25
Jan

my first eGTP test – take 2 – CreateSessionResponse

   Posted by: cristina_crow   in technical

As I was telling you about in a previous post – my first eGTP test, the reply (first reply) to a CreateSessionRequest message is a CreateSessionResponse message, described below. This message contains:

- GTPversion 2, Message Type information, in this case, this is a Response, the length of the message, the sequence number (1) and the TEID (tunnel Endpoint Identifier) – which is copied from the CreateSessionRequest message

- the Cause field indicates this is a Response for an Accepted Request – in case there would be any error, the Cause Source field would indicate the cause of the error

- PDN Address Allocation (PAA) – field which is completed at this moment  (in the CreateSessionResponse) by the SGW with the IP address of the PDN – should you remember, in the CreateSessionRequest message, this field indicated the type of address (IPv4) and value 0.0.0.0; as per 3GPP TS 29.274 – this value is a fixed IPv4/IPv6 address as indicated by the HSS registers, or it leaves the value to 0.0.0.0 indicating that the PDN GW address is assigned dynamically

- F-TEID (Fully Qualified Tunnel Endpoint Identifier) – as mentioned also in the previous post, there are 2 F-TEIDs: one for the S11 interface, and another one for the S5/S8 interface, both source IP addresses of GTP-C:

— one for the S11 interface (the one between MME and SGW) – the SGW end – the IP of the SGW from the S11 interface

— one for the S5/S8 interface (the one between SGW and PGW) – the IP address of the APN server

- APN Restriction header – as per 3GPP TS 29.274, it “denotes the restriction on the combination of types of APN for the APN associated with this EPS bearer Context.” – haven’t  used it yet, so I cannot say too much about it

- Bearer Context – information I have neglected to describe in sufficient detail in the CreateSessionRequest description. Here, in the CreateSessionResponse message, the Bearer Context header has 6 sub-headers:

— EPS Bearer ID

— Charging ID

— F-TEIDs : here both of the identifiers contain the IP address of the SGW’s S11 interface – the source GTP-U interface

— Cause : here is Request Accepted, no Cause Source

— Bearer QoS, which contains the QCI label and some other QoS identifiers that shall be described – hopefully I’ll be able to see them at work till then

Tags: , , , , , , , ,

This entry was posted on Monday, January 25th, 2010 at 7:48 pm and is filed under technical. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

4 comments so far

Yuanliang Zhang
 1 

Hi, Critina, can I have two questions?
(1): for the PAA in this post, my understanding is about the UE’s IP address with this PDN, not the PDN GW address.

(2): For the F-TEID of the GTP-U in the Bear Context, why the PGW (S5/S8)GTP-U and the SGW (S1-U) GTP-U have the same IP address, while the IP address in the correspoding GTP-C is different?

May 24th, 2010 at 10:17 am
 2 

@Yuanliang: Hello, thanks for reading this so thoroughly :)

(1): indeed, this is the PDN address, not to PDN GW – I am changing it right now

(2): that was a bug in my script, so it always puts that IP there; the IP address of GTP-U of the S5/S8 PGW IP is different than the SGW GTP-U for S1-U IP address, just as the GTP-C IP addresses are also different; in the real world, these IP addresses are actually different

Thank you again for your interest :)

May 24th, 2010 at 11:15 am
Yuanliang Zhang
 3 

You are welcome, cristina. I’m a new comer for the LTE call processing and your posts and logs are quite helpful for me, thanks!

One more suggestion, for the Create Bearer request message in the logs, I would like to suggest to change the EBI IE to be the name “LBI”, since although the IE type is EBI, it is actually the LBI of the dedicated EPS bearer to be created, please correct me if I’m wrong.

May 25th, 2010 at 4:27 am
 4 

@Yuanliang: good point, though the spec states that type of IE is EBI. I’ll further look into it :) Thanks

May 25th, 2010 at 9:30 am

Leave a reply

Name
Mail (will not be published)
URI
Comment