Transcript Agenda of CPBW Workshop (3/`04)

Parul Gupta, Smruti Sarangi, Shivkumar Kalyanaraman [IBM Research – India]
Zhen Bo Zhu, Lin Chen, Yong Hua Lin, Ling Shao [IBM Research – China]
The Future of Software Radio:
Wireless Network Cloud
© 2009 IBM Corporation
Outline
 Cellular wireless systems and convergence trends (esp in emerging markets)
 Today’s 2G/3G architectures and the trend towards 4G (all-IP and OFDMA for PHY/MAC
layers)
 Increasing computational costs, low utilization with future generations of technology,
upgrade cycles, growing maintenance / OPEX costs.
 SDR evolution:
– Firmware OTA upgrades on vendor platforms,
– From DSP/FPGA platforms to Hybrid IT platforms,
– Multi-technology / multi-operator support
– Virtualization & cloud given fiber-to-tower availability
2
Version 0.1
© 2009 IBM Corporation
Wireless: short technical summary
Rate
4G
802.11b
WLAN
3G
2G
1. Scarce bandwidth
2. Spectral Efficiency:
(10-100 MHz/operator) MHz -> Mbps (signal to noise ratio is key!)
Other Tradeoffs:
Rate vs. Coverage
Rate vs. Delay
Rate vs. Cost
Rate vs. Energy
2G Cellular
Mobility
3. Tradeoffs: Rate vs X
(no free lunch!)
Today
With femto cells & MIMO
antennas
Wireless networks are designed to maximize spectral efficiency, support mobility,
coverage, and Quality-of-Service under severe spectrum/bandwidth constraints
3
In emerging markets the dependence on wireless is high and growing.
ARPU pressures are severe: $2 incremental ARPU today.
Operators aggressively outsource their (wireless/wired) networks and IT.
Version 0.1
2009 IBM Corporation
Wireless IT©convergence
Wireless Convergence: Closed Vertically Integrated to Horizontal Open
Circuit Voice;
Data overlay
Integrated MAC,
network functions
Apps: Unified Comms (multimedia),
Smarter Planet, mobile VAS, Internet
apps, Spoken Web/SMS platforms…
Middleware: IMS, SDP
Convergence at
Application/Solutions Level
Convergence at
Network Services Level
Network layer: IP
TDMA/CDMA
Radio layer: OFDMA/MIMO
2G/3G wireless
Convergence at
Systems Level
4G wireless
 System: IT and wireless systems are approaching similar system architectures
 Network Services: Convergence of enterprise wireless and operator wireless services
 Application/Solutions: Seamless integration and interaction of wireless infrastructure and
mobile applications
There is another interesting twist in this wireless / IT convergence at the systems level
4
Version 0.1
© 2009 IBM Corporation
2G-3G
4G
Wireless
wireless
Network
network
over
architecture
Wireless Network Cloud
Cloud
Access
of Network
Wireless Access Network
Core Network
+ Core Network
Mobile switch center
SMS/MMS
SMS/MMS
controller
Content Service
Edge gateway
Management
Server
BS
BS
Service Network
IMS WAP GW
WiMAX
GSM
GSM
BS cluster
Radio network
TD-SCDMA
BS
PSTN
Service support
Radio network
Gateway
node Edge
controller
Billing
gateway
Web Service
Internet
BS
5
LTE
LTE
WiMAX
WiMAX
BS cluster
Version 0.1
© 2009 IBM Corporation
2G/3G/4G Wireless over Wireless Network Cloud
Cloud of Wireless Access Network + Core Network
PSTN
Service Network
SMS/MMS
Service on Edge
IMS
WiMAX
GSM
GSM
TD-SCDMA
BS cluster
Content Service
Edge gateway
Management
Server
Billing
Web Service
Service support
node Edge
gateway
Internet
LTE
LTE
WiMAX
WiMAX
BS cluster
6
Version 0.1
© 2009 IBM Corporation
Mobile Infrastructure Network Hierarchy
Radio Access Network
7
Core Network
Challenges: The radio access network is a costly and continuous investment ($100B+).
With reducing ARPUs and need for broader nationwide coverage, there are more
initiatives for sharing infrastructure
© 2009 IBM Corporation
Version 0.1
Various Forms of Infrastructure Sharing in Wireless Networks
Network Sharing (eg: rural)
Base Station Sharing (leads to cloud)
Owner #1
Network
Owner #1
Retail
MSC
BSC
BSC
SDR
RRU
BTS
Owner #2
Network
Owner #2
Retail
Base Band Unit
BSC
BTS
Antenna Sharing
Owner #1
Network
Tower Sharing (very popular)
BSC
Owner #1
Network
BTS
BSC
BTS
BSC
BTS
O
Owner #2
Network
8
BSC
BTS
Version 0.1
Owner #2
Network
© 2009 IBM Corporation
Towers: Passive vs Active Infrastructure Sharing
9
Version 0.1
© 2009 IBM Corporation
Passive Sharing and Tower Companies
Eg: Indus Towers (JV controls towers of Bharti, Vodafone, Idea) has 100K towers.
Tata Teleservices, Aircel have signed deals with BSNL for sharing 60K+ towers.
10
Version 0.1
© 2009 IBM Corporation
Towards Active Sharing: Unbundling Base Stations: RRU + BBU
 Distributed base station
– RRU (Remote Radio Unit)
– BBU (Base Band Unit)
 Two key standards enable distributed base
station development
– CPRI
– OBSAI
Traditional Integrated Macro BS
 Benefits of distributed base station
– Reduce cost of facilitate infrastructure
– Reduce power consumption
– Easy of installation
– Flexible deployment model
RRU
BBU
Distributed BS: RRU + BBU
11
Version 0.1
© 2009 IBM Corporation
Distributed base station deployment #1: under the tower
Scenario #1: Unbundle at the tower
MSC
RRU
RRU
<100m
BSC
5-10Km
5-10Km
BBU

