept PoC team (involved partners) Type
Developed in
PoC#1
On chip frequency generation
Standalone
WP2 PoC#2
Active SIW antennas with integrated power amplifiers for K/Ka frequency bands Integrated
WP2 PoC#3
PAPR reduction and power amplifier predistortion
Integrated
WP2, WP3 PoC#4
Multiband transmitter
Integrated
WP2, WP3 PoC#5
Full duplex FBMC transceiver
Integrated
WP3, WP4 PoC#6
High-speed low power resilient LDPC decoder Standalone
WP4 PoC#7
HW/SW function split for energy aware communications
Integrated
WP4, WP5 PoC#8
Reconfigurable programmable radio platform (terminal side) and SW programming performed and injected by the network Integrated
WP4, WP5 PoC#9 Flexible, scalable and reconfigurable small cell platform
Standalone
WP5 PoC#10
Flexible resource allocation in CRAN/vRAN platform
Standalone
WP5 PoC#11 Multi-Chain MIMO Transmitter
Integrated
WP2, WP3 PoC#1 PoC#2 PoC#3 PoC#4 PoC#5 PoC#6 PoC#7 PoC#8 PoC#9
PoC#10 PoC#11
13 IEEE 5G Summit (Berlin, November 2, 2016)
14 IEEE 5G Summit (Berlin, November 2, 2016)
Scenario
broadband connections.
Operators needs
coverage, capacity and QoS (Quality of Service). More attention to energy efficiency and
Emerging trends
market is already moving progressively toward virtual implementations of EPC, and recently operators are considering with increasing attention the centralization (C-RAN) and virtualization (NFV) of mobile networks functionalities in a standard IT platform.
Benefits
throughput by taking care of costs and utilization efficiency, and also energy efficiency and sustainability of the mobile network evolution.
network deployments, suitable for additional revenue generating services.
15 IEEE 5G Summit (Berlin, November 2, 2016)
As a consequence, a possibility for operators is to evaluate virtualization
But, before talking about vRAN, it is worth to consider C-RAN evolution
In
NOTE: C-RAN architecture can be applied for both macro-only and
16 IEEE 5G Summit (Berlin, November 2, 2016)
pRBS
Compacted architecture with
distributed BB
pRBS
Splitted architecture with
BB hotelling
C-RAN architecture
BB pooling
17 IEEE 5G Summit (Berlin, November 2, 2016)
Operating System. This is usually done by an “hypervisor”. Some functionalities are executed running as Virtual Machines (VMs). Examples of VMs in CRAN environment are represented by single RATs or by sub-systems of the protocol stack of a RAT.
eNB
UE Virtual BaseBand pool
RAP RAP RAP
E-UTRAN
18 IEEE 5G Summit (Berlin, November 2, 2016)
19 IEEE 5G Summit (Berlin, November 2, 2016)
20 IEEE 5G Summit (Berlin, November 2, 2016)
21 IEEE 5G Summit (Berlin, November 2, 2016)
Mobile edge computing (MEC) permits to extend cloud platform for applications to the edge of the network (e.g. Base Stations), in order to improve the user experience (also thanks to the standardization of proper measurements and network APIs).
HSS MME S/P-GW Application Server PCRF
eNB
UE Virtual BaseBand pool
RAP RAP RAP
EPC E-UTRAN + MEC server + MEC server Web
This view is not in contrast with C-RAN architecture: MEC will add flexible decentralization and proper dynamic instantiation of applications, in close proximity to terminals (that are also empowered with additional capabilities through computation offloading).
22 IEEE 5G Summit (Berlin, November 2, 2016)
23 IEEE 5G Summit (Berlin, November 2, 2016)
Name of the PoC (in collaboration with UniPisa):
resource allocation in CRAN/vRAN platform
eNB
UE Virtual BaseBand pool
RRH RRH RRH
E-UTRAN
Features:
robustness and reliability, using SotA optimization techniques to strike the best trade-offs between performance and energy, scale vs. resources Goal:
solutions in 5G networks for centralized RAN environments, through virtualization of BSs
Main ideas:
24 IEEE 5G Summit (Berlin, November 2, 2016)
USRP USRP PC Phone/dongle Phone/dongle USRP Phone/dongle
C-RAN/vRAN server used for the BBUs
(Micro and Macro). It contain also a Coordinator and the Intelligent Programs as developed in WP5.
considering an hexagonal grid (assuming 3 sectors per site);
The surrounding noise will be generated with additional LTE node (acting as intercell interference generator).
Name of the PoC (in collaboration with UniPisa):
25 IEEE 5G Summit (Berlin, November 2, 2016)
HSS Coordinator EPC (MME+GW) OAI eNodeB OAI eNodeB OAI eNodeB
ETH Switch
Eth Cable USB3.0 Cable Implemented in WP5. Integration within the PoC in next months.
Target KPI Related Use Case(s) Comment
Energy Consumption (NRG)
Dynamic Hotspots 50+ Mbps Everywhere
% of Energy saving in comparison with the solution where all the hotspots are transmitting.
User data rate (UDR)
Dynamic Hotspots 50+ Mbps Everywhere
User/cell edge Throughput (comparison with and without CoMP technique).
Flexibility, versatility, re-configurability (FVR)
Dynamic Hotspots 50+ Mbps Everywhere
Time requested to adapt the system to a change of network configuration (switch on/off of one or more cells and reallocation of the resources).
First evaluations planned according to the target KPIs
26 IEEE 5G Summit (Berlin, November 2, 2016)
Coordination: Flexible CoMP/ Coordinated Scheduling (Global Scheduler)
Efficiency: Flexible node activation/ deactivation (Global Power Manager)
show promising gains in terms of CQI, and cell tput. Next results from January.
Device (Remote) Network Controller Sensors / Actuators / Hardware
Dynamic Functional Re-composition (DFR) Monitoring Service
Performance- aware Resource Management
Monitoring library
Modular and flexible node operation Multi-node coordination Intelligent programs
Radio Engine Radio Program Local Resource Controller Radio Engine Manager Monitoring Agent
Global Scheduler
Local Scheduler Local Power Manager
Global Power Manager Parameter Value
Carrier frequency 2 GHz Bandwidth 10 MHz (50 RBs) Path loss model ITU Urban Macro Fading model Jakes Macro eNB Tx Power 40 dB Micro eNB Tx Power 26 dB eNB antenna gain 18 dB Noise figure 5 dB Cable loss 2 dB UE mobility model Stationary UE traffic model CBR, 160 Kbps
27 IEEE 5G Summit (Berlin, November 2, 2016)
28 IEEE 5G Summit (Berlin, November 2, 2016)
virtual RAN) should move toward a fully virtualized network, including the addition of mobile edge computing technologies.
well as Flex5Gware PoCs during the final demonstration event (Turin, June 2017).
facilities and TI Test Plant in Turin, as already done in the past for the EARTH project demonstrator.
Turi urin, , 21-22 June
ne 2017
dario.sabella@telecomitalia.it