. 69
( 132 .)



Physical L1 Physical L1

Figure 6.64 RRC control

the assignment of radio resources. These radio resources will include physical resources
such as codes and frequencies, and transport resources such as transport block sizes and
transport format sets. A conceptual model of RRC is shown in Figure 6.64.
This section describes the different RRC states that a UE can be in. It then explains
how the UE is addressed in these different states. Subsections 1.16.3“1.16.11 describe
various procedures that are frequently performed by RRC signalling. The RRC speci¬-
cation (TS25.331) is rather lengthy, running into many hundreds of pages, thus only a
sample of all of the procedures are explained. These examples are in no particular order.
However, an example of a call life cycle in Section 6.22 places them in context with each
other, and the complementary signalling required at other levels as well. The procedures
illustrated are:

• RRC connection
• Signalling radio bearers

• RRC security mode control
• RRC paging
• Radio bearer establishment
• Transfer of NAS messages
• Cell/URA update
• Measurement reporting
• Active set update.

6.16.1 RRC mobile states
At any given instant, the mobile device may or may not have a signalling connection.
If no connection exists, the mobile device is referred to as being in idle mode, and
must establish a connection before it can proceed with any further communication to the
network. In idle mode, only the core network knows of the existence of the user, and this
is to a location area (LA) or routing area (RA). In idle mode, the location of the mobile
device is known only to the accuracy of an LA or RA and the UTRAN does not know
anything about the UE. The UE will monitor the LA or RA and if it notices that it has
changed, it must inform the core network. To do this, it must enter connected mode. The
mobile device will enter the idle state either when the signalling connection is released
or when is a failure of the radio resource connection.
When a signalling connection does exist, this is known as RRC connected mode.
However, as has been seen, depending on the activity of the mobile device this connection
can be transported either over dedicated or common/shared channels. If the mobile device
has a lot of activity with the network, then its location may be tracked to the cell level.
However, if the device sends data infrequently, then it may be tracked on a larger area
consisting of a number of cells. In this latter case if a mobile device has to be located for
a mobile terminated call, it is necessary to page the whole range of cells. However, this
tracking over a larger area does minimize the amount of location update messages sent
by the UE.
If a signalling connection does not exist then the mobile device will require to be paged
across a location area (routing area) to receive a mobile terminated call. This page will
instruct the mobile device to establish a dedicated signalling connection. If the mobile
device is to be able to receive mobile terminated calls, it must have already registered
with the core network so that paging (to indicate an incoming call, for example) can be
directed to the correct geographical area.
In the connected state the location of the mobile device is known within the core
network to the accuracy of the serving RNC (SRNC). The SRNC will then know the
location of the mobile device to either the cell or the UTRAN registration area (URA)
level. The connected state actually consists of four separate states known as CELL-DCH,

In this state a dedicated physical channel is allocated for the mobile device in the uplink
and downlink. In the uplink a dedicated transport channel, DCH, is used and in the

downlink, this may be a DCH and/or DSCH. The location of the device is known by
its SRNC to a cell level. The UE enters this state from idle mode through establishment
of an RRC connection, and from CELL-FACH through the establishment of a dedicated
channel. It may move to any other state from CELL-DCH by explicit signalling. The UE
does not generally listen to the broadcast channel in this state and any major changes to
system information will be indicated on the FACH channel.

No dedicated channel is allocated to the mobile device in this state, the RACH and
FACH are used instead for both signalling and small amounts of user data. This is ideal
for such services as SMS and multimedia messaging service (MMS). The mobile device
can monitor the BCH and may also be asked to use the CPCH to transfer data in the
uplink. The UE must constantly monitor the FACH in the downlink.

Again, the mobile device can be located to a speci¬c cell but it cannot transfer data and
must be paged by the network. If the mobile device moves cell and a cell reselection is
required, the mobile device will change to the CELL-FACH state to perform the procedure
and will then return to the CELL-PCH state. CELL-PCH allows the mobile device to use
DRX. This allows the mobile device to ˜sleep™ and only wake up at speci¬c times to mon-
itor the paging indication channel (PICH). This enables the device to consume less power.
A counter within the network can monitor the number of cell updates a mobile device
makes and if a threshold is met the mobile device may be moved to URA-PCH mode to
conserve its power and to reduce the amount of signalling. No uplink activity is possible
in the CELL-PCH state. The dedicated control channel cannot be used in this state.