RRU-BBU Distance <100m
BBU
70% - 80% power consumption is RRUs
 3 RRU: 100 – 150W/RRU
 1 BBU: 100W

Requirement & Challenges to BBU
 light weight < 10Kg
 small size (1U – 2U)
 low power consumption (< 100W)
12
Version 0.1
© 2009 IBM Corporation
Software Radio & Software Defined Radio: One way of BBU impln
13
Version 0.1
© 2009 IBM Corporation
Multi-Technology Software Radio: 1 BBU Bladecenter vs 5 boxes
14
Version 0.1
© 2009 IBM Corporation
Multi-Operator Base Station with Software Radio
15
Version 0.1
© 2009 IBM Corporation
Active Infrastructure Sharing: Field Trial in India (IBM/partner)
16
Version 0.1
© 2009 IBM Corporation
Unbundled SDR BS w/ Open Wireless Interfaces & IT Platforms
...
RRU adaptor
RRU
GPS module
Base band
processing server
(PHY, MAC, C&M)
...
Base band
processing server
(PHY, MAC, C&M)
GE switch
RRU adaptor
PCIe/IB switch
General purpose servers
E1/T1/STM-1
Base band
processing
accelerators
SWR Base Station
CPRI/
OBSAI/
Ir
17
IO & GPS
module
Version 0.1
GE/E1/
T1/STM-1
To RNC/ASN-GW/
AGW
© 2009 IBM Corporation
Distributed base station #2: distributed RRU + centralized BBU pool
Scenario #2: central deployment

RRU
MSC
BBU
BBU
BBU
BBU
Benefits
– Fit for super urban, urban with high
density of traffic
– Highly scalable
– Improve utilization by resource sharing
– Reduce management cost
RRU
10KM
BSC

Requirements & Challenges to BBU
– High density
– Resource sharing with BBU pool
– Low power consumption
BBU Pool

Key barriers:
– Fiber distance (<10Km)
– Increasing IO data throughput >10Gbs
with LTE
– Fiber construction cost
– Synchronization in long distance
network

Case in China:

World largest TD-SCDMA BBU pool

Max support 72 RRUs

Power: 400W
A city like Bangalore or Delhi could be served from <10-15 pooled sites.
18
Version 0.1
© 2009 IBM Corporation
Wireless Network Cloud: Convergence of IT Platforms, SDR & RRH,
Cloud Computing Principles & Fiber-to-the-tower
Software Radio Technology/
IT & Cloud Computing
Techniques
Hybrid IT Systems
GSM RF header
GSM RF header
Server for BS
GSM
WCDMA
Remote Radio Header Technology
Antenna + Remote
Radio Header
WCDMA RF header
WiMAX
GSM
Server for Access
GW
WCDMA
GSM
Resource
manager
LTE
WiMAX
WCDMA
Server for Access
GW
Timing Network
over IP/Eth
WCDMA RF header
Fiber (> 10Km)
BaseStation Pool
WiMAX RF header
LTE RF header
End-to-End IP Infrastructure in 4G
19
Version 0.1
© 2009 IBM Corporation
Wireless Network Cloud Potential: Distributed Interference Management.
Eg: Collaborative MIMO for Elastic Capacity Allocation
Joint processing
Multi-cell environment with
frequency reuse factor 1
Optical fiber
Optical fiber
interference
I
I
Optical fiber
I
I
I
I
 Multiple points collaborate to mitigate ICI
or align interference for cancellation.
20
Version 0.1
© 2009 IBM Corporation
An e2e Demo has been setup in IBM CRL/IRL ([email protected])
Collaboration of IBM China and India Labs:
Multiple base-stations on common IT platform, USRP & e2e flows
Demo Application
VoIP
VoIP Stream
Web Browsing
Internet
Web Browsing
WiMAX
BS
Edge
Gateway
WiMAX
BS
WiMAX
BS
RRH
(USRP)
SIP Server
RRH
(USRP)
MS
(laptop simulated)
WiMAX BS /Gateway
(multi-core server)
I-vieW
Soft-phone
Web Server
Multi-core Utilization Analysis Tool
21
Version 0.1
Radio Signal Analysis Tool
© 2009 IBM Corporation
BS SDR System Architecture
MAC: Software Components
RRH adaptor
RRH
...
RRH
RRH adaptor
RRH
Signal processing
server
(PHY) (P7/x86/Cell)
Local Switch Fabric
(GbE switch)
RRH
Local Switch Fabric
(PCIe Switch)
Prism based platform
...
Signal processing
server
(PHY) (P7/x86/Cell)
Wireless MAC modules
& PHY-MAC adaptors
BS edge router
Control & Manage
BS edge router
Control &
management
Switch of BS
system
MAC
instance
MAC
instance
MAC
instance
Adapter
Adapter
Adapter
To other BS chasis
To ASN gateway or core
network
To internet directly
MAC and adapter Stacks
Downlink
Uplink
Defragment &
unpacking
Msg.
Packets
Decryption
Msg.
MAC Management
Msg.
Scheduling
Fragment &
packing
Packets
Encryption
Packet Extract
Key technical challenges being addressed (IRL+CRL)
• How to map the wireless software radio (SWR)
stack/workload to massive multicore and hybrid
architectures?
• How to meet QoS and real-time requirements for the
VoIP application, especially since the wireless software
stack (such as PHY and MAC layer) will be implemented
in software using a regular OS?
Concatenation
UL Adapter
22
DL Adapter
Version 0.1
Fast path data
processing
Slow path
message
processing
© 2009 IBM Corporation
Laundry List of Challenges
 Difference from regular computational, storage cloud: “real-time computational cloud”
– Focus on very high performance, real-time synchronized behavior.
– Cooperative techniques require greater degree of rigor in performance management
– Wireless = Critical infrastructure. Availability / reliability equally important as real-time
performance support.
 Choice of underlying platforms: hybrid systems, commodity servers and mapping it to VMs
(eg: MAC VM may work well on system A, and PHY VM work well on system B).
– Need real-time virtual switches that can tie together such component VMs into a pipeline
(Network -> MAC -> PHY)
– Cooperative techniques require redesign of protocols / implementation
 Multi-tenancy, elastic provisioning of real-time resources, tracking performance / availability
risks (eg: 4-5 nines)
– Providing backup for virtual base-stations from multiple data center sites: “cloud” attribute.
 Helping the industry move from an integrated “box” model to a software + outsourced
services model.
– Hypothesis: ARPU pressures faced by providers will ultimately drive such a move.
Aggressive outsourcing happening in markets like India.
23
Version 0.1
© 2009 IBM Corporation
Summary
 Software radio is an emerging technology.
 The long term potential of software radio involves:
– Unbundling base stations into hardware, software, RF components.
– Application of IT platforms, open wireless interfaces to SDR; opening up a opensource community of developers
– Allowing flexibility for BS software to be virtualized, and consolidated into pools for
reduced CAPEX/OPEX, higher utilization and change business models
– Fiber to towers will allow pooling and application of the Cloud model.
 Wireless network cloud can provide new benefits.
– Elastic capacity allocation & higher utilization/lower costs
– Distributed Interference Management: Collaborative MIMO etc (5th Generation Wireless)
– Integration of edge-based services at the cloud site. Eg: caching, content delivery,
unified communications, enterprise app delivery, cloud-based application delivery etc
 Perfect storm of “cloud” challenges:
– Real-time, synchrony/performance-critical, ultra-high reliability requirements.
24
Version 0.1
© 2009 IBM Corporation
Terminology












25
RAN – Radio Access Network
CN – Core Network
BTS – Base Station = BBU + RRU
BSC – Base Station Controller
BBU – Base Band Unit
RRU – Remote Radio Unit
RNC – Radio Network Controller, BSC in 3G
NodeB – BTS in 3G
eNB – Base Station Node in LTE
LTE – Long Term Evolution (E-UTRAN)
AIPN – All IP Network
NEP – Network Equipment Provider
Version 0.1
© 2009 IBM Corporation