fronthaul architecture towards 5g -...
TRANSCRIPT
![Page 1: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/1.jpg)
Aleksandra Checko, MTI
8/22-24/2016
In collaboration with:
MTI Radiocomp: Andrijana Popovska
Avramova, Morten Høgdal, Georgios
Kardaras
DTU: Michael Berger, Henrik L. Christiansen
EU project HARP consortium
Fronthaul architecture towards 5G
Multiplexing gains analysis
Challenges/solutions for fronthaul network
![Page 2: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/2.jpg)
Compliance with IEEE Standards Policies and Procedures
Subclause 5.2.1 of the IEEE-SA Standards Board Bylaws states, "While participating in IEEE standards development activities, all participants...shall act in accordance with all applicable laws (nation-based and international), the IEEE Code of Ethics, and with IEEE Standards policies and procedures."
The contributor acknowledges and accepts that this contribution is subject to
• The IEEE Standards copyright policy as stated in the IEEE-SA Standards Board Bylaws, section 7, http://standards.ieee.org/develop/policies/bylaws/sect6-7.html#7, and the IEEE-SA Standards Board Operations Manual, section 6.1, http://standards.ieee.org/develop/policies/opman/sect6.html
• The IEEE Standards patent policy as stated in the IEEE-SA Standards Board Bylaws, section 6, http://standards.ieee.org/guides/bylaws/sect6-7.html#6, and the IEEE-SA Standards Board Operations Manual, section 6.3, http://standards.ieee.org/develop/policies/opman/sect6.html
2
![Page 3: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/3.jpg)
Fronthaul architecture towards 5G
Multiplexing gains analysis
Challenges/solutions for fronthaul network
Date: 2016-08-23
Author(s):
Name Affiliation Phone [optional] Email [optional]
Aleksandra CheckoMTI (Microelectronics
Technology, Inc.)
aleksandra.checko
@mtigroup.com
IEEE 1914.1 TFNGFI
Bomin Li ([email protected])
![Page 4: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/4.jpg)
5G requirements and challenges
Source: netmanias
To carry 20Gb/s as I&Q signals (64-QAM ¾) over CPRI will require a CPRI data rate of at least 325Gb/s
Not acceptable for NGFI
New functional split is needed
– User processing in BBU
– Cell processing in RRH
– Split functions, still benefit from C-RAN (e.g. CoMP)
Variable bit rates on fronthaul
Agenda for today
– How big are the multiplexing gains in C-RAN?
– How to optimize fronthaul network?
![Page 5: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/5.jpg)
Studying multiplexing gains
- towards quantifying benefits of C-RAN
Multiplexing and pooling gain defined
Exploring the tidal effect
Exploring different application mixed and measurement methods
Cloud
RRH 1
RRH 2
RRH n
...
BBU Pool
Mobile Backhaul Network
Aggregated Traffic (h)
24 h
Cloud-RAN
∑ < n
RRH 1
RRH 2
RRH n
...
Mobile Backhaul Network
BBU 2
BBU nBBU 1
RAN with RRHs
∑ = n
![Page 6: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/6.jpg)
Multiplexing gains are available for any shared resources
0
20
40
60
0 6 12 18 24
Load
Time (h)
Data from China Mobile [1]Office Residential
How big are the multiplexing gains in C-RAN?
Cloud
RRH 1
RRH 2
RRH n
...
BBU Pool
Mobile Backhaul Network
Aggregated Traffic (h)
24 h
Cloud-RAN
∑ < n
RRH 1
RRH 2
RRH n
...
Mobile Backhaul Network
BBU 2
BBU nBBU 1
RAN with RRHs
∑ = n
𝑀𝐺𝑝𝑟𝑖𝑛𝑡𝑒𝑟𝑠 =#𝑝𝑒𝑜𝑝𝑙𝑒
#𝑝𝑟𝑖𝑛𝑡𝑒𝑟𝑠
![Page 7: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/7.jpg)
Multiplexing gains are available for any shared resources
Variable bit rate on the fronthaul network
– Bursty traffic
BBUs in the pool – Pooling Gain (PG)
– Processing resources
– Power
1
1
1
1
RRH 1
RRH 2
RRH n
...
Mobile Backhaul Network
BBU 2
BBU nBBU 1
RAN with RRHs
∑ = n
RRH 1
RRH 2
RRH n
...
Mobile Backhaul Network
BBU 2
BBU nBBU 1
RAN with RRHs
∑ = n
RRH 1
RRH 2
RRH n
...
