Sie sind auf Seite 1von 8

Data Offload

Use Case

AWTG Limited
7th Floor, Westgate House, Westgate Road,
London, W5 1YY.
Tel. 0208 799 0368, Fax. 0208 728 9610

www.awtg.co.uk

Copyright AWTG Ltd Page 0 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Contents

Executive Summary ..................................................................................................................................... 2

Introduction ................................................................................................................................................. 2

Use Case ...................................................................................................................................................... 3


Objective & Scope of Work .....................................................................................................................3
Demonstration and Results .....................................................................................................................4
Step 1: 3G connectivity- EAP/AKA authentication: .............................................................................4
Step 2: WiFi connectivity- EAP/AKA and EAP-SIM authentication......................................................5
Step 3: 3G to WiFi handover, EAP-AKA authentication ......................................................................6
Step 4: WiFi to 3G handover, EAP-AKA authentication ......................................................................6

Conclusion ................................................................................................................................................... 7

Future work ................................................................................................................................................. 7

References ................................................................................................................................................... 7

Copyright AWTG Ltd Page 1 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Data Offload
Use Case
Executive Summary
WiFi enabled Smartphones, Tablets and Mobile Broadband
enabled Laptops have led to an increased appetite for
mobile data bandwidth hungry applications. Many of today’s
devices are dual mode, for example a 3G/LTE handset often
has a WiFi chipset onboard. With the amount of cellular
traffic being consumed by these Smart devices set to
increase further, mobile operators are increasingly turning
to WiFi as a complementary technology to work alongside
traditional mobile access technologies.

Figure 2. Global Handset Data Traffic

One of the available options for operators to answer the


data hungry users is to offload 3G data to WiFi networks.
There are different approaches for mobile operators to
offload data onto WiFi and these approaches differ based on
the level of integration of WiFi into the mobile core network,
each approach gives the operator different levels of control
over the subscriber.
Figure 1. Wireless data traffic generated of out of home and place of work
by network type, developed markets
In the integrated model proposed by 3GPP, WiFi is inter-
(Source :Analysis Manson 2013) networked with the core mobile network through a bridge
for controlling and managing the mobile data offloading. In
With the rise in number of Carrier-class WiFi networks and this approach, operators can seamlessly move the mobile
the uptake of WiFi by mobile carriers as shown in Figure 1, it device from the cellular network to the WiFi network and
becomes extremely important to ensure that end user seamlessly transfer the data through WiFi. However, this
experience is not degraded when a mobile handset switches approach adds more complexity to the 3G/4G core network
between access technologies. architecture in terms of mobile and WiFi network integration
challenges. The offloading solution may also require access
AWTG has carried out a 3G data offloading demonstration layer applications to be installed on handsets to make this
under lab conditions to study the results and challenges of technology available to more customers.
seamless offloading, the demonstration was defined based
on real life scenarios and used industrial standards as a There are already handsets on the market which can use 3G
basis. The network design and topology was based on Cisco offloading solutions without additional applications on the
and Ericsson solutions. client side, this can help operators connect these clients to
their WiFi network seamlessly and offload data traffic to the
Introduction WiFi. This model is called clientless offload solution.
During recent years, WiFi enabled Smart phones and Mobile
Clientless offload solutions provide operators with relatively
Broadband enabled laptops have increased the data usage
quick solutions with low cost to their capacity issue by
demand from bandwidth-hungry, real-time streaming video
having solutions without heavy investment in complex
and data applications. This unexpected demand has
integration level issues. Another advantage of this approach
increased the operator’s network usage exponentially as
is that it does not require any special application running on
shown in figure 2 whilst the operator’s network was not fully
the mobile devices to enable operators to define and enable
ready to support the amount of requested data.
QoS, security policies and accounting services.

