Friday, 25 February 2011

Attach Sequence for LTE Radio

I have in past posted a complete Attach Sequence on the 3G4G website for LTE Radio Signaling but included the signalling on a few nodes. Recently I came across a signalling example in NTT Docomo technical journal which was less detailed but at a higher level and detailed signalling on these other nodes. It may be worthwhile brushing up the LTE Architecture diagram before diving into this.

With EPC, when a terminal connects to the LTE radio access system it is automatically connected to the PDN and this connected state is maintained continuously. In other words, as the terminal is registered on the network (attached) through the LTE radio access system, a communications path to the PDN (IP connectivity) is established.

The PDN to which a connection is established can be preconfigured on a per-subscriber basis, or the terminal can specify it during the attach procedure. This PDN is called the default PDN. With the always-on connection function, the radio link of the connection only is released after a set amount of time has elapsed without the terminal performing any communication, and the IP connectivity between the terminal and the network is maintained. By doing this, only the radio link needs to be reconfigured when the terminal begins actual communication, allowing the connection-delay time to be reduced. Also, the IP address obtained when the terminal attaches can be used until it detaches, so it is always possible to receive packets using that IP address.

The information flow for the terminal attaching to the network up until the connection to the PDN is established is shown in Figure 2 below.

Steps (1) to (3): When the terminal establishes a radio control link for sending and receiving control signals with the eNodeB, it sends an attach request to the MME. The terminal and MME perform the required security procedures, including authentication, encryption and integrity.

Steps (4) to (5): The MME sends an update location request message to the Home Subscriber Server (HSS), and the HSS records that the terminal is connected under the MME.

Step (6): To begin establishing a transmission path to the default PDN, the MME sends a create session request to the S-GW.

Steps (7) to (8): When the S-GW receives the create session request from the MME, it requests proxy binding update to the P-GW. The P-GW allocates an IP address to the terminal and notifies the S-GW of this information in a proxy binding acknowledgement message. This process establishes a continuous core-network communications path between the P-GW and the S-GW for the allocated IP address.

Step (9): The S-GW prepares a radio access bearer from itself to the eNodeB, and sends a create session response signal to the MME. The create session response signal contains information required to configure the radio access bearer from the eNodeB to the S-GW, including information elements issued by the S-GW and the IP address allocated to the terminal.

Steps (10) to (11) and (13): The MME sends the information in the create session response signal to the eNodeB in an initial context setup request signal. Note that this signaling also contains other notifications such as the attach accept, which is the response to the attach request. When the terminal receives the attach accept in Step (11), it sends an attach complete response to the MME, notifying that processing has completed.

Step (12): The eNodeB establishes the radio data link and sends the attach accept to the terminal. It also configures the radio access bearer from the eNodeB to the S-GW and sends an initial context setup response to the MME. The initial context setup response contains information elements issued by the eNodeB required to establish the radio access bearer from the S-GW to the eNodeB.

Steps (14) to (15): The MME sends the information in the initial context setup response to the S-GW in a modify bearer request signal. The S-GW completes configuration of the previously prepared radio bearer from the S-GW to the eNodeB and sends a modify bearer response to the MME.

Through these steps, a communications path from the terminal to the P-GW is established, enabling communication with the default PDN.

If the terminal performs no communication for a set period of time, the always-on connection function described above releases the radio control link, the LTE radio data link, and the LTE radio access bearer, while maintaining the core network communications path.

After the terminal has established a connection to the default PDN, it is possible to initiate another connection to a different PDN. In this way it is possible to manage PDNs according to service.

For example the IMS PDN, which provides voice services by packet network, could be used as the default PDN, and a different PDN could be used for internet access.

To establish a connection to a PDN other than the default PDN, the procedure is the same as the attach procedure shown in Fig.2, excluding Steps (4) and (5).


Attach: Procedure to register a terminal on the network when, for example, its power is switched on.

Detach: Procedure to remove registration of a terminal from the network when, for example, its power is switched off.

Integrity: Whether the transmitted data is complete and has not been falsified. Here we refer to pre-processing required to ensure integrity of the data.

Bearer: A logical transmission path established, as between the S-GW and eNodeB.

Context setup: Configuration of information required for the communications path and communications management.


Noah said...

The call flow you've posted uses PMIP-based S5. I suggest you look at GTP-based S5 for how most LTE attaches will function.

kamal said...

During handover, the UE detaches from source eNB and attaches to target eNB.

Are these type of attach and detach the same as stated here?

Pravi said...

There are some General High level points for understanding which you have highlighted very nicely....I found it very useful and understood better!!