RAN Sharing - HQoS Guidelines - Ver. 3.1
RAN Sharing - HQoS Guidelines - Ver. 3.1
RAN Sharing - HQoS Guidelines - Ver. 3.1
HQoS guidelines
Issue 3.1
Date 2016-08-26
and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and
the customer. All or part of the products, services and features described in this document may not be
within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements,
information, and recommendations in this document are provided "AS IS" without warranties, guarantees or
representations of any kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.
Website: http://www.huawei.com
Email: support@huawei.com
Author
Prepared by Mauro Grandi (employee ID 00737927) Date 2016-05-13
Reviewed by Date
Granted by Date
Change History
Date Revised Version Change Description Author
Mauro Grandi (employee
2016-05-13 01 Completed the first draft
ID 00737927)
2016-06-23 02 Added some figures and parameters Mauro Grandi (employee
ID 00737927)
2016-08-17 03 Modified some figures and paragraphs Mauro Grandi (employee
ID 00737927)
2016-08-26 03.1 Changed WRED thresholds and added SVR Mauro Grandi (employee
ID 00737927)
Contents
3 Configuration ............................................................................................................................... 15
3.1 MW link configuration ................................................................................................................................... 15
3.2 Service configuration ..................................................................................................................................... 16
3.3 HQoS tests...................................................................................................................................................... 18
3.4 HQoS configuration ....................................................................................................................................... 21
3.4.1 U2000 and RTN950 SVR ..................................................................................................................... 21
3.4.2 DS profile .............................................................................................................................................. 22
3.4.3 WRR Scheduling Profile ....................................................................................................................... 27
3.4.4 Service WRED Profile .......................................................................................................................... 29
3.4.5 V-UNI Egress profile ............................................................................................................................ 32
3.4.6 HQoS port and service setting............................................................................................................... 35
3.4.7 HQoS Group ......................................................................................................................................... 37
3.4.8 Egress DSCP Mapping .......................................................................................................................... 39
4 Appendix ...................................................................................................................................... 40
4.1 Specific request of HQoS measurements from a customer ............................................................................ 40
4.2 Testbed setup .................................................................................................................................................. 40
4.3 Testing ............................................................................................................................................................ 41
4.3.1 Testing configuration ............................................................................................................................ 41
4.3.2 Baseline MW link ................................................................................................................................. 42
4.3.3 Oversubscribe BE on all VLANS ......................................................................................................... 43
4.3.4 Oversubscribe BE on all VLANS and Reduce CIR/PIR VLAN 30 ...................................................... 43
4.3.5 Reduce BE Traffic in VLAN 30 ............................................................................................................ 44
1 Introduction
1.1 Objective
This document is based on Customer T (called in the following CT) transport network for mobile
technologies: 2G, 3G and 4G. They have RAN sharing with Customer V (called in the following CV). This
means that in the same mobile site there are 2 customers using the same transport network. In this case in
order to guarantee each operator a certain amount of data bandwidth: HQoS must be implemented.
Referring to network depicted in Figure 1, in Figure 2 is reported traditional DiffServ QoS with the
limitations above mentioned that are solved with the implementation of HQoS.
In Figure 3 is depicted how HQoS technology is implemented in Huawei MW. It schedules Ethernet services
that are transmitted through five levels, finely controlling the service quality of different subscriber data and
service data.
Level 5: subdivides the services of a subscriber into voice, video, Internet traffic, and others.
Controls the bandwidth of each service type of the subscriber.
Level 4: identifies each subscriber and controls the bandwidth of each subscriber.
Level 3: identifies each subscriber group and controls the bandwidth of each subscriber group. (For
example, the subscribers using different types of base stations can form different subscriber groups.)
Level 2: limits the rate of each queue at an egress port.
Level 1: limits the rate of each egress port.
The following chapters report some HQoS configurations based on customers requirements.
Table 1 reports for each service (4G, 3G, 2G) the correspondent DSCP and the PHB expected by the
transport network. OAM is treated as 2G.
MW network (3rd column) must classify packets based on DSCP (2nd column). Further DSCP must not be
modified by RTN network. In the same way RTN must leave VLAN-priority bits (COS) as they are at the
ingress of the network.