Copyright AWTG Ltd Page 2 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Use Case There are various ways to authenticate and manage
Objective & Scope of Work subscriber access to WiFi networks. Transparency in the
Data traffic has increased significantly on today’s 3G authentication method could affect the subscriber
network, this situation is not only creating the problem but confidence in the network, the more transparent the
also creating an opportunity for carriers. Providers are authentication method, the larger the probability that the
focusing on WiFi networks and using 3G data offloading subscriber will connect to the network. There are two
solutions. 3G data offloading has numerous challenges. different authentication methods, portal based
authentication and EAP authentication. Portal based
The demonstration has been defined to simulate a 3G data authentication is a complicated method which requires HTTP
offloading solution based on the combination of Cisco (WiFi and layer-3 connectivity before the authentication process
AP and ASR5K used as Gateway) and Ericsson (SGSN, HLR) starts. The alternative way is EAP authentication which
hardware as shown in figure 3. The detailed scope of work provides easy and transparent access. In this demonstration
for this document includes user authentication, testing both we simulate EAP authentication. EAP-AKA (EAP
trusted and untrusted WiFi, clientless and client based and Authentication and Key Agreement) and EAP-SIM are two
also seamless mobility for real time applications. different standards used for the encapsulation, we test both
EAP-AKA and EAP-SIM authentication with a WiFi Network.

Another crucial challenge in 3G data offloading is the


behaviour of near real time applications such as YouTube,
FTP and real time applications such as VoIP and VPN during
handover between 3G and WiFi, we go further and test
seamless mobility for these near real time applications and
Real Time applications for mobility between 3G and WiFi.

Based on the defined scope of work we have set the


objectives of the demonstration and track the results and
also ensure that the chosen architecture and message flows
to authenticate (using EAP-SIM and EAP-AKA) between the
Mobile Packet Core and the WiFi Core are inline with the
scope.
Figure 3. Demonstration Topology
We analyze a clientless solution to monitor its limitations in
seamless mobility and then prove functions of a client based
User authentication is an important aspect in 3G data
solution for seamless mobility.
offloading, in the demonstration we capture and verify the
message flows for authentication and mobility between 3G
Various test scenarios are run to show that there is no
and WiFi networks.
impairment in user experience when the user moves
between WiFi and the 3G Packet Core. The network used for
Untrusted WiFi networks were introduced in the early phase
the 3G offload solution simulation has been illustrated as
of WiFi specification in 3GPP release 6, access to these
below:
networks includes any type of WiFi access that is not
controlled by the operator. Another scenario for untrusted
access could be that the network might be under control of
the operator but does not present a satisfactory level of
security. Examples of Untrusted Access are Home WLAN and
publically open hotspots.

Trusted IP Access was introduced only with the LTE model in


3GPP release 8. Trusted WiFi Access refers to provider or
mobile operator owned or controlled WiFi access providing
over the air encryption and secure authentication methods.
Figure 4. Network layout used for the 3G offload simulation

Mobile operators are not only focusing on deploying their


own WiFi networks but also willing to expand their WiFi
access via untrusted networks. In this demonstration we
have set up the solution for both trusted and untrusted WiFi
access and tested both clientless and client based solutions.

Copyright AWTG Ltd Page 3 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Lab Setup Demonstration and Results
This section describes the devices, applications and The demonstration has been divided into four main
protocols we have used to simulate 3G data offloading. steps as below:
 Cisco Access Point
 Wireless LAN Controller Step 1: 3G connectivity- EAP/AKA authentication:
 Cisco ASR5K used as Gateway After setting up our lab we verified the 3G connectivity. We
 Ericsson SGSN have the 3G network setup in the lab and verified
 Ericsson HLR connectivity by switching on the UE to get the coverage and
 GTP and Mobile IP protocols are used to maintain the network related configurations from the core network to
IP address for UE when moving between technologies establish the connection with the GGSN.
 Both Trusted (VF-X1-TEST) and Untrusted WiFi (VF-X1-
TTG) SSIDs configured in the lab WiFi network Pre-Conditions
 Four Samsung phones (UE) are used in this testing with Before we set up 3G connectivity we have to verify some
SIM IDs of 311480000000226, 259, 260 and 267. These settings which should be present in the UE. The APN for
are configured on both 3G and WiFi core so that vzwinternet-wifi was configured on the Samsung Galaxy 3
interworking can be tested. phone with a 226 SIM. The SIM was also configured for EAP-
 Third-Party client is used to test client based solution AKA authentication. The 3G connectivity test was an
for seamless mobility. important prerequisite for the 3G to WiFi offloading test.