BBU Pool
Mobile Backhaul Network
Aggregated Traffic (h)
24 h
Cloud-RAN
∑ < nCell-processing
User-processing
![Page 8: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/8.jpg)
Multiplexing gains in C-RAN
Sources
– Aggregation of bursty traffic
– Tidal effect
– Different functional splits
• On BBU/fronthaul
Impact energy and cost savings
– On user-data dependent resources
– Processing: ctrl + cell + user
• 3-12% on downlink, 17-33% on uplink of total baseband processing [2]
– Power: ctrl + cell + user
• 2-24 % of total base station consumption [2]
MG
MG
Bit-level
processing
MAC
RLC
PDCP
CP
Resource mapping
IFFT
QAMAntenna mapping
Use
r p
roce
ssin
g
Variable
bit rate
Constant
bit rate
PGPDCP
PGRLC
PGMAC
PGBLP
PGQAM
PGRM_IFFT
PGCP
P
H
Y
Ce
ll
pro
ce
ssin
g
![Page 9: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/9.jpg)
Exploring tidal effect – analytical approach
0.8
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
0 10 20 30 40 50 60 70 80 90 100M
ult
iple
xin
g ga
inOffice cells (%)
Multiplexing gain for different locations, for different mix of office and residential
cellsLondon/LosAngeles
New York
Hong Kong
China Mobile
China Mobile,simulated
0
0.5
1
1.5
2
2.5
0 6 12 18 24
DL
dat
a
Time (h)
Traffic in London, from MIT/Ericsson [3]
City of London Newham
0
0.5
1
1.5
2
2.5
0 6 12 18 24
DL
dat
a
Time (h)
Traffic in New York, from MIT/Ericsson [3]
Lower Manhattan Ridgewood
[4]
![Page 10: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/10.jpg)
Exploring application mix – OPNET simulations [5]
Results for 30% office, 70% residential cells, web and video traffic
Traffic burstiness contributes to multiplexing gain up to 6
Switch
….
3 office cells
….
7 residential
cells
MAC
IP
TCP
Application
MAC
IP
TCP
Application
MAC MAC
Traffic (h)
24 h
Traffic (h)
24 h
Aggregated
Traffic (h)
24 h
BBU Pool
0
1
2
3
4
5
6
7
0 10 20 30 40 50 60 70 80 90 100
Mu
ltip
lexin
g g
ain
Web traffic (%)
100 ms
1 s
![Page 11: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/11.jpg)
Discussion
• PG from analyzing compute resource utilization, in Giga Operations Per Second (GOPS): 1.09 - 1.37, source: Werthmann et al. [6]. Tidal effect not accounted.
• PG based on population in different districts in Tokyo: 4, source: Nambaet al. [7]
• MG from tidal effect: 1-1.3
• MG from traffic burstiness: up to 6
• Fraction of it impacts baseband resources (3-33%) to achieve PG
MG up to 6 achievable on fronthaul, fraction of it achievable on BBU side
New functional split should result in bursty traffic being as “low” as possible (closest to BB-RF – traditional RRH BBU split) to benefit from C-RAN (e.g. spectral efficiency)
![Page 12: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/12.jpg)
Fronthaul transport network
- Towards NGFI
Transport options
Synchronization challenge: application of IEEE 1588
Delay challenge: application of TSN
![Page 13: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/13.jpg)
Possible transport solutions
Technology is ready
(capacity not necessarily)
CPRI / OBSAI
CPRI / OBSAI
WDM
CPRI / OBSAI
CPRI / OBSAI
Microvawe
Compression Compression
Eth/CPRI/ OBSAI
Eth/CPRI/ OBSAI
Ethernet
CPRI/ OBSAI
CPRI/ OBSAI
OTN
BBU Pool
RRH
CPRI / OBSAI Point to point
WDM-PON
![Page 14: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/14.jpg)
Widely deployed (reuse!)
Dedicated links
Shared links
Aggregation
Multiplexing gains on BBU and links
Switching
! Fronthaul cost savings vs problems with delays and synchronization
! Synchronous CPRI vs asynchronous Ethernet
! Data delay: 100-400 us, ≈constant
BBU Pool
GPS/1588 Slave
RRH
RRH
Sync delivered together with data (CPRI)
Ethernet network
Eth switches
BBU Pool
1588 Master
1588 Slave
RRH
CPRI2Ethgateway
RRH
CPRI2Ethgateway
1588 Slave
Shared Ethernet for cost-saving and flexibility [8]
![Page 15: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/15.jpg)
Timing in fronthaul
Timing is really important
Frequency of transmission
Handover, coding, cooperative techniques, positioning
00
0102
00
01
02
Device A
Device B
Current solutions for timing distribution GPS
PHY layer clock – SyncEth
Packet-based timing
IEEE 1588v2 (PTP)
Multiple
Requirements (4G) Frequency error LTE – A TDD/FDD: ±50
ppb
Phase error LTE-A with eICIC/CoMP: ± 1.5 - 5 μs, MIMO: 65 ns, positioning: ± 30 ns
What are the requirements for RoE and 5G?