RTN
Row DSCP Queue WRR IPVPN
4G Service Classifier Coloring
# Marking Type % CoS
and Queue
1 Not used for 4G services (56) CS7 SP NA NA BE
2 Not used for 4G services NC (48) CS6 SP NA NA BE
QCI-1 (Conversational
3 EF (46) EF SP NA NA EF
voice)
4 M-plane AF11 (10) AF1 WRR 5% NA AF11
QCI-2 (Conversational
5 AF31 (26) AF2 Green AF11
video)
QCI-3 (Real-time
6 AF42 (36) AF2 Yellow AF12
gaming) WRR 30%
QCI-4
7 (Non-conversational AF32 (28) AF2 Red AF12
video)
QCI-5 (IMS signaling)
8 AF41 (34) AF3 WRR 10% NA AF11
and C-plane
9 QCI-6 ('Gold') AF21 (18) AF4 WRR 15% NA AF11
10 QCI-7 ('Silver') AF23 (22) BE Green AF11
QCI-8 ('Bronze') and
11 DF & CS0 (0) BE WRR 30% Yellow DE
ICMP
12 QCI-9 ('Oak') CS1 (8) BE Red DE
RTN
DSCP Queue WRR IPVPN
3G Service Classifier Coloring
Marking Type % CoS
and Queue
13 Not used for 3G services (56) CS7 SP NA NA BE
14 Not used for 3G services NC (48) CS6 SP NA NA BE
Network
15 EF (46) EF SP NA NA EF
Control/S-Plane/RT R99
C-plane/HSPA FP
16 AF41 (34) AF3 WRR 45% NA AF11
Control
NRT R99/Streaming
17 AF31 (26) AF2 WRR 45% NA AF11
HSPA/PS data
18 M-plane AF11 (10) AF1 WRR 3% NA AF11
19 NRT HSPA and ICMP DF & CS0 (0) BE WRR 7% NA DE
RTN
DSCP Queue WRR IPVPN
2G Service Classifier Coloring
Marking Type % CoS
and Queue
20 Not used for 3G services (56) CS7 SP NA NA BE
21 Not used for 2G services NC (48) CS6 SP NA NA BE
Network
22 Control/S-Plane/CS-Voic EF (46) EF SP NA NA EF
e
23 C-plane AF41 (34) AF3 WRR 45% NA AF11
24 PS Data AF31 (26) AF2 WRR 45% NA AF11
25 M-plane AF11 (10) AF1 WRR 3% NA AF11
26 ICMP DF & CS0 (0) BE WRR 7% NA DE
Table 1 DSCP mapping for transport network
Queue AF2 and BE for 4G services (rows 5-7, 10-12) are supposed to work with colored packets (color
aware with DSCP), in order to use this information WRED must be implemented on these queues. RTN with
simple classification can work with colored packets in queues AF4-AF1, so BE is not working on this mode
and in order to accomplish what requested by CT we have suggested to swap AF1 with BE queues with
correspondent weights, both queues are in WRR the behavior will remain the same.
The 4 different types of services 4G, 3G, 2G and OAM have the mapping between common DSCP and
queue that perfectly overlaps, so it will be possible create only one simple classification map in ingress. In
Table 2 are reported the DSCP values against the PHB in the RTN network. Each packet entering the RTN
network is classified based on DSCP and it is scheduled according with values reported in the table. As
already mentioned DSCP is color aware; it means that on the same AF queue, there are packets with high
(RED), medium (YELLOW) and low (GREEN) discarding probability in case of congestion. In the table are
reported only DSCP and queues used by CT.
DSCP PHB
46 EF
18 AF41
34 AF31
26 AF21
36 AF22
28 AF23
22 AF11
0 AF12
8 AF13
10 BE
Table 2. DSCP mapping in RTN
For the 4 services the scheduling type queues are the same: CS7, CS6 and EF in SP, the remaining queues in
WRR. The weights of WRR between 4G and 2G/3G/OAM services are different. CS7 and CS6 at the
moment are not used by CT MW network.
Both sites have the following data: OAM, 2G, 3G and 4G for each customer. In the core network there are 4 VPN
OAM, 2G/3G, 4G CT, 4G CV. All services above mentioned are reported in Table 4.
In the next chapter is reported the MW configuration for the network depicted in Figure 6, with HQoS settings
and with some test cases in order to verify HQoS behavior.
3 Configuration
Link 2 Link 1
In Figure 16, Figure 17 and Figure 18 are reported 3 test cases in order to verify that HQoS is working as
expected.
Figure 16 Test 1
Figure 17 Test 2
Figure 18 Test 3
3.4.2 DS profile
In the following are listed the steps in order to create a DS profile template. In Figure 22 and Figure 23 the values
in IP DSCP column are configured according to Table 2.
It is also possible to directly download and configure the DS profile on the ports. In Figure 26, Figure 27, Figure
28, Figure 29 are reported all the steps required to do it.
Figure 28 DS profile download to port: 3rd step, application object selection on the port
In the same way 2G-3G V-UNI Egress profile has been created.
In Figure 44 how to download the V-UNI Egress profile template on the NEs.
On each NE there is V-UNI Egress profile (see Figure 45 and Figure 46).
In Figure 49 there is an example of UNI QoS configurations over one e-line service done directly from
U2000. In the example there is the HQoS configuration for 4G CT Site 2. In Figure 50 is reported the same
configuration above mentioned on a single NE.
4 Appendix
4/1 VLAN 3
VLAN 20
4/2 4/4 VLAN 30
4/3
Node 2 Node 1
VLAN 3 NE3
VLAN 30
IDU
IDU 22 IDU
IDU 11
NE1
Throughput 100Mbps
4.3 Testing
4.3.1 Testing configuration
The IDU will be set with 100Mbps on the MW link and CIR/EIR/PIR will be set as follows on the MW link:
The test will be based on the following traffic analysis from MBNL:
IPAbis 3G 4G
EF % 39.00% EF % 9.00% EF % 2.50%
AF4 % 1.00% AF4 % 1.00% AF4 % 0.00%
AF3% 0.00% AF3% 0.00% AF3% 0.00%
AF2% 1.00% AF2% 0.00% AF2% 0.00%
AF1% 57.00% AF1% 0.00% AF1% 2.00%
BE % 2.00% BE % 90.00% BE % 95.50%
This is to ensure that the additional BE load is able to burst into unused BE capacity in other VLANs but that
native prioritisation is not affected.
This test can be repeated for the other PHBs if required and different burst rates measured.