Test Execution Methodology Execution Steps


To simulate 3G data offloading and the scenarios that we Ensuring that there is no data connection present at the UE,
have discussed in this document, we have executed we monitored the activity on the GGSN to confirm that we
following test plan. are not receiving any PDP context requests. Then we
switched on the 3G data connection and checked to see if
To simulate the scenario we first turned off both WiFi and the phone connects to the 3G UMTS network
3G on the UE and then whilst the WiFi was off we turned on
3G connectivity. Once the UE successfully connected to 3G Results
we verified the IP address allocated to the UE from the
The 3G connectivity flow was as expected, Ericsson’s SGSN
GGSN/HA pool. After verification that the UE is connected to
received the data connectivity request and the SGSN
3G we started an application under test for mobility from 3G
communicated with the HLR to authenticate the subscribers
to WiFi. While the application was running we turned on
from a 3G perspective. Once authentication was completed,
WiFi and verified that UE was connected to the WiFi network
the SGSN initiated a PDP context request to the ASR5K GGSN
and received the IP allocated to the UE from the client pool
and in turn the GGSN accepted the request and sent PDP
and also that the application was not disrupted. We turned
context response. The UE was assigned an address from the
off the WiFi to move back to 3G and again confirmed the IP
GGSN pool between (192.168.20.192 - 222) stating IUCNAI
address allocation and checked the application status and
for UMTS. We verified the address allocation on the ASR5k
confirmed that the application was not disrupted at this
by issuing a "show subs all” command. This info was stored
stage.
in a log file for all the test cases.

To summarise, the following steps were completed during


the procedure:

 Turn off WiFi, attach the UE to the 3G network and


start an application under test for mobility
 Check the IP address allocated to the UE from the
GGSN/HA pool Figure 5. UE assigned IP Address
 Turn on the WiFi and see that another IP address is
allocated to the UE from the client pool Call flows were verified using Wireshark traces stored during
 Check that the application is not disrupted testing, as seen in figure 6 the successful RANAP attach and
 Move back to 3G and check that the earlier IP PDP context activation message goes to Cisco GGSN using
addresses are maintained and the application GTP control messages.
session is not disrupted

Copyright AWTG Ltd Page 4 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Figure 6. Wireshark Trace for PDP context request and response

Step 2: WiFi connectivity- EAP/AKA and EAP-SIM Results:


authentication The connection was established after EAP authentication
Following on from the testing of 3G connectivity, we test and IP addresses assigned from the client pool to the mobile
WiFi connectivity. To simulate WiFi connectivity on the UE wireless interface as can be seen in figure below.
the WiFi network was switched on and the connection from
UE to GGSN was traced and monitored.

Figure 7. Wireshark Trace showing connection established and IP addresses assigned from the pool

Copyright AWTG Ltd Page 5 of 7 August 2013


Document V 1.11 Use Case: Data Offload
Step 3: 3G to WiFi handover, EAP-AKA authentication
During this step the EAP-AKA authentication flow for 3G to
WiFi handover was captured. To observe the impact on data
services on the UE and also observe the packet flow of 3G to
WiFi offloading we took following steps:

 Connect to 3G network
 Monitor EAP-AKA authentication
 Start ICMP traffic
 Switch on the WiFi
 Monitor WiFi Authentication
 Monitor effect on ICMP traffic

