Before we jump in to SIPTO it would be suggested that you visit Traffic Break-Out in LTE for better understanding of concept, let see how service (i.e.
Internet) is given in LTE. After subscriber authorization and authentication
S-GW and P-GW (SE-GW) are used to provide service to subscriber. Now it is just
common sense that it would be better for service, for resource (RAN) management
point of view and less in cost as well if service is given by SE-GW that is
nearest (Geo-graphically) to UE (subscriber). Now SIPTO is a way to route
traffic from closest point to UE. Only specific traffic is routed/served from
closest point that is why it called as SELECT IP Traffic Offload (SIPTO).
General architecture of SIPTO |
Let’s take an example, there is a subscriber of an operator in India uses video.internet.com APN to view videos by using SE-GW of India, now this operator has roaming agreement with various countries. Subscriber travelled to Japan then to get internet access it should use SE-GW of Japan rather than India if it is allowed for SIPTO. In Lay man term, an e-node B (Tower) in Japan should access Service Providing Nodes in Japan once authentication and authorization is done otherwise it shall automatically increase round trip time if it uses Indian Serving nodes. More Simpler it would be better to get electricity from nearest Poll or getting water from nearest water tank rather than own state poll or tank while you are on travel.
Formally, SIPTO function enable an operator to offload certain type of traffic to Service providing node that is closed to UE in ACCESS network. In LTE, SIPTO over RAN is possible be selecting S-GW and P-GW that is geographically closest to UE.
Complete offload is generally done with the help of LIPA
that we shall discuss later.
Permissions for SIPTO to an APN are stored in HSS. SIPTO can
be applied in both Roaming and non-Roaming cases. If HSS indicates that VPLMN
address allocation is not allowed then SIPTO is not provided.
As LTE and technology
is growing there for various concept are invented such as H-Enode B and L-GW
that can be used for SIPTO. As we Know
there are two major entities in SIPTO (1) Gateways (2) E-NodeB, thus optimization
can be done at two places either Gateways or E-NodeB so concept of L-GW comes
because of optimization/Enhancements at Gateways and H-Enode B comes because of
optimization/Enhancements at E-NodeB.
SIPTO at the Local Network with L-GW function,Here traffic is offloaded with architecture-2 that is self explanatory from below Image.
SIPTO at the Local Network with L-GW function,Here traffic is offloaded with architecture-2 that is self explanatory from below Image.
Your Comments /Suggestions and Questions are always welcome, shall clarify with best of knowledge. So feel free to put Questions.
How can The SIPTO work how can it decided and how it is work ?
ReplyDeleteHi Munther Numan.
DeleteWhen, How and Where to apply SIPTO is stored by operator in Subscriber Profile for specific APN. This Subscriber Profile is downloaded to MME (Controlling Node) then MME takes decision about SIPTO based on SIPTO details received from HSS and MME(SGW-PGW) local configuration.
Thanks for your query.
Happy to help you again.
Team-Diameter
Hi,
ReplyDeleteI have LBO functionality to enable for 4G network to one of our client, does making LIPA-Permission to LIPA-Only will work or should anything else to be done.
Hi Sunil Reddy,
DeleteLIPA is basically not about functionality whether or not supported by LIPA Moreover whole idea of LIPA is to break traffic of a given functionality and it depends on your network deployment.
Following links might help you.
1) http://diameter-protocol.blogspot.in/2015/07/traffic-break-out-in-lte.html
2) http://diameter-protocol.blogspot.in/2015/07/lipa-local-ip-access-function.html
Thanks for your query.
Happy to help you again.
Team-Diameter
Hi,
ReplyDeleteCan you please explain the detailed signal flow for LBO/SIPTO with proper inclusion of interface details.
How charging is done in such scenarios? whether it occur at h-pcrf or v-pcrf?