Will explain “3rd generation”-->1.1 Historical
Will explain “IMT-2000 defined by ITU”-->1.2 Standardization
The UMTS Forum is an international and independent body, uniquely committed through the building of cross-industry consensus to the successful introduction and development of UMTS/IMT-2000 ’’third generation’’ mobile communications systems
www.umts-forum.org
Congestion
more than 300 million wireless subscribers worldwide -->booming market -->congestion of 2G (Japan case )-->need to increase system capacity
Limited mobility around the world
great amount of 2G systems not compatible with each other-->need for a global standardisation
Limited offer of services
more than 200 million internet users communications are not limited to speech anymore 2G are too limited to offer data services (low bit rate, circuit switching) Need for new multimedia services and applications (video telephony, e-commerce...)
Note: Alcatel will skip HSCSD!
EDGE mainly concerns the modulation scheme on the GSM timeslots. The modulation technique that GSM uses is called Gaussian Minimum Shift Keying (GMSK). With GMSK, one bit per symbol can be transmitted (21=2 phase states). EDGE will extend these boundaries by applying a new alternative modulation technique, that is 8 Phase Shift Keying. 8PSK provides for the transmission of 3 bits per symbol (23 phase states) , that is three times the transmission rate of GMSK.
In these examples, the useful rate is supposed to be :
9.6 Kbps for GSM
50 Kbps for GPRS
150 Kbps for EDGE
2 Mbps for UMTS
same examples with different rates for GPRS :
Downloading a Map: 13 s with GPRS CS-2 and 3 Time Slots (~30Kbps)
Downloading a Word Document: 135 s with GPRS CS-2 and 3 Time Slots
*A recommendation is not a specification.
IMT-2000: International Mobile Telecommunications-2000
ITU:International Telecommunication Union (www.itu.int)
Problem:
2GHz is already used by 2G systems in US : shall the frequency carriers of 2G be reframed? Isn’t EDGE the most suitable technology for 3G systems?
ITU is an international organisation composed of members of governments all over the world.
ETSI, ARIB, TIA… are regional standardization bodies composed of companies such as manufacturers and operators.
IMT-2000 is a result of the collaboration between the ITU and several regional standardization bodies, which are located mainly in Europe, in Japan and in the US
In the first phase of 3rd generation standardization, each region carried out its own standardization process to meet the IMT-2000 requirements but also to take into account its own 2nd generation mobile systems.
As similar technologies were being standardized in several regions around the world, initiatives were made to create a single forum for WCDMA standardization for a common WCDMA specification, e.g 3GPP (Third Generation Project Partnership), 3GPP2
Each Consortium has proposed one or more Radio Interfaces for IMT-2000, which have been approved for ITU. UMTS contains the two interfaces standardized by 3GPP: IMT-DS and IMT-TC.
Which radio technologies belong to UMTS?
UMTS contains the two interfaces standardized by 3GPP: IMT-DS also called UMTS FDD and IMT-TC also called UMTS TDD. UMTS core network is the evolved GSM network.
Different regions of the world will adopt different radio interface technologies according to the existing 2G system.
The connection of these different radio technologies to different core networks will require cooperation between the current standardization bodies. UMTS Release 99 does not contain these options.
ERAN: EDGE Radio Access Network
Note: CDMA in yellow is cdmaOne (IS-95)
Market share between digital systems
GSM = 48%
CDMA = 28%
TDMA = 15%
PDC = 9%
Western Europe:GSM = 100%
US & Canada:GSM = 12%CDMA = 49% TDMA = 39%
China:GSM = 87%CDMA = 13%
Japan:CDMA = 36% PDC = 64%
RoW:GSM = 41%CDMA = 35% TDMA= 24%
For information:
1999 total market (including analog systems): 41.8 B$
(US & Canada = 8.9 B$ Western Europe = 8.8 B$ China = 4.8 B$ Japan = 4.6 B$)
What about Global Roaming?
ITU leads this process of harmonizing, which is necessary for a global terminal roaming and to offer operators some degree of flexibility in selecting their 3rd generation technology.
However because of different radio technologies global roaming will continue to require specific arrangements between operators, such as multi-mode and multi-band handsets and roaming gateways between the different core networks.
We can also imagine a compatibility of SIM cards instead of multi-mode handsets (ie using a UMTS SIM card in a CDMA2000 terminal)
In fact, Global Roaming is not the issue :
The challenge is roaming and seamless services across boarders of heterogeneous private and public, fixed and mobile access networks rather than Global Roaming.
3GPP is a joint organization of standardization bodies of Europe, Japan and US
To meet new market requirements, 3GPP specifications are continually being enhanced with new features. In order to provide developers with a stable platform for implementation while at the same time allowing the addition of new features, the 3GPP uses a system of parallel "releases”: release 99, release 4, release 5, ...
R99, The first Release of the 3rd generation specifications was essentially a consolidation of the underlying GSM specifications and the development of the new radio access network. The foundations were laid for future high-speed traffic transfer in both circuit switched and packet switched modes.
R99 is based on ATM transmission technology architecture through the RAN towards CN
3GPP is large organization, which was created in 1998.
Detailed technical work is carried out in 5 Technical Specification Groups (TSG) divided into subgroups.
Many people are involved: it is estimated that more than one thousand people contribute in one way or another. This is an unprecedented number of experts working on the same project.
3GPP has delivered almost stable specifications, accepted by the majority of major industrial players, in only two years.
For information
NB : the TS 21.101 lists the existing Technical Specifications for the release R 99.
NB : the TS 21.102 lists the existing Technical Specifications for the release R 4.
NB : the TS 21.103 lists the existing Technical Specifications for the release R 5.
GPRS implementation:
TMN: November 2000: 1st European operator
Telering:January 2001
UMTS:
field trials starting from end 2001
EDGE
HSDS (High Speed Data Service) is available with Evolium™ BSS in B8 release