Location, Routing, Service and UTRAN Area Planning Aspects in

WCDMA

Version: Draft

2/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Revision History

Version Date Change Notes

0.1 10/08/2010 Draft version

Authors The following persons have collaborated on this document:

Name Department

Pavel Romanenko NWS LTE RA E2E SA NE

3/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Contents

1. Introduction ............................................................................4

2. Concept of LA, RA and URA areas in UMTS .........................5

2.1. General Induction ..........................................................................................5

2.2. Location area definition .................................................................................6

2.3. Routing area definition ..................................................................................7

2.4. UTRAN registration area definition ..............................................................9

2.5. LAC, RAC and SAC definition ....................................................................15

2.6. Relationship between the different areas ..................................................17

2.7. Paging ...........................................................................................................19

2.8.1. RANAP paging .............................................................................................20

2.8.2. Paging type 1 ...............................................................................................21

2.8.3. Paging type 2 ...............................................................................................23

2.8. Location area update and routing area update.........................................24

2.9.1. Location and routing areas update ............................................................25

2.9.2. Routing area update ....................................................................................30

2.9.3. Combined updates.......................................................................................31

2.10. Summary of mobility management ............................................................31

3. Practical guide for UMTS LA, RA and URA areas Planning and Optimization..................................................................33

3.1. Location and Routing Areas Planning and Optimization .........................33

3.2. Service Area Planning and Optimization ...................................................42

3.3. UTRAN radio Area Planning and Optimization.........................................43

4/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

1. Introduction

The main task of this document is to provide description in detail of Location Area (LA), Routing Area (RA) and UTRAN registration area (URA) planning.

The document audience is NSN internal employers, interested in Network Planning, Initial Tuning and Optimisation.

The document operates and by theoretical foundations, dimensioning simulations results and by practical cases and examples. The document presents related parameters policy, collected from different projects shared experience. The concept of planned LA size is dependant on many factors, including with NW size and load. But, the approached understanding of areas zones planning is coming from general zones definitions that provided in chapter 2.

Chapter 2 summarizes the purpose of Mobility Management (MM) and defines the Location and Routing Area.

The chapter 2 explains theory aspects of three WCDMA areas, their interdependencies between each other and paging, and will give general concepts of their planning.

The practical approach of three area concept planning depends on current NW case, and sampled in chapter 3 of current document.

In many cased UMTS zones area parameterization is part of not initial tuning and planning, but optimization phase, required after NW implementation and loading.

The document is actual for most of cases in RU (Releases from RAS06 to RU20,) and in I-HSPA Rel.2, just some RU20 and I-HSPA features potentials are not included to practical chapter in the reason of lack of practical output’s availability from them.

5/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2. Concept of LA, RA and URA areas in UMTS

2.1. General Induction

3GPP definition

A UMTS network consists of three interacting domains:

• Core Network (CN)

• UMTS Terrestrial Radio Access Network (UTRAN)

• User Equipment (UE)

The main function of the core network is to provide switching, routing and transit for user traffic. Core network also contains the databases and network management functions. The basic Core Network architecture for UMTS is based on GSM network with GPRS. All equipment has to be modified for UMTS operation and services. The UTRAN provides the air interface access method for User Equipment. Base Station is referred as Node-B and control equipment for Node-B's is called Radio Network Controller (RNC). UMTS system page has an example, how UMTS network could be build. It is necessary for a network to know the approximate location in order to be able to page user equipment. Here is the list of system areas from largest to smallest.

• UMTS systems (including satellite)

• Public Land Mobile Network (PLMN)

• MSC/VLR or SGSN

• Location Area

• Routing Area (PS domain)

• UTRAN Registration Area (PS domain)

• Cell

• Sub cell

Definitions of LA, RA and URA areas are tightly connected with basic UMTS

network structure, and intended both in Core and Radio Network Planning. UMTS network domains interconnection visually effects on UE states and is

necessary for keeping in mind before making planning in any NW element. The Core Network (CN) logical structure consists of two services domains:

� Circuit-switched (CS) service domain � Packet-switched (PS) service domain.

For the first approach of possible UE states within different domains, the figure below gives an overview of the UE registration and connection principles within the UMTS, when the CN consists of separate PS and CS service nodes or one combined CS and PS service node.

6/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Important note is that 3G-MSC and 3G-SGSN sub domains may be physically co-located at one territory site.

Figure 2.1.UE registration and connection set-up for 3G-MSC and 3G-SGSN

2.2. Location area definition

3GPP definition

Location Areas (LA) is used to inform the core network of a UE’s location for CS services. A "Location Area" is a set of base stations that are grouped together to optimize signaling. Typically, tens or even hundreds of base stations share a single Radio Network Controller (RNC) in UMTS, the intelligence behind the base stations. RNC handles allocation of radio channels, receives measurements from the mobile phones, and controls handovers from base station to base station, included to LA. To each Location Area, a unique number called a "Location Area Code" is

7/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

assigned. The location area code is broadcast by each base station, known as Node B in UMTS, at regular intervals.

In UMTS NW structure, a location area is defined as a cluster of cells. The

cells are typically adjacent between each others. For CS services, the whole network is divided into location areas. For

example, location area is used at CN initiated paging related to CS services. A CS service related temporary identity might be assigned to the UE. This temporary identity is then unique within a LA. A location area consists at minimum of one cell and at maximum of the entire VLR area.

In UMTS, some operators define location areas equal to RNC or even MSC areas, but this is not mandatory, the total LA planning is more complex.

The optimal size of a location area in terms of cells is limited by the LA update load on one side and on the paging load on the other side. A small location area will require less signaling load for paging on RNC and MSC, but high load for LA updates in RNC and MSC. A large location area will result in a high paging load for the RNC and MSC, but low load for LA updates will appear in the BSC/RNC and the MSC. This leads to a typical optimization problem.

2.3. Routing area definition

3GPP definition

Routing Area (RA) is used to inform the core network of a UE’s location for PS services. The Routing Area is the PS domain equivalent of the location area. A "routing area" is normally a subdivision of a "location area". Routing areas are used by mobiles which are GPRS-attached. GPRS ("General Packet Radio Services") new data transmission technology is optimized for "bursty" data communication services, such as wireless internet/intranet, and multimedia services. It is also known as IP ("Internet Protocol") because it will connect users directly to Internet Service Providers (ISP). The bursty nature of packet traffic means that more paging messages are expected per mobile, and so it is worth knowing the location of the mobile more accurately than it would be with traditional circuit-switched traffic. A change from routing area to routing area (called a "Routing Area Update") is done in an almost identical way to a change from location area to location area. The main differences are that the "Serving GPRS Support Node" (SGSN) is the element involved.

CN uses routing areas (RA) for PS services. The routing area is used at CN

initiated paging related to PS services. A PS service related temporary identity might be allocated to the UE. This temporary identity is then unique within a RA.

8/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

The same as LA, routing area is defined as a cluster of cells. The cells are typically adjacent because the target is to maximize the dwelling time within one routing area (which implicitly prevents a large number of routing area updates).

Routing areas are used for the packet switched part of the network to localize the subscriber. Routing area consists of at minimum one cell and at maximum of the entire location area. More than one routing area can be defined per location area.

Minimum and maximum values in terms of number of cells are limited by paging results in signaling load on RNC and SGSN on the one side. On the other side the routing area update load of the SGSN limits the number of cells, i.e. the size of a routing area.

In the packet oriented (PO) domain the change of a UE from one cell to another can initiate two different types of ‘handover’ (either one of it or both) in the CN (Core Network): a RAU (Routing Area Update) or SRNS relocation.

� Routing Area Update For inter-system changes, routing area changes inside GPRS, routing area changes inside UMTS in UMTS PMM idle state and after SRNS relocation with routing area change in UMTS PMM connected state. In the last case only a subset of the routing area update procedure will be executed to update the HLR.

� SRNS Relocation In UMTS PMM connected state. Execution decided by UTRAN. After SRNS relocation, a RAU in PMM connected state is triggered in order to update the location towards HLR if the routing area has been changed.

In principle, the handover handling of the 3G SGSN has the same function as in the 2G SGSN to provide a smooth service in case of roaming. There are some differences, which results from the Iu interface and from the used handover strategy in UMTS.

In GPRS the MS decides by itself to perform a cell change without interaction to the BSS. This is valid especially for a change between cells of two BSS. The SGSN takes care of undelivered or duplicated user packets. Different to that, in UMTS the UTRAN will be responsible for the decision of a cell respectively RNC change if the UE will be in the PMM connected state. The UTRAN may decide to change the point of attach to the core network (SGSN) at any time which leads to an RNC relocation.