In the URA-PCH state, the mobile device does not update its location every time it selects
a new cell. The URA is a new area that has been introduced into the UMTS network
to complement the LA and RA, which are still used. Unlike the LA and RA, which
are controlled from the core network, the URA is controlled within UTRAN. The core
network only knows to which RNC to send data; it is up to the UTRAN to locate a
speci¬c mobile device. Each cell in the UTRAN will belong to at least one URA and
the mobile device will be allocated a URA to which it is attached. The mobile device
will monitor the cell broadcast channel to see if the current cell broadcast is in the same
URA as the one it has been allocated. If they are the same then there is no update; if they
are different then the mobile device will enter the CELL-FACH state to perform a URA
update, which is similar to a cell update. It should be noted that a cell may broadcast a
number of URAs. By introducing this overlapping feature, the UTRAN can reduce the
number of URA updates. Thus, two mobile devices may have different URAs assigned
to them but be attached to the same cell. The dedicated control channel cannot be used
in this state. The four states and the transitions between them are shown in Figure 6.65.
CELL-PCH and URA-PCH states are really designed for packet traf¬c. For a phone
call, there is a very clearly de¬ned start and end of the transaction, so returning to idle

RRC Connected



Cell-FACH Cell-PCH

Figure 6.65 RRC Modes

mode following call hang-up could be seen as logical. However, this does mean that the
subscriber will need to be paged for a mobile terminated call over the LA; if it had been
transferred to Cell-PCH for a certain period of time it takes little in the way of resources
and can be paged directly in a single cell. However, for packet traf¬c, this is not the case
as it is impossible for the network to decide if a user has completed their data transfer.
A typical example would be web traf¬c where there are periods in between information
exchange where the user is reading the pages. It is useful to keep the UE ˜connected™ but
release any resources it is utilizing by moving the UE to one of these two states.
A scenario may go as follows. The UE is transferring a reasonable amount of data, say
email or ¬le transfer, and the UTRAN allocates a dedicated channel, moving the UE to
CELL-DCH. Subsequently, the RNC observes that the quantity of data being transferred
has signi¬cantly reduced, and instructs the UE to release its dedicated connection and
instead use the FACH/RACH or CPCH. The UE then moves into CELL-FACH. After a
further period of inactivity, the RNC instructs the UE to move to CELL-PCH, where all
resources are released; however, the RNC is still aware of the UE, as it is performing
cell updates. The RNC will generally use an inactivity timer to decide when the UE
should change states. However, this is part of the RRM strategy, which is the remit of
the manufacturer. This state would be ideal for a stationary data user, since the burden of
signalling is reduced to periodic cell updates between the UE and RNC, and DTX mode
can be used. If the user then moves, these cell updates will happen more frequently, and
the RNC may decide to reduce this further by moving the UE into the URA-PCH state.
The precise mechanisms controlling how and when a UE moves states, and into which
state, are governed by the radio resource policies in the RNC. However, these policies
may be currently considered quite subjective, with the absence of an adequate working
test environment in which to evaluate.
As previously stated, if the mobile device is in idle mode then it will still update
the core network of its location to the LA or RA. However, if the mobile device is in
connected mode, the UTRAN itself deals with the location of the device, removing some
of the burden from the core network. The core network only needs to know the location

of the device to the correct RNC; it is up to the RNC to locate the mobile device within
a cell or number of cells. If a mobile device has a connection to the packet switched
core network and another subscriber wishes to make a mobile terminated voice call to the
same mobile device, the page request is directed to the correct RNC from the MSC/VLR.
The RNC in this case will know exactly where the mobile device is since it is already in
packet switched connected mode. The RNC will use the IMSI number of this device to
coordinate the use of the existing connection for the page. If the circuit switched call is
accepted, the UE will now have two connections, one from the RNC to the CS-CN and
one to the PS-CN. There will, however, only be one instance of RRC between the mobile
device and the RNC. The IMSI is only stored in the RNC for the duration of the RRC
connection. It should be noted that if the mobile device is in CELL-PCH or URA-PCH
then paging is done in a similar method to that of paging a mobile device in idle mode.

6.16.2 UTRAN UE identi¬ers
Throughout the UTRAN, once a user is RRC connected, a number of radio network
temporary identities (RNTI) are used to identify the UE and are contained in signalling
messages between the UE and UTRAN. There are four types of RNTI:

• Serving RNC RNTI (S-RNTI): this is allocated to a UE by the SRNC when it establishes
an RRC connection (see Section 6.16.3) to the network. It is unique within that SRNC.
Should the RRC connection change, a new S-RNTI is allocated. It is used subsequently
by the UE to identify itself to the SRNC, and by the SRNC to address the UE. It is
also used by the DRNC to identify the UE to the SRNC when the Iur is being used.
• Drift RNC RNTI (D-RNTI): this is allocated to a UE by the DRNC when the UE
establishes a connection to it. The SRNC maintains a mapping between S-RNTIs and
D-RNTIs. It is only used by the SRNC to identify the UE to the DRNC and is not used
by the UE at all.
• Cell RNTI (C-RNTI): this is allocated by the controlling RNC when the UE accesses a
new cell, and is unique within that cell. It is used on common channels for the UE to
identify itself to the CRNC, and also by the CRNC to address the user.
• UTRAN RNTI (U-RNTI): this is allocated to a UE having an RRC connection and
identi¬es the UE within the whole UTRAN. A U-RNTI consists of the RNC ID of
its SRNC and its S-RNTI. The RNC ID is a unique identi¬er for the RNC within
the UTRAN.

6.16.3 RRC connection
From idle mode, once the UE is ready to connect to the network, for example in response
to a page or to perform a location update, it will send an RRC connection request message
to the network. Since this is an initial access and the UE has no other connection to use,
it will be sent on the RACH using the logical common control channel (CCCH). At

the physical layer there is a chance for collision on this channel, since more than one
mobile could attempt to access the network at the same time. The physical layer provides
a number of mechanisms to reduce the collision probability and deal with collisions,
should they occur.
The RNC will respond to an RRC connection request with an RRC connection setup
message, again on the logical CCCH, which in turn is con¬rmed by the UE with an RRC
connection setup complete message. Since the connection setup message informs the user
of the connection that the RNC has established, the UE will then use this channel (DCCH)
to reply to the RNC with the complete message. Figure 6.66, shows an example of the
establishment of a dedicated signalling connection. Here, once established, the UE can
send subsequent RRC messages over its dedicated transport channel. Note that RRC
messages are transparent to the BTS.
The RNC will allocate resources and establish bearers for the signalling using
the Node B application part (NBAP) and the access link control application protocol
(ALCAP). This will establish a connection for the subscriber between the RNC and BTS.
NBAP and ALCAP signalling is only required if a DCH is being set up; if the user
is offered a signalling connection over the RACH/FACH then this is not required. This
NBAP and ALCAP signalling is shown in Figure 6.66 as signalling bearer establishment.
Once completed, it will respond to the UE with the RRC connection setup message.
The parameters of the RRC connection request are very straightforward. It uses the
RLC transparent mode and contains the initial UE identity, which is the IMSI, TMSI or
P-TMSI, and an establishment cause to indicate why this RRC connection is being made.
It could indicate, for example, registration, for a location update, or originating interactive
call, etc. It will also include some measurements made on the pilot channel, such as the
received power level, to assist in power control procedures. A common pilot channel
measurement is the CPICH Ec/No. This has a range of values from 0 to 47, which map
to 0.5 dB measurement ranges of the CPICH Ec/Io. Table 6.16 shows these mappings.
This information is used by the BTS to select a suitable power level on which to reply
to the UE.


RRC Connection Request

signalling bearer

RRC Connection Setup


. 69
( 132 .)