Format of gsm/gprs target cell id in GTPcV2 in case of srvcc handover

Question

Hello,

I am trying to test our SRVCC implementation.

In the "PS to CS Request" message from MME to EMSC, there is no specific format in which target cell id is sent.

Do you have any inputs how i can encode and send and decode the same. ?

29.280 sec 6.9 says refer to MAP spec 29.002, But MAP spec doesnot specify anything.

Is there some format like "PLMN id", LAC etc passed in ..If so why this is not specified in specs !

thanks a lot
pdk

Comments

From the current gsm/gprs handover logs ( Bit painful, but had  setup our setup for current intermsc handover): I could see the following format in the wireshark

Cell Identifier (Target) - LAC (0x000e)/CI (1414)
    Element ID: 0x05
    Length: 8
    0000 .... = Spare bit(s): 0x00
    .... 0000 = Cell identification discriminator: The whole Cell Global Identification, CGI, is used to identify the cells. (0)
    Mobile Country Code (MCC): Unknown (1)
    Mobile Network Code (MNC): Unknown (14)
    Cell LAC: 0x000e (14)
    Cell CI: 0x0586 (1414)

Hex outpout:
05 08 00 00 f1 41 00 0e 05 86

I can related the last three bytes to the LAC and CI (0e 05 08)..Not able to map the MCC and MNC to the hex dump. Could some experts please help me with this.This forum has been really helpful in that respect
PS: When i highlight the MCC (00 f1 ) is selected. For MNC (f1 41) is selected. on wireshark

thanks a lot
~pdk

Answer 1

What is the query, I could not get the question. Please explain a bit more -
1. Which is the protocol (I believe its GTP V1).
2. It seems u answered yourself other then MCC and MNC part see the following which should be useful (not sure at ur protocol if it is big or little endian so providing both encoding.

/* MCC/MNC dissection - little endian MNC encoding */
 * +---+---+---+---+---+---+---+---+
 * |  MCC digit 2  |  MCC digit 1  |  octet x
 * +---------------+---------------+
 * |  MNC digit 3  |  MCC digit 3  |  octet x+1
 * +---------------+---------------+
 * |  MNC digit 2  |  MNC digit 1  |  octet x+2
 * +---------------+---------------+
 */

/* MCC/MNC dissection - big endian MNC encoding */
 * +---+---+---+---+---+---+---+---+
 * |  MCC digit 2  |  MCC digit 1  |  octet x
 * +---------------+---------------+
 * |  MNC digit 1  |  MCC digit 3  |  octet x+1
 * +---------------+---------------+
 * |  MNC digit 3  |  MNC digit 2  |  octet x+2
 * +---------------+---------------+
 */

Comments

Thanks a lot for the quick response and right answer. Sorry about my Question, struggling to setup thinkgs, in my company with no documents and startup is bit nightmare. that too coming from UE, sometimes we keep somethings we just used to some format and diffuclt when looking from other side :)

Btw, it looks like the PLMN is the only field which looked bit confusing, as there were lot of hardcodings within our code (which made me bt confused about the PLMN format ) Sorry about this basic question...

---

Sorry forgot to answer your query, the protocol is GTPcV1.

Lets say it is big endian: It should have been encoded as "00 01 41". But not sure why it is encoded as "00 f1 41" in the Hex dump though !

---

No clue from where F is coming, unfortunately I dont have access to wireshark so difficult to guess here.

---

Sorry for the delay. I am not sure there is also a hardcoding in the code to check this value. May be specified in some specs some where but not able to find..But anyway now it is easier to decode ( big endian/little endian encoding with your inputs)..thanks a lot

---

Your most welcome :)