Results:
Figure 8. Only four packets lost during 3G to WiFi handover
As expected, when we switched on the WiFi after initiating
ICMP traffic while the 3G connection was established, the UE
connects to WiFi by going through EAP-AKA authentication Step 4: WiFi to 3G handover, EAP-AKA authentication
in the WiFi core. A GTP tunnel is established between the In step 3, we monitored the packet flow and service impact
eWAG and the Cisco GGSN. The UE is allocated another IP of 3G to WiFi handover. In this step we observe the
address from the client pool, we verify the address allocated signalling flow of WiFi to 3G handover.
to the handset by issuing the "show subs all” command on To test WiFi to 3G handover we take the following steps.
the ASR5K, we note that the IP addresses are maintained  Ensure handset is connected to WiFi and 3G is also
from the previous addresses allocated when connected to enabled.
3G. This info is stored in the Log file for all the test cases.  After verification of connectivity, verify that the
During the 3G handover to WiFi we found some ICMP traffic handset is receiving an IP address from WiFi client IP
packets were lost. We use Wireshark traces to verify the call
Address pool.
flows for GTP, IP and Radius messages stored in the file
“WiFi connectivity 226”  Initiate the ICMP traffic.
 Switch off WiFi and observed the packet flow and
impact on ICMP traffic.

Figure 9. WiFi to 3G handover flow

Copyright AWTG Ltd Page 6 of 7 August 2013


Document V 1.13 Use Case: Data Offload
Results:
Results of WiFi to 3G handover were as expected, the UE maintained from the previous address allocated when
connects to WiFi by going through EAP AKA authentication. connected to 3G. This info is stored in the Log file for all the
A GTP tunnel is set between the SGSN and the CISCO GGSN test cases also verify the call flows in the Wireshark traces
and the UE gets another IP address from GGSN. for GTP, IP and Radius messages stored in the file. During the
handover we found a small number of packets were lost.
We verified the address allocated to the UE by checking on
Only two packets were dropped during handover from WiFi
GGSN and confirmed that the IP addresses are to 3G.

Figure 10. Wireshark Trace showing IP addresses are maintained


Conclusion
The main aim of this demonstration was to test a 3G  The call flows on the 3G and WiFi technologies allow us
offloading solution based on Cisco’s ASR5K solution in to verify our network architecture and the expected
AWTG’s lab and to study the challenges and observe the messaging for RANAP, GTP, RADIUS and Mobile IP
outcomes. protocols.
AWTG’s lab setup enabled us to replicate a 3G offloading
scenario based on a multivendor network consisting of a Future work
combination of Cisco and Ericsson hardware. During these AWTG constantly investigate future technologies to help our
demonstrations we verified vendor claims and also tested clients improve their offloading solutions and such we
connectivity and seamless handover between two continue to develop more use cases in our lab covering the
technologies. following bullets:
We demonstrate that WiFi is inter-networked with the core  Test untrusted WiFi offloading scenario which includes
mobile network through a bridge for controlling and the TTG function and a client on the UE.
managing the mobile data offloading. We showed that
 Testing offloading while reducing the power of 3G or
operators can seamlessly move the mobile device from the
cellular network to the WiFi network and seamlessly transfer WiFi AP to force handover between 3G and WiFi rather
data through WiFi. than switching on/off
 Change the timer value for IP allocation on ASR5000 to
The summary to take from these demonstrations and the
related logs are: study the impact on end user experience
 Successful end-to-end 3G network connectivity using  Testing trusted and untrusted scenarios with Ericsson
Ericsson SGSN and HLR for authentication with UE GGSN to simulate a multivendor offload environment.
address allocation from Cisco’s GGSN.  Implement 4G to WiFi offload and test seamless
 Successful WiFi network connectivity and IP address mobility
allocation using GTP and Mobile IP with Cisco GGSN/HA
/FA whilst retaining IP address from 3G for mobility. References
 When users move between 3G and WiFi, the IP http://www.prnewswire.com/news-releases/strategy-analytics-handset-
data-traffic-to-grow-over-300-by-2017-to-21-exabytes-214113401.html
addresses allocated to the handset by the network are
maintained as users move between the technologies

Copyright AWTG Ltd Page 7 of 7 August 2013


Document V 1.13 Use Case: Data Offload

Das könnte Ihnen auch gefallen