![Page 16: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/16.jpg)
How to reduce queueing delays?
Preemption (switch upgrade required)
Scheduling and source scheduling
A
BB A
A
BB A
No scheduling – B delay non deterministic
Scheduling – B delayed initially, delay deterministic
BBU pool BBU pool
A
B
A
AB B
B
Input (high priority)
Input (low priority)
Output (no preemption)
Output (with preemption)
Preemption
A
B
AB
AB B
IEEE 802.1, Time Sensitive Networking task force
Frame preemption (802.1Qbu)
Scheduled traffic (802.1Qbv)
Time-Sensitive Networking for Fronthaul (profile definition, 802.1CM)
![Page 17: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/17.jpg)
Exemplary architecture
With control solution – e.g. SDN
With synchronization solution – e.g. IEEE 1588
With delay minimization solution – e.g. TSN
SDN
Ethernet network
Eth switches
BBU Pool
1588 Master
1588 Slave,CPRI2EthGateway,
source scheduling
RoE RRH, 1588 Slave, Source scheduling
Legacy RRH
![Page 18: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/18.jpg)
Conclusions and final remarks
![Page 19: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/19.jpg)
Costs
– 2x2 MIMO, 20 MHz LTE,
15+1 CPRI 2.5 Gbps
– 3 sectors? 7.5 Gbps
– Tens of BS over long
distance? 100+ Gbps
Costs vs savings
C-RAN benefits?Fronthaul
cost?
Savings
– Equipment
– Energy
Benefits from cooperative
techniques
Re-definefronthaul
New function split
Ethernet-based FH
Cost savings
Joint design, joint benefits
![Page 20: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/20.jpg)
Conclusions, proposals to 1914
Optimal functional split is needed to reduce data rate and benefit from
multiplexing gains on fronthaul, while exploiting benefits of C-RAN
One split probably won’t fit all – possible reconfiguration options are
interesting
Multiplexing gains are possible on BBU resources (on 3-33% of
resources), and for variable bit rate split also on fronthaul.
Industry shows a strong interest in packet-based fronthaul.
Ethernet-based fronthaul with traffic scheduling and/or preemption has
the potential to meet mobile networks’ requirements while being cost-
efficient.
Thank you for your attention
![Page 21: Fronthaul architecture towards 5G - IEEEsites.ieee.org/sagroups-1914/files/2016/08/tf1_1608_Checko_Front... · Fronthaul architecture towards 5G ... – User processing in BBU](https://reader034.vdocuments.site/reader034/viewer/2022052303/5b87b4777f8b9a3d028b819e/html5/thumbnails/21.jpg)
References
[1] C-RAN The Road Towards Green RAN. Tech. rep. China Mobile Research Institute, October 2011
[2] C. Desset, et al. “Flexible power modeling of LTE base stations”. In: Wireless Communications and Networking Conference (WCNC), 2012 IEEE, Apr. 2012, pp. 2858–2862.
[3] Many cities. MIT Senseable City Lab. [cited: January 2016]. URL: http://www.manycities.org/
[4] A. Checko, H. Holm, and H. Christiansen. “Optimizing small cell deployment by the use of C-RANs”. In: European Wireless 2014; 20th European Wireless Conference; Proceedings of. 2014 VDE
[5] A. Checko1st, A. P. Avramova1st, H. L. Christiansen, and M. S. Berger. “Evaluating C-RAN fronthaul functional splits in terms of network level energy and cost savings”. in Journal of Communications and Networks, vol. 18, no. 2, pp. 162-172, April 2016.
[6] T.Werthmann, H. Grob-Lipski, and M. Proebster. “Multiplexing gains achieved in pools of baseband computation units in 4G cellular networks”. In: Personal Indoor and Mobile Radio Communications (PIMRC), 2013 IEEE 24th International Symposium on. Sept. 2013, pp. 3328–3333.
[7] S. Namba, et al. “Colony-RAN architecture for future cellular network”. In: Future Network Mobile Summit (FutureNetw), 2012. July 2012, pp. 1–8
[8] A. Checko, A. Juul, H. Christiansen, M. S. Berger, "Synchronization Challenges in Packet-based Cloud-RAN Fronthaul for Mobile Networks“, IEEE ICC 2015
A. Checko, H. Christiansen, Y. Yan, L. Scolari, G. Kardaras, M.S. Berger and L. Dittmann „Cloud RAN for Mobile Networks - a Technology Overview”, IEEE Communications Surveys & Tutorials