The above-mentioned RNC relocation takes place only if an established radio access bearer exists (UMTS PMM connected state). Otherwise (UMTS PMM idle state) a normal routing area update will be done which may lead to an SGSN change similar to that in the 2G SGSN.

Routing area update in UMTS PMM connected will follow the relocation procedure if the RA was changed.

9/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2.4. UTRAN registration area definition

3GPP definition

“UTRAN registration area" is defined as an area, covered by a number of cells, known internally only in UTRAN to provide a layer of abstraction between cells and the routing area. URA contains one or more cells and a routing area contains one or more URA. The URA is used to track the location of an UE within UTRAN. URA is uniquely identified using the URA Identity parameter. Upon reception of RRC message UTRAN Cell Update or UTRAN Registration Area Update from a UE the drift RNS inserts necessary information received in the RRC message to the CELL/URA UPDATE INDICATION message and sends the message to the serving RNS.

The location of the UE is known by UTRAN Registration Area level according

to last RA update. Some examples with URA state are provided below First figure shows the RRC states in UTRAN RC Connected Mode, including

transitions between UTRAN RRC connected mode and GSM connected mode for CS domain services, and between UTRA RRC connected mode and GSM/GPRS packet modes for PS domain services. It also shows the transitions between Idle Mode and UTRAN RRC Connected Mode and furthermore the transitions within UTRAN RRC connected mode.

Figure 2.2 RRC States and State Transitions including GSM

10/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

CELL_DCH state is characterized by: � a dedicated physical channel is allocated to the UE in uplink and downlink � the UE is known on cell level according to its current active set � dedicated transport channels, downlink and uplink (TDD) shared transport channels, and a combination of these transport channels can be used by the UE CELL_FACH state is characterized by: � No dedicated physical channel is allocated to the UE. � The UE continuously monitors a FACH in the downlink. � the UE is assigned a default common or shared transport channel in the uplink (e.g. FACH) that it can use anytime according to the access procedure for that transport channel � the position of the UE is known by UTRAN on cell level according to the cell where the UE last made a cell update � in TDD mode, one or several USCH or DSCH transport channels may have been established CELL_PCH state is characterized by: � no dedicated physical channel is allocated to the UE � the UE selects a PCH with the algorithm, and uses DRX for monitoring the selected PCH via an associated PICH � no uplink activity is possible � the position of the UE is known by UTRAN on cell level according to the cell where the UE last made a cell update in CELL_FACH state URA_PCH State is characterized by: � no dedicated channel is allocated to the UE � the UE selects a PCH with the algorithm, and uses DRX for monitoring the selected PCH via an associated PICH � no uplink activity is possible � the location of the UE is known on UTRAN Registration area level according to the URA assigned to the UE during the last URA update in CELL_FACH state The next example is Cell reselection procedure:

11/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Figure 2.3 States and procedures in the cell reselection process in connected

mode

When a cell reselection is triggered, the UE evaluates the cell reselection criteria based on radio measurements, and if a better cell is found that cell is selected, procedure Cell reselection. If the change of cell implies a change of radio access technology, the RRC connection is released, and the UE enters idle mode of the other RAT. If no suitable cell is found in the cell reselection procedure, the UE eventually enters idle mode. When an Initial cell reselection is triggered, the UE shall use the Initial cell reselection procedure to find a suitable cell. One example where this procedure is triggered is at radio link failure, where the UE may trigger an initial cell reselection in order to request re-establishment of the RRC connection. If the UE is unable to find a suitable cell, the UE eventually enters idle mode.

The URAs can be overlapping or even hierarchical. The same cell may belong to several different URAs, and the UEs in that cell may have been registered to different URAs. SIB 2 contains a list of URA identities indicating which URAs this cell belongs to. This arrangement is done to further reduce the amount of location update signaling because now the UEs moving back and forth in the boundary area of two URAs do not have to update their URA location information if the boundary cells do belong to both URAs.

12/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Figure 2.4 Overlapping UTRAN Registration Areas

The figure above illustrates a definition of three overlapping URAs with the primary and the secondary URA IDs given for each cell:

• URA1 includes the cells A, B, D, E, F, G, H and I • URA2 includes the cells B, C, F, G, I and J • Cells F, G and I are URA2 cells, which overlap on URA1 (ID of the

secondary URA defined as URA1) • Cell B is the URA1-cell overlapping on URA2, etc. When the UE performs an URA Update through the cell D, the URA identity

of URA1 is assigned to this UE (the UE is registered on URA1). If the UE is paged, the paging procedure must be executed, not only through the cells with the Primary URA-ID defined as URA1, but also through the overlapping (Secondary URA-ID) URA2-cells (F, G and I).

If the UE moves from URA1 to URA2 via the cells F and G (route: D=>E=>F=>G), the RRC:URA Update procedure is not initiated since the URA identity of URA1 is broadcast in these URA2-cells and the UE stays registered on URA1.

When the UE registered on URA2 moves to the opposite direction (route: G=>F=>E=>D), the URA Update is initiated when the UE selects the Cell E and the RNC registers the UE on URA1.

13/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

With Radio Network Planning of the URAs the following issues are possible:

� URA can cover geographical areas e.g. highways planned to be in same URA.

� URA is typically including several Cells, but smaller than Location / Routing Area.

� One cell can belong to several URAs. � URA may overlap (geographically) and for that purpose up to 8 URA identities can be broadcasted in one cell.

� UE in URA_PCH state has always only one valid URA, which is the primary URA of the cell (which UE has entered URA_PCH state, or in which the UE has initiated the last URA Update procedure).

� UE does not initiate URA Update procedure in the cells where the assigned

� URA is included in the URA-list. � When defining overlapping URAs it should be noted the UEs will be paged via all cells on these URAs, and this may notably increase the paging load in

� UTRAN (reduction of Cell Update signaling and the impact caused to the paging-load must be kept in balance).

Activating URA_PCH RRC state feature, gives following benefits to operator:

� high Mobility UE’s (which are moving fast) � URA_PCH state is used instead of the Cell_PCH for the “high mobility” UE

� amount of cell updates decreased � three times longer UE stand-by time with NSN WCDMA RAN � better UE battery performance with always-on applications � with Cell_PCH/URA_PCH implemented, a short activation timer can be used in Cell_FACH state while maintaining good browsing experience

� keep-alive sent over Cell_FACH, state transition to Cell_DCH avoided � small signaling messages could be transferred via RACH/FACH � without dedicated radio / Iub resource allocation

URA_PCH RRC Connected Mode was innovated by URA_PCH (RAN833) feature (introduced in NSN from RU 10). That feature makes Reduction of RACH signaling load and UE power consumption.

Thanks to feature signaling load caused by the cell updates can be reduced when the fast moving UEs are transferred to the URA_PCH state.

Starting from RU 10 NSN implementation supports all states within RRC connected mode. URA_PCH helps to reduce the signaling load by reducing the requirement for cell update procedures resulting from UE mobility. UE are moved from CELL_PCH to URA_PCH if they complete more than a specific number of cell re-selections within a specific time interval. Benefiting from URA_PCH requires the radio network planner to define UTRAN Registration

14/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Areas (URA) in addition to Location Areas and Routing Areas. The drawback of using URA_PCH is a slightly increased paging load, i.e. paging messages are broadcast across URA rather than across individual cells. URA can be defined to overlap with one another to avoid ping-pong scenarios for UE. A single cell can belong to a maximum of 8 URA. The first URA listed within SIB2 represents the primary URA while the remaining URA represent secondary URA. The primary URA is selected when a UE first enters URA_PCH and also subsequent to a cell update procedure. The UE only needs to complete a cell update procedure if the selected URA does not appear within the list of URA within SIB2.

In URA_PCH state the location of the UE is known by the URA level. There is no dedicated channel allocated for the UE in URA_PCH state. The mobility in this state is handled by RRC in URA Update procedure. In this state the UE uses discontinuous reception (DRX) and monitors paging messages from RNC. If the network wants to initiate any downlink activity, it needs to make a paging request in all cells on the PCCH logical channel within the URA where the location of the UE is known. Any Activity causes the UE to be transferred to the Cell_FACH state, where the uplink access is performed on RACH.

In URA_PCH state, the UE:

� listens to the PCH transport channel for the decoding of paging and notification messages sent by the RAN,

� listens to the BCH transport channel of the serving cell for the decoding of system information messages,

� initiates a URA updating procedure on URA change, � initiates a periodic URA update after timer T305 has expired in UE (if configured),

� initiates a Cell Update in case of paging response (via FACH state) � initiates a Cell Update in case of UL data transmission (If the UE needs to transmit anything to the RNC, it moves to the Cell_FACH state and executes a RRC: Cell Update procedure)

Summarized, comparing between CELL_PCH and URA_PCH is:

� CELL_PCH alone minimizes paging load but causes more cell updates than URA_PCH

� URA_PCH alone causes very high paging load if URA = RNC area and provide paging channel capacity approx 400/s

� if URA is smaller than RCN area, then separate planning is needed for URA

� the optimal solution combines CELL_PCH + URA_PCH to minimize paging and cell update load; URA_PCH alone is not a good solution without Cell_PCH

15/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

NSN solution:

� both Cell PCH and URA PCH are supported � NSN RNC can detect if UE is moving and keep that UE in URA PCH

2.5. LAC, RAC and SAC definition

The routing area code (RAC) combines an integer value between 0 and 255 defined inside a LA.

The routing area identity (RAI), defined by an operator, identifies one or several cells.

RAI is broadcast as system information and is used by the UE to determine, when changing cell, if an RA border was crossed. If that is the case, the UE initiates the RA update procedure.

A routing area is a subset of one, and only one, location area (LA), meaning that a RA cannot span more than one LA. A RA is served by only one SGSN.

The following rules apply for the Routing Area Identity (RAI):

� RAC is only unique when presented together with LAC � RAI is only unique when presented together with LAI � LAI = MCC + MNC + LAC � RAI = MCC + MNC + LAC + RAC

Where, “MCC + MNC” is the PLMN identity.

In order to respect the functional separation of core network and radio network and to confine the radio (and thus cell coverage) aspects within UTRAN, 3GPP WG RAN3 has not introduced the concept of UMTS Cell or UMTS Cell Global Identity over Iu.

Instead of this, the concept of Service Area (SA) has been introduced. Such a Service Area can be mapped onto one or more “cells” within a location area. This mapping onto cells, however, is invisible to core network and will be handled within UTRAN. A cell is allowed to belong to more than one service area.

A Service Area Identifier (SAI) is described by following main rules:

� SAI = MCC + MNC + LAC + SAC � The Service Area Code (SAC) has a length of two octets and is unique within the location area

� The other codes denote the Mobile Country Code (MCC), the Mobile Network Code (MNC), and the Location Area Code (LAC)

The service area can be configured in the way that it matches the cell for service information broadcast.

Release 99 requires that for the broadcast (BC) domain the service area matches one cell. For the CS and PS domain, the service area can consist of

16/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

more than one cell, e.g. cover the cells of one base station or more base stations.

Practically, Service Areas are used for emergency calls location based services and the Service Area Broadcast (SAB) feature which makes use of a third S-CCPCH and Service Area Codes for SAB (SACB).

This system feature specification defines upper layer coordination for the Service Area Broadcast. WCDMA Service Area Broadcast (SAB) service corresponds to the SMS Cell Broadcast service defined for the GSM. Since cells are not known by the CN, service area definition is used instead. The new terminology has not changed the functionality of the service. As in 2G based documents the term "Cell Broadcast" is still widely used, the terms "Service Area Broadcast" and "Cell Broadcast" can be understood as synonyms so far.

The Cell Broadcast Service (CBS) is a service, which enables a provider of information to submit short messages for broadcasting to a specified area within the PLMN. These messages could be used for informing of, for example, PLMN news, emergencies, traffic reports, road accidents, delayed trains, weather reports, theatre programs, telephone numbers or tariffs.

CBS permits a number of unacknowledged general CBS messages to be broadcast to all receivers within a particular region. CBS messages are broadcast to defined geographical areas known as Service Area. Service Area will consist of one cell. CBS messages originate from the Cell Broadcast Center and Radio Network Controller can interface to one or more Cell Broadcast Center.

The cell broadcast architecture in UMTS: � The basic network structure in the RAN is containing the RNC and BTS.

� The Cell Broadcast Center (CBC) is part of the core network and connected via lu.

� Service Area Broadcast Protocol (SABP) is used between the CBC and RNC for CBS message transferring.

� Broadcast/Multicast Control protocol is used between the RNC and MS for the message broadcasting on the radio interface.

� Architecture for the Cell Broadcast Service is given in Figure 2.5.

Uu

Cell Broadcast

Center

(CBC)

RAN

RNC

BTS

BTS MS

MS

Iub

IuBC

Figure 2.5 Architecture for the Cell Broadcast Service in UMTS

17/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Broadcast messages which are transmitted by the RNC to the MS include two

types of messages: BMC CBS Message which contains user information and BMC Schedule Message which contains scheduling of the BMC CBS messages. BMC Schedule messages will be used to implement DRX function for the Cell Broadcast Service and will be generated by the RNC.

The distribution of broadcast information relates on the mapping between service area and cell will be controlled by the RNC. The provision of this mapping information is an O&M function.

2.6. Relationship between the different areas

3GPP definition

• A RA consists of a number of cells belonging to RNC that are connected to the same CN node

• A LA consists of a number of cells belonging to RNC that are connected to the same CN node

• A RA is handled by only one CN serving node, i.e. by one 3G-SGSN;

• A LA is handled by only one CN serving node, i.e. by one 3G-MSC/VLR

• A service area is a subset of a location area

• A URA is a subset of a routing area

The mapping between LA and RNC is handled within the MSC/VLR to which

the LA is assigned. The mapping between RA and RNC is handled within the SGSN/SLR owning this RA. The mapping between LA and cells and respectively between RA and cells is handled within RNC.

Three areas definitions picture is shown below in hierarchical area concept figure.

18/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Figure 2.6 Hierarchical area concepts.

The size of LA/RA areas depends on specific NW case, those areas

can be different based on NW circumstances, and current picture shows just that they are responsible for different services, signaling messages.

Figure 2.7 LA/RA areas interdependence

19/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2.7. Paging

3GPP definition

In computer operating systems there are various ways in which the operating system can store and retrieve data from secondary storage for use in main memory. One such memory management scheme is referred to as paging. In Mobile Telecommunication, the thermion paging means indication to the user about the need for transaction. Paging procedure is always initiated by the network and is made in order to find out the actual location of the user. The network request is carried out in the logical channel Paging Control Channel (PCH).

This procedure is used to transmit paging information to selected UE in idle

mode using the paging control channel (PCCH). A normal paging message to the RNC contains information on the area in

which the paging message shall be broadcast. This is indicated with the Paging Area ID parameter. Value of LA of CS services (or RA for PS services) is taken from the “Cell Identifier List”. If WCDMA cell is paged, the Cell Identifier List contains just one dummy cell to derive the LA. Contrary to GSM, the RNC itself builds up the list of cells to be paged.

Important is that paging is processing completely independently for CS and PS services.

Paging is necessary for the CN to reach the UE from the specific location or routing area. In idle mode, paging is always initiated by the CN. In CS paging, the CN and further the RNC broadcast paging messages through base stations of the location area in which the UE is situated. In PS paging, the CN and further the RNC broadcast paging messages through base stations of the routing area in which the UE is situated.

To save on power consumption, the UE can use discontinuous reception (DRX). This means that the network pages the UE according to a preset sequence of frames during specific intervals. This preset frame sequence interval is called a DRX cycle. DRX cycles can vary in length, and in case the UE is connected to two different CN domains with different DRX cycle lengths, the UE uses the shortest DRX cycle. The UE can store each domain-specific DRX cycle of each CN it is currently attached to.

If the CN sends a specific DRX cycle length coefficient, the RAN uses it in the paging message. If there is no CN-specific coefficient, the RAN uses the default value for CN-specific DRX cycle length coefficient in the radio network database.

The RNC sends paging requests to all WCDMA BTSs which belong to the paging area where the UE is currently registered. The cells in a single BTS can belong to different paging areas.

20/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Each paging message on the Iu interface involves only the UE and therefore, the RNC has to pack the pages into the relevant radio interface paging message. The paging procedure is divided into two parts:

� the part from Core Network (CN) to RNC on the Iu interface (RANAP) � the UTRAN internal part on the Iub interface (NBAP) part

2.8.1. RANAP paging

3GPP definition

The CN initiates the procedure via sending a message PAGING. This message shall contain information necessary for RNC to be able to page the UE, like following:

1) CN Domain Indicator 2) International Mobile Subscriber Identity (Permanent NAS UE Identity) 3) Temporary Mobile Subscriber Identity (Temporary UE Identity) 4) Paging Area 5) Paging Cause 6) Non-Searching Indicator

More detail description is:

1) The CN Domain Indicator IE is used by the RNC to identify from which CN the PAGING message originates.

2) The Permanent NAS UE Identity (IMSI) shall be used by the UTRAN paging coordination function to check if a signaling connection towards the other CN domain already exists for this UE. In that case, the radio interface paging message can be sent via that connection instead of using the paging broadcast channel.

3) The Temporary UE Identity IE (TMSI) is the identity of the user that shall be used over the paging channel. If the Temporary UE Identity IE is not included in the PAGING message, the RNC shall use the Permanent UE Identity instead.

4) The Paging Area IE shall be used by the RNC to identify the area in which the radio interface paging message shall be broadcast in case no signaling connection, as described above, already exists for the UE. If the Paging Area IE is not included in the PAGING message, the whole RNC area shall be used as Paging Area.

5) The Paging Cause IE shall indicate to the RNC the reason for sending the PAGING message. The paging cause is transferred transparently to the UE.

6) The Non Searching Indication IE shall be used by the RNC to decide whether the UTRAN paging co-ordination function needs to be activated or

21/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

not. In the absence of this IE, UTRAN paging co-ordination shall be performed.

It should be noted that each PAGING message on the Iu interface relates to only one UE and therefore the RNC has to pack the pages into the relevant radio interface paging messages.

The core network is responsible for the paging repetition over the Iu interface in case of non-successful paging.

For each of UTRAN state, sub-cases can be individuated (with essential counters). From the picture above we have evidence that some overlapping zones are possible (e.g. latency in RRC setup or mismatch from repetition cycles in CN and UTRAN or LA/RA with more than 1 RNC.

Figure 2.8.Paging types 1 and 2 UTRAN state sub-cases with essential counters

Different paging procedures are available for idle and connected cases. UEs in Idle Mode (or Cell-PCH or URA-PCH) listen to PICH once per DRX-cycle UEs in Connected Mode can receive the paging directly on DPCH and it is not necessary listen PCH. Moreover when a UE in idle mode state is paged, the message is broadcasted to all UEs in the LA or RA because the position of UE is known at this level from the network; while if the UE is in Connected mode the position in note by the UTRAN Registration Area that is typically a subset of the LA and RA.

2.8.2. Paging type 1

This procedure is used to transmit paging information to selected UE in idle mode, CELL_PCH or URA_PCH state using the paging control channel (PCCH). Upper layers in the network may request paging, to e.g. establish a signaling connection. UTRAN may initiate paging in CELL_PCH or URA_PCH state, to trigger UE state.

In addition, UTRAN may initiate paging in idle mode, CELL_PCH and URA_PCH state to trigger reading of updated system information.

22/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

The UE shall in idle mode, CELL_PCH state and URA_PCH state receives the paging information for all monitored paging occasions. For an UE in idle mode, the paging occasions are specified in TS 25.304 and depend on the IE "CN domain specific DRX cycle length coefficient". For an UE in CELL_PCH state and URA_PCH state the paging occasions depend also on the IE "UTRAN DRX Cycle length coefficient" and the IE "DRX indicator.

Figure 2.9 RRC Idle Paging Type1 – Signaling

Excluding the RRC connection failures, the paging procedure can fail for

the following reasons: 1. UE out of coverage 2. signaling in the transmission network failure (negligible); 3. overload in RNC units (ICSU and RRMU, negligible); 4. PCH blocking

When the UE receives a PAGING TYPE 1 message, it checks each occurrence of the IE "Paging record".

For each included paging record the UE shall compare the included identity with the identity of the UE according to different scenarios in different cases, which are listed: 1) An idle mode UE scenario is used:

• if the IE "paging originator" is CN, compare the included identities of type CN UE identity

• with all of its allocated CN UE identities

• for each match, forward the identity and paging cause to the upper layer entity indicated

• by the IE "CN domain identity"

• if the IE "paging originator" is UTRAN, ignore that paging record

23/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2) A connected mode UE scenario is used:

• if the IE "paging originator" is UTRAN, compare the included identities of type "UTRAN

• originator" with its allocated U-RNTI

• for each match, the UE shall enter CELL_FACH state and perform a cell update procedure

3) "Paging response" case scenario is used If the IE "paging originator" is CN, ignore that paging record

2.8.3. Paging type 2

This procedure, also called UE dedicated paging, is used to transmit dedicated paging information to one UE in connected mode in states CELL_DCH and CELL_FACH. Upper layers in the network may request initiation of paging, e.g. to establish a signaling connection.

Since paging type 2 is only required for UE in CELL_DCH and CELL_FACH state in which the location of the UE is known on cell level, paging type 2 is not in the scope of mobility management for location and routing area planning.

Generally, signaling messages scenarios for CELL_DCH and CELL_FACH states are quite similar, but have some specified difference, which comes from existing if CELL_FACH scenario RB Reconfiguration step.

In CELL_DCH case, the potential reasons of paging procedure failure are: 1. signaling in the transmission network failure (negligible) 2. overload in RNC units (ICSU and RRMU, negligible)

Figure 2.10 Cell-DCH Paging Type2 - Signaling

24/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Figure 2.11Cell-FACH Paging Type2 – Signaling

In Cell-FACH case, the additional messaging is adding, so the following

reasons can initiate failure of paging procedure: 1 signaling in the transmission network failure (negligible) 2 overload in RNC units (ICSU and RRMU, negligible) 3 RB reconfiguration failure

2.8. Location area update and routing area update

3GPP definition

In Cellular networks each base station covers a small geographical area which is part of a uniquely identified location area. By integrating the coverage of each of these base stations, a cellular network provides radio coverage over a much wider area. A group of base stations is named a location area, or a routing area.

The location update procedure allows a mobile device to inform the cellular network, whenever it moves from one location area to the next. Mobiles are responsible for detecting location area codes. When a mobile finds that the location area code is different from its last update, it performs another update by sending to the network, a location update request, together with its previous

25/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

location, and it’s Temporary Mobile Subscriber Identity (TMSI). There are several reasons why a mobile may provide updated location

information to the network. Whenever a mobile is switched on or off, the network may require it to perform an IMSI attach or IMSI detach location update procedure. Also, each mobile is required to regularly report its location at a set time interval using a periodic location update procedure. Whenever a mobile moves from one location area to the next while not on a call, a random location update is required. This is also required of a stationary mobile that reselects coverage from a cell in a different location area, because of signal fade. Thus a subscriber has reliable access to the network and may be reached with a call, while enjoying the freedom of mobility within the whole coverage area.

When a subscriber is paged in an attempt to deliver a call or SMS and the subscriber does not reply to that page then the subscriber is marked as absent in both the MSC/VLR and the HLR (Mobile not reachable flag MNRF is set). The next time the mobile performs a location update the HLR is updated and the mobile not reachable flag is cleared.

LA update procedure is processing by the WCDMA mobile subscriber

(MS) via the 3G-MSC if the WCDMA UE changes the location area or if a timer has expired. The procedure determines the location of the UE (VLR address) for the HLR and the authentication parameters of the UE for VLR concerned. Therefore, a LA update is performed when:

• The location area of the UE has changed

• A periodical update is required due to expiration of a time trigger

The routing area update procedure is initiated by the WCDMA UE via the 3GSGSN/SLR if the WCDMA UE changes the routing area. The procedure determines the routing area of the WCDMA UE for the HLR and the authentication parameters of the WCDMA UE for the 3G-SGSN/SLR concerned. Hence, a routing area update is performed when the routing area of the UE changed.

2.9.1. Location and routing areas update

The location updating procedure is always initiated by the mobile station. In the case that the mobile station is initiating an emergency call but, due to

cell re-selection or redirection by the network, it moves to a different LAI then the mobile station may delay the location updating procedure in the new LA until after the emergency call is completed.

LAU (Location area update) is initiated by the UE to inform the CS service domain of the CN that the UE has entered a new location area. In case the new LA is in an area served by another CN node, the location area update also triggers the registration of the subscriber in the new CN node and a LA update for CS services towards the HLR.

26/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

LAU is only initiated by the UE when the UE is in state CS-IDLE, and this independently of the PS state. If the UE is CS-IDLE but RRC connected, which means that the UE is in PS-CONNECTED state, location area update is initiated by the UE when it receives information indicating a new location area.

The location updating procedure is a general procedure which is used for the following purposes:

- Normal location updating - Periodic updating - IMSI attach The normal location updating procedure is used to update the registration of

the actual Location Area of a mobile station in the network. The location updating type information element in the LOCATION UPDATING REQUEST message shall indicate normal location updating. Only applicable for mobile stations supporting VGCS listening or VBS listening: A mobile station in RR group receive mode is in the MM IDLE state, sub state RECEIVING GROUP CALL (NORMAL SERVICE) or RECEIVING GROUP CALL (LIMITED SERVICE). To perform a location updating, the MS in RR group receive mode shall leave the group receive mode, establish an independent dedicated RR connection to perform the location updating as described above and return to the RR group receive mode afterwards.

The normal location updating procedure shall also be started if the network indicates that the mobile station is unknown in the VLR as a response to MM connection establishment request.

To limit the number of location updating attempts made, where location updating is unsuccessful, an attempt counter is used. The attempt counter is reset when a mobile station is switched on or a SIM/USIM card is inserted.

Upon successful location updating the mobile station sets the update status to UPDATED in the SIM/USIM, and stores the Location Area Identification received in the LOCATION UPDATING ACCEPT message in the SIM/USIM. The attempt counter shall be reset.

The Mobile Equipment shall contain a list of "forbidden location areas for roaming", as well as a list of "forbidden location areas for regional provision of service". These lists shall be erased when the MS is switched off or when the SIM/USIM is removed, and periodically (with period in the range 12 to 24 hours). The location area identification received on the BCCH that triggered the location updating request shall be added to the suitable list whenever a location update reject message is received with the cause "Roaming not allowed in this location area" or with the cause "Location Area not allowed". The lists shall accommodate each 10, or more location area identifications. When the list is full and a new entry has to be inserted, the oldest entry shall be deleted.

In a shared network, the MS shall choose one of the PLMN identities as specified in 3GPP TS 23.122 [14]. The MS shall construct the Location Area Identification of the cell from this chosen PLMN identity and the LAC received on the BCCH. If the constructed LAI is different from the stored LAI, the MS shall initiate the location updating procedure. Whenever a LOCATION UPDATING REJECT message with the cause "PLMN not allowed" is received by the MS, the PLMN identity used to construct the LAI which triggered the location

27/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

updating procedure shall be stored in the "forbidden PLMN list". Whenever a LOCATION UPDATING REJECT message is received by the MS with the cause "Roaming not allowed in this location area", "Location Area not allowed", or "No suitable cells in Location Area", the constructed LAI which triggered the location updating procedure shall be stored in the suitable list.

The Mobile Equipment shall store a list of "equivalent PLMNs". This list is replaced or deleted at the end of each location update procedure, routing area update procedure and GPRS attach procedure. The stored list consists of a list of equivalent PLMNs as downloaded by the network plus the PLMN code of the registered PLMN that downloaded the list. The stored list shall not be deleted when the MS is switched off. The stored list shall be deleted if the SIM/USIM is removed. The maximum number of possible entries in the stored list is 16.

Periodic updating may be used to notify periodically the availability of the mobile station to the network. Periodic updating is performed by using the location updating procedure. The location updating type information element in the LOCATION UPDATING REQUEST message shall indicate periodic updating.

The procedure is controlled by the timer T3212 in the mobile station. If the timer is not already started, the timer is started each time the mobile station enters the MM IDLE sub state NORMAL SERVICE or Attempting TO UPDATE. When the MS leaves the MM Idle State the timer T3212 shall continue running until explicitly stopped.

The timer is stopped (shall be set to its initial value for the next start) when: - LOCATION UPDATING ACCEPT or LOCATION UPDATING

REJECT message is received; - AUTHENTICATION REJECT message is received; - the first MM message is received, or security mode setting is

completed in the case of MM connection establishment, except when the most recent service state is LIMITED SERVICE;

- Mobile station has responded to paging and thereafter has received the first correct layer 3 messages except RR message;

- Mobile station is deactivated (i.e. equipment powered down or SIM/USIM removed).

When the timer T3212 expires, the location updating procedure is started and the timer shall be set to its initial value for the next start. If the mobile station is in other state than MM Idle when the timer expires, the location updating procedure is delayed until the MM Idle State is entered.

If the mobile station is in service state NO CELL AVAILABLE, LIMITED SERVICE, PLMN SEARCH or PLMN SEARCH-NORMAL SERVICE when the timer expires, the location updating procedure is delayed until this service state is left.

In A/Gb mode and GERAN Iu mode, the (periodic) location updating procedure is not started if the BCCH information at the time the procedure is triggered indicates that periodic location shall not be used. The timeout value is broadcasted in the L3-RR SYSTEM INFORMATION TYPE 3 message on the BCCH, in the Control channel description IE.

28/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

In UTRAN Iu mode, the (periodic) location updating procedure is not started if the information on BCCH or in the last received dedicated system information at the time the procedure is triggered indicates that periodic location shall not be used. The timeout value is included in the CS domain specific system information element.

The T3212 timeout value shall not be changed in the NO CELL AVAILABLE, LIMITED SERVICE, and PLMN SEARCH and PLMN SEARCH-NORMAL SERVICE states.

When a change of the T3212 timeout value has to be taken into account and the timer is running (at change of the serving cell or, change of the broadcast value of T3212), the MS shall behave as follows:

Let t1 be the new T3212 timeout value and let t be the current timer value at the moment of the change to the new T3212 timeout value; then the timer shall be restarted with the value t modulo t1.

When the mobile station is activated, or when a change of the T3212 timeout value has to be taken into account and the timer is not running, the mobile station shall behave as follows:

Let t1 be the new T3212 timeout value, the new timer shall be started at a value randomly, uniformly drawn between 0 and t1.

The IMSI attach procedure is the complement of the IMSI detach procedure. It is used to indicate the IMSI as active in the network.

In A/Gb mode and GERAN Iu mode, a flag (ATT) is broadcast in the L3-RR SYSTEM INFORMATION TYPE 3 message. It indicates whether attach and detach procedures are required to be used or not.

In UTRAN mode, a flag (ATT) is included in the CS domain specific system information element. It indicates, whether attach and detach procedures are required to be used or not.

The IMSI attach procedure is invoked if the detach/attach procedures are required by the network and an IMSI is activated in a mobile station (i.e. activation of a mobile station with plug-in SIM/USIM, insertion of a card in a card-operated mobile station etc.) within coverage area from the network or a mobile station with an IMSI activated outside the coverage area enters the coverage area. The IMSI attach procedure is used only if the update status is UPDATED and if the stored Location Area Identification is the same as the one which is actually broadcasted on the BCCH of the current serving cell. In a shared network, the MS shall choose one of the PLMN identities as specified in 3GPP TS 23.122 [14]. The MS shall use the IMSI attach procedure only if the update status is UPDATED and the stored Location Area Identification is equal to the combination of the chosen PLMN identity and the LAC received on the BCCH. Otherwise a normal location updating procedure is invoked independently of the ATT flag indication.

IMSI attach is performed by using the location updating procedure. The location updating type information element in the LOCATION UPDATING REQUEST message shall in this case indicate IMSI attach.

Example of LAU is shown on picture below. There thermions “Old” and “New” mean network elements from/to which LA

update is coming.

29/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Figure 2.12 Location area update example

30/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2.9.2. Routing area update

Routing area update is initiated by the UE to inform the PS service domain of the core network that the UE has entered a new routing area. In case the new routing area is in an area served by another CN node, the routing area update also triggers the registration of the subscriber in the new CN node and a location update for PS services towards the HLR. Routing area update is initiated by the UE when the UE is in state PS-IDLE, independently of the CS state. If the UE is PS-IDLE but RRC connected, which means that the UE is in CS-CONNECTED state, routing area update is initiated by the UE when it receives information indicating a new routing area.

When the UE is in PS-CONNECTED state the UE initiates RA update when RAI in Mobility Management (MM) system information changes.

Figure 2.13 Routing area update example

31/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

The load resulting from a location/routing area update in relation to a paging message can in a first step be estimated based for instance on the number of messages between the network elements which is about 7 to 1.

The paging load per area within the UTRAN corresponds to the number of paging requests in this area. The larger the area, the more paging messages are necessary. Although the number of paged MS is constant in the whole network, the MS has to be page within the complete paging region resulting in a paging load in each cell of the region.

2.9.3. Combined updates

WCDMA paging for CS and PS is completely separated. A combined updated can be initiated by the SGSN. Information is exchanged with the 3G-MSC/VLR via the Gs interface. Combined mobility management between CS and PS domain enables a more efficient usage of network resources. The Gs interface interfaces the MSC / VLR to the SGSN is used to convey some CS related procedures via the SGSN. The aim is a more efficient coordination of PS and CS services and functionality. Combined mobility management procedure means that a combined attach, combined RA / LA update and CS paging via SGSN only need to be performed in the PS domain. If the circuit-switched MSC triggers paging over the Gs interface, knowing that a subscriber is also attached to an SGSN, only a routing area or even one cell has to be paged via the SGSN. During a procedure, the SGSN informs the MSC/VLR of the action.

Special mobile stations such as class A and class B mobile stations can initiate combined mobility management. Class A mobile stations (MS) support the usage of data connections at the same time as speech connections, whereas class B mobile stations (MS) only support one of the two connection types at a time. The Gs interface allows a mobile subscriber having a class B mobile station with an ongoing data connection to disconnect the data connection and to accept an incoming speech call.

If the Gs interface is not available, the SGSN informs the mobile station that the combined mobility management has failed. The mobile station will then again perform the same procedure in the circuit-switched network. This would cause an additional signaling load on the radio interface.

2.10. Summary of mobility management

Mobility management is divided into the following functions:

• Location management: update the UE location in CS domain (location area) or in PS domain (routing area)

• Attach/detach: establish or delete an MM context for an UE in the network node (MSC/VLR or SGSN)

• Paging and searching

• Subscriber database

• Security

32/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Here, subscriber database has two options: 1) VLR in CS domain and SLR in PS domain. For temporary storage of

subscriber data is delivered from HLR and mobility data delivered by MM functions.

2) HLR contains data on subscription restrictions, services assigned to the mobile subscribers and the current subscriber status including information on the current location

In security realized guard against unauthorized service usage –

authentication and service request validation, ciphering

Figure 2.14.Mobility Management Summary structure

33/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

3. Practical guide for UMTS LA, RA and URA areas Planning and Optimization

3.1. Location and Routing Areas Planning and Optimization

LA and RA planning principles are normally very similar and close to each other, because of their definitions:

� Location Areas (LA) are used with task to inform the core network of a UE’s location in CS services

� Routing Areas (LA) are used with task to inform the core network of a UE’s location in PS services

� The Routing Area is the PS domain equivalent of the location area

� A "routing area" is normally a subdivision of a "location area"

Based on those 3GPP definitions, most of concepts for LA and RA planning are the same. Normally, planning of Routing Area can be done based on Location Area planning results. So, logically, LA planning is the main, the first planned “player”, which usually determines RA planning.

Specific planning of UMTS areas size is based on specific NW circumstances and conditions, but the initial idea that is coming from their basic fundamentals, which details are presented in chapter 2, and shortly summarized for LA and RA in table below:

Short Location Area Concepts

LA and RA are used by core network to follow subscriber’s track for CS and PS domains accordingly

One RA is to be configured inside a single LA

One RA is handled by only one CN node, i.e. one SGSN

LA and RA updates (LAU and RAU) can happen normally, like mobility result or periodically, based on timers, configured in RNC for LA, and in SGSN for RA

The smallest possible size of a LA/RA is a single cell

The biggest possible size of a LA/RA is a set of cells, connected to single VLR

34/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Changing of LA/RA size makes effect to level of signaling load

Planning the same LAC/RAC for 2G and 3G NWs is possible, but not recommended

LA/RA areas’ borders should not be located in zones with high number of updates

Some of issues, introduced above are more particularly described below, and general practical projects experience based summary recommendation is presented in the following subchapters.

1. One RA is to be configured inside a single LA

The reason of this rule is coming from 3GPP [WG RAN3]: A routing area is a subset of one, and only one, location area (LA), meaning that a RA cannot span more than one LA. Also, important, that a RA is served by only one SGSN. The following rules apply for the Routing Area Identity (RAI): - RAC is only unique when presented together with LAI. - LAI = MCC + MNC + LAC - RAI = MCC + MNC + LAC + RAC, where “MCC + MNC” means the PLMN identity.

Theoretically, that be explained by MM and GMM messages establishments processing’s (3GPP TS 24.008 V8.2.0). Practically, RAC has to be assigned only to singe RAC, but can be several RACs inside one LAC.

Previously, some mobile operators configured more than one Routing Area inside one Location Area. That was explained by SGSN limits:

In the previously used releases, one Packet Processing Unit (PAPU) was required to handle one area. Only with the Large Routing Area Support (LRAS) feature more than one PAPU unit can serve the same area. This feature makes it possible to define PAPU groups where multiple PAPUs can serve the same routing areas or share the same Gb/Iu connections. But, from the SG6.0 release onwards the LRAS feature is used in both 2G and 3G environments, so, starting from mentioned Core Release, the best RA size planning method is to define that the same as LA.

Because of that assumption, current document considers location and routing areas practical planning together, and designates those by combined abbreviation LA/RA.

It is only necessary to decrease the size of a RA area relative to a LA if there is a large quantity of paging from the PS service domain; such case is really unusual and can happen in much PS-oriented NWs.

35/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

2. Normal and periodical LA and RA updates

LA and RA updates (LAU and RAU) can happen normally, like mobility result or periodically, based on timers, configured in RNC for LA, and in SGSN for RA.

Normal updates are processing as result from UE mobility, i.e. when a UE moves across a LA or RA boundary.

Periodic location updating is used to inform the network that the mobile station is available; the network requires from the mobile station to 'report in' at certain time intervals. If there are no messages from the mobile station, the NW assumes that the MS is out of the coverage area, or that it has not been turned on, and incoming calls to the MS are not paged. This saves radio resources.

The time interval for periodic location updating is defined in the RNC set base station parameters. The time-out value is continuously broadcast, so when a mobile station enters the VLR area, it will automatically define how often it has to report to the network. When the timer of the MS reaches the configured timeout value, the MS initiates a periodic location update. Every time location updating occurs, the MS and the VLR reset their timers.

In a periodic location update, the VLR area does not change, and location updating in the HLR is not needed.

Periodic LA updates are completed with a frequency defined by the timer T3212. This timer is broadcasted within System Information Block 1 (SIB1). Periodic RA updates are completed with a frequency defined by the timer T3312. This timer is sent to the UE within the Attach Accept and Routing Area Update Accept messages. A UE will only initiate a LA update in the case it is in CS-IDLE state. The LA update can be either normal or periodic and the UE can be in either RRC Connected mode or RRC Idle mode. A UE can initiate a RA update if it is in PS-IDLE state.

Optimization of timer for periodic LA is in deep dependency with normal Location Update procedure success. Therefore, if Location Area is bigger (rather big), so Periodic Location Updates happen more often, and vice verse, as smaller LA size, so smaller need to make Periodic Update often.

The mentioned timer is configured on RNC by parameter CS_ T3212:

36/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Parameter name T3212

Abbreviated name CS_T3212

3GPP name T3212

Managed object IUCS

Parameter group SystemInfo

Category Telecom

Multiplicity 1

The timeout value, in decihours (6 min), for periodic location updating. Given only for the CS domain. (Part of

the CS domain specific NAS System information in SIB1).

The value 0 is used for an infinite timeout value, that is, periodic location updating is not used.

Description

This parameter is part of System Information Block 1.

Range 0..255

Default value 0 decihours

RNC <-> UE

RAC <-> RNC

EM <-> RNC Interfaces

RACApp <-> RAC

References 3GPP Reference: TS 24.008, Mobile radio interface layer 3 specification; Core Network Protocols, TS

25.331, RRC Protocol Specification

In the case of really small and un-loaded NW, for example in recently implemented or Trial NWs, is recommended to switch off Periodic Location Update at all, so to configure parameter CS_T3212 as 0, so keep the default parameter value.

In SGSN databuild parameters relevant to location area, routing area and cell identity planning are presented in Error! Reference source not found..

Parameter name Periodic RA update timer (PRAU)

periodicRaUpdateTimerMin Q3 Name:

periodicRaUpdateTimerSec

Description The parameter indicates the interval of the MS's periodic RA update

Enter the value in minutes and seconds (min sec). The values range: 2.62 seconds with a 2 second step 1.31 minutes with a 1 minute step 36.180 minutes with a 6 minute step

Range

Value 0 means that this parameter is not in use

37/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

MML Default 54 minutes

Interfaces

The value set for this parameter has to be smaller than the value set for the MS reachable timer (MSRT) parameter. If enter value 0 for this parameter, the value of the MSRT parameter has to be 0. Optionally, with MS reachable timer (MSRT) parameter and periodic RA update timer (PRAU) parameter values means that the parameters, which are configured with ZEJF command are not used at all if 250k PAPU mode configuration is installed. Even if it is possible to configure those values with ZEJF command, the fixed MS reachable timer and fixed periodic RA update timer parameter values are always used (marked with an asterisk *) in 250k mode.

References 3GPP Reference: 3GPP TS 24.008, Mobile radio interface Layer 3 Specification; Core network protocols; Stage 3.

Similar to LA case, in really small and un-loaded NW can recommended to switch off Periodic Routing Update at all, so to change default parameter value to 0, just important also to change MSRT parameter.

3. The biggest possible size of a LA/RA is a set of cells, connected to single VLR

That is mandatory rule, coming from general Core Areas basics, Location/Routing areas have to be inside single VLR, and, normally, they are smaller that VLR area size.

In planning stage, important to remember to configure LA/RA inside single VLR, and never involve cells from other VLRs. Actually, VLR is not reference point for LA/RA planning, the usual orienteer is RNC, many operators normally plan LA/RA the same as RNC size.

4. Changing of LA/RA size makes effect to level of signaling load

The mentioned effect is the main in LA/RA planning/optimization approach. For better understanding of the dependency mechanism, let’s come to paging messages theory. That mechanism is specified by following basics:

� NSN RAN provides an 8 kbps PCH transport channel on the S-CCPCH � 8 kbps is sufficient to include a single paging record per 10 ms � A single cell can thus page 100 UE per second

38/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

� S-CCPCH can be shared with the FACH-c and FACH-u but PCH always has priority

� High paging load has an impact upon FACH capacity when single S-CCPCH is configured

� Paging completed over either a Location Area, Routing Area, RNC or Cell � Utilisation of paging capacity is maximised when paging is completed over a Cell (UE in CELL_PCH)

� Allowing location areas to include multiple RNC reduces the effectiveness of CELL_PCH reducing the paging load

Paging is common for every BTS within the LA, but LUP (Location area

update) is local to the border cells. Moreover, there are many project specific conditions for the actual amount of LUP in the network, such as geographical LA border design, dual band strategy, etc., which can’t be easily standardized.

A location area can be figured out as the minimum unit for delivering single paging message. LUP will be switched on if LA border is crossed, so Uu interface channel resources will be used for the LUP. Considering paging effects of big LA and a small LA sizes, the main is how much paging’s load this LA is going to carry and how much LUP will be generated by this LA design.

LA size practically differs:

1) The case of “too small LA size” can have the effect of too much LA borders are configured and too much of location area updates can exist in NW. That can generate continues unnecessary high load in Core and RAN both. Therefore really small size of LA can unreasonably load Signaling and HW resorse. Location border areas can suffer of too frequent location updates procedures, LUP “ping-pong effect”. In too small LA area size, paging might be lost if the subscriber changes the LA faster than the LUP procedure is able to update the VLR/HLR with the correct LAI. 2) The case of “too big LA size” can have the effect, that paging messages numbers, sent from WBTS in single LA are too high. The worst possible effect can be that some paging messages might be lost due to paging queues congestion in the WBTS, as far as INITIAL DIRECT TRANSFER COMMAND (or REJECT) messages might be discarded.

Although both too small LA and too large LA size can cause problems, relatively large LA size is more preferable than relatively small, but recommended also to calculate the maximum paging load for a Location Area before designing the number of LACs needed in the network.

In really rough, but useful in practical planning approach:

• In the case of planned LA/RA relatively larger, they may serve larger numbers of UE and increase paging traffic level

• In the case of planned LA/RA smaller, quantity of normal LA updates increases, and increases number of signaling messages For practical usage, some assumptions can be admitted:

39/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

1) NSN assumed, that single cell can page absolute maximum of 100 UE per second. Paging messages procedure collapse close reaching the limit of 100 UE per second. However, practically observed the tangible collision will be experienced even in much smaller paging load. Paging channel has effective capacity configuration in approximately 35 of paging messages in second. So, for practical planning 35 UE per second in paging is recommended to be regarded.

2) 35 UE per second means 126000 UE per hour maximum. That is not absolute maximum, getting much is possible, but not suggested because of increasing risk of paging message lost.

3) Paging update command is included in normal mobile terminated call signaling flow. Some assumption of percent of mobile terminated calls inside total number of call attempts has to be agreed. That percent depends on NW specific and can be provided by as customer requirement, or forecasted, based on other customers’ wishes. In the lack of information, NSN initial planning figure can be 30 %. There is no planning and approximate even assumption method to calculate connected mode and SMS paging numbers. Those figures depend mostly on LA size (how often UE cross LA border) and can be really different for different NWs. In some obtained NWs monitoring (and big and small LA sizes were presented) were exceedingly smaller than mobile terminated calls. The extremely highest number of such paging is close to 10 of mobile terminated calls paging. So, assumption of paging events is 33 % of total call attempts. That sample approach does not provide exact number, but can be useful for approximate initial calculation.

4) Based on previous points info, the following formulas can be used for LAC calculation: Maximum number of Wcells in single LA = Maximum number of RNCs in single LAC *Maximum numbers of carriers

Maximum number of RNCs in single LA = Maximum Number of pagings in cell per hour / / [(Relation of Paging Messages to total number of call attempts)* Busy hour call attempts] = 126000 / (0.33* Busy hour call attempts)

Capacity of different RNC types is easily founded in Customer Documentation Libraries, and summarized in the tables below.

40/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

RNC196 48M 85M 122M 159M 196M 300M 450M 1000M

Number of subscribers

59 000 122 000 181 000 240 000 300 000 300 000 360 000 1 000 000

BH call attempts

52 000 108 000 160 000 216 000 272 000 272 000 320 000 1 000 000

Number of carriers

384 576 768 960 1 152 1 152 1 152 1 800

Number of BTS

170 256 340 420 512 512 512 600

Table 3.1 RNC 196 capacity steps

RNC450 RNC450/150 default RNC450/300 RNC450/450

Number of subscribers 181 000 284 000 360 000

Busy hour call attempts 240 000 375 000 576 000

Number of carriers 600 900 1 152

Number of BTSs 200 300 512

Table 3.2 RNC 450 capacity steps

High capacity RNC2600 Step 1 Step 2 Step 3

Number of subscribers 363 000 636 000 909 000

CS Busy hour call attempts 575 000 1 000 000 1 440 000

Number of carriers 1 440 2 100 2 800

Number of BTSs 1 440 2 100 2 800

Table 3.3 RNC 2600 capacity steps

Assumption of paging events as 33 % of total call attempts can be used in most cases, but recommended to be analyzed based on NW circumstances. Let’s sample RNC450/300, the figures are:

Maximum number of RNCs in single LA = 360000 / (0.33* 375000) ~1

So, started from RNC2600 capacity step 2 is strongly not recommended to plan more than one RA inside LA zone.

Maximum number of Wcells in single LA = 1*900=900

In that RNC capacity number of cells and WBTSs is the same in RNC. Introduced LA size calculation method is useful to avoid grow of signaling. Important note: Introduced calculation assumption does not include 24 kbps Paging Channel (RAN1202) feature’s benefits, from which offering

41/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

increased paging capacity is planned for RU20 and Paging Optimization feature in I-HSPA Rel. 3.

5. Planning the same LAC/RAC for 2G and 3G NWs is possible

That is possible to plan 2G and 3G networks using shared LAC/RAC (further in document is used LAC only thermion, as RAC, fixed with LAC assumed). Below will be summarized effects of such planning.

Positive effect Negative effect

-reducing the risk of missed paging messages -reduces number of area updates for UE, moving between systems, that possibly may reduce the signaling load

-increasing the network paging load as a result of all paging messages being broadcast on both system -making LA planning and optimization more complicated, as far as no differentiation between 2G and 3G paging and make the requirement to re coordinate cell identities

The biggest problem in planning 2G and 3G shared LA is that 2G paging load

becomes 3G loading also, and vice verse. That effect will require reducing LA size also.

In the case if shared approach is adopted then the 2G and 3G cells must be planned with different Cell Identities.

Shared strategy may be not possible also to realize in practice because they may not have the same coverage areas and not all sites being co-sited.

Summarized, shared strategy can be acceptable in small multiband NWs trials, in which LAL planning and paging load issues are not really considerable. All other cases, assigning of 3G location areas and routing areas different identities to 2G location areas and routing areas is recommended.

Problems with connections establishing while completing inter-system cell re-selection are not the mandatory reason to plan the same 2G/3G LA, just reason to analyze problems source deeply.

6. LA/RA areas’ borders should not be located in zones with high number of updates

That statement is starting point for the definition of 3G boundaries. In the planning stage, the following territories should be avoided in LA/RA

boundaries: � major roads � railways � any transportation links

42/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

� potential high traffic areas (markets, business-areas)

First three rules are coming from high number of normal updates UE mobility, last rule, - from high number of periodic updates in one small area.

In NW monitoring/tuning/optimization faze high number of LA/RA updates can be the reason to change LA/RA zones’ boundaries.

In NW acceptance procedures, NW clusters’ for borders should be defined for drive-tests not in LA/RA boundaries. That will help to verify the success of update procedures and their impact for End-User Experience.

3.2. Service Area Planning and Optimization

Service Area can be configured for number of “cells” within a single location area. That mapping is invisible for core network and handled within RAN. A cell is allowed to belong to more than one service area.

Service Areas are used for emergency calls, location based services and the Service Area Broadcast (SAB) feature performing.

In parameterization, two different are defined, SAC, used for emergency calls, and SACB, assigned for SAB services.

Obviously, that SACB is actual one only when SAB services are really in use. SAC can be assigned with the same values for different cells, but SACB has

to be identifiable for each cell. Practically, impact from SAC planning is visible only in emergency calls

performing. The main rule is to plan uncial SAC for every cell; the smartest way to simplify system design is to assign SAC and SACB the same as CI (Cell Identify).

MO Class Parameter Name

Abbreviation Multiplicity Range and

Step

WCEL Service Area

Code SAC 1 0..65535, step 1

WCEL Service Area Code for SAB

SACB 1 0..65535, step 1

The possible reason to change SAC can be two or more cells with very close

coverage, here can be the idea to configure emergency calls in one cell only. In such case, Service Area can be defined different from cell area. Assigning to the cell SAC equal to other cell’s CI means that emergency calls are switched off on that cell. However, the practical case of such modification is normally not exigible, because of small amount of emergency calls and load, generated by them.

43/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

3.3. UTRAN radio Area Planning and Optimization

Each cell has to be identified by URA identifier parameter. Correct URAId leads reducing of signaling load in RNC and UE power consumption, by URA state usage.

URAId planning include also URA group definition, those parameters are shown below in fragment of typical WCDMA Radio Network Configuration Parameter file.

MO Class

Parameter Name

Abbreviated Name

Multiplicity Range and

Step

WCEL List of URA identities

URAIdList 1

WCEL URA identity URAId 8 1..65535, step

1

Here List of URA Identifies is not one parameter, but set of parameters, -

Secondary URA Identifies. URAId itself is Primary Identifier, which is assigned to the UE in this cell when the UE is registered on this URA.

Secondary URA IDs in the URA list are URA IDs of the neighboring/overlapping URAs to which the cell belongs to, and they are broadcast on BCH of this cell.

URA planning is always limited by RNC area planning, and, same as LA/RA planning, balancing between paging and signaling load: 1) Too small URA area leads to the case, when UE has to complete cell updates

more frequently and the signaling load and also UE power consumption increases. Also, can increase the risk of ping-pong effect on URA borders.

2) Too big URA area leads the case when, paging messages have to be broadcast across an increased number of cells and the paging load increases.

In initial planning step, URAId is recommended to define equal to RNC area.

In the further optimization, based on NW analysis, URA area can be updated, based on network statistics for mobility and signaling activity.

If operator does not have risks of URA overlapping, Secondary URA

Identifiers are not even used, so in RNC damp file that parameter value is shown as “-1”.

44/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

Some other parameters are also related to URA planning topic:

MO Class

Parameter Name Abbreviation Range and

Step Default Value

RNC Cell Reselection Observing Time

CellReselectionObservingTime

1..60 min, step 1 min

16 min

RNC Max cell reselections MaxCellReselections 0..100 times, step 1 times

3 times

WCEL Maximum number of cell reselections

NCr 1..16, step 1 8

WCEL Maximum number of cell reselections in non-HCS case

NonHCSNcr 1..16, step 1 8

WCEL

Evaluating period for amount of

reselections in non-HCS

NonHCSTcrMax

Not used (0), 30 s (1), 60 s (2), 120 s (3), 180 s (4), 240 s

(5)

Not used (0)

WCEL Cell reselection hysteresis 1

Qhyst1 0..40 dB, step 2

dB 0 dB

WCEL Cell reselection hysteresis 1 for

FACH Qhyst1FACH

0..40 dB, step 1 dB

0 dB

WCEL Cell reselection

hysteresis 1 for PCH Qhyst1PCH

0..40 dB, step 1 dB

0 dB

WCEL Cell reselection hysteresis 2

Qhyst2 0..40 dB, step 2

dB 2 dB

WCEL Cell reselection hysteresis 2 for

FACH Qhyst2FACH

0..40 dB, step 1 dB

2 dB

WCEL Cell reselection

hysteresis 2 for PCH Qhyst2PCH

0..40 dB, step 1 dB

2 dB

WCEL Evaluating period for

amount of cell reselections

TCrmax

Not used (0), 30 s (1), 60 s (2), 120 s (3), 180 s (4), 240 s

(5)

60 s (2)

45/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

WCEL Cell reselection triggering time

Treselection 0..31 s, step 1 s 2 s

WCEL Cell reselection triggering time for

FACH TreselectionFACH

0..6.2 s, step 0.2 s

2 s

WCEL Cell reselection triggering time for

PCH TreselectionPCH 0..31 s, step 1 s 2 s

Normally, in planning those parameters are not needed to have values,

different from defaults. For optimization some be tuned, let’s detail treat to of them:

1) “Max cell reselections” This parameter specifies the maximum allowed

number of Cell Reselections in the CELL_FACH or CELL_PCH state before transition to the URA_PCH state. The amount of cell reselection is counted in both the CELL_FACH and CELL_PCH states. When the UE is in the CELL_FACH state, the value of the counter MaxCellReselections cannot be used as a trigger for the CELL_FACH to URA_PCH transition, but it is used when deciding a target state after the MAC-c entity has sent an inactivity indication to Layer 3.

2) “Cell Reselection Observing Time”. The timer is set when the first cell state Update message due to 'cell reselection' is received while the UE is in CELL_FACH or CELL_PCH state. In the expiry of the timer, the counter “Max cell reselections” is reset. If there are too many reselections coming from the UE within the observation time, the UE is switched to the URA_PCH state. Otherwise the Cell_PCH state is chosen. Below is an example of the target RRC state selection when value 3 is used for “Max cell reselections”:

Velocity, km/h Cell radius, km Cell Update, frequency

Target RRC, minutes state

50 10 12 CELL_PCH

50 20 24 CELL_PCH

75 10 8 URA_PCH

75 20 16 CELL_PCH

100 10 6 URA_PCH

3) “Evaluating period for amount of cell reselections”. This parameter defines the duration for evaluating the allowed amount of cell reselections. If the number of cell reselections during the time period defined by that parameter exceeds “Maximum number of cell reselections”, high mobility has been detected. The parameter is needed only if HCS is used. The parameter is used for cell selection and re-selection in both idle and connected mode.

4) “Cell reselection triggering time”. This parameter defines the way UE triggering the reselection of a new cell if the cell reselection criteria are fulfilled

46/44 Wroclaw

NWS LTE RA E2E Mgmt

SA NE

Pavel Romanenko

8/10/2010

during the time interval, assigned by parameter. The use of 2s reselection time may avoid too many cell reselections between cells and hence LA/RA updates when crossing LA/RA border. Thus, there are less signaling and less call failures at LA/RA border due to LA/RA update. The reselection time of 0s can be used in areas of high mobility, for example highways.

5) “Cell reselection triggering time for FACH”. This parameter is used for cell selection and re-selection in FACH. The UE triggers the reselection of a new cell if the cell reselection criteria are fulfilled during the parameter defined time interval. The reselection time of 2s may avoid too many cell reselections between cells and hence LA/RA updates when crossing the LA/RA border. Thus, there are less signaling and less call failures at the LA/RA border due to the LA/RA update. The reselection time of 0s can be used in areas of high mobility, such as highways.

6) “Cell reselection triggering time for PCH”. The parameter is used for cell selection and re-selection in PCH. The UE triggers the reselection of a new cell if the cell reselection criteria are fulfilled during the time interval. The reselection time of 2s may avoid too many cell reselections between cells and hence LA/RA updates when crossing the LA/RA border. Thus, there are less signaling and less call failures at the LA/RA border due to the LA/RA update. The reselection time of 0s can be used in areas of high mobility, such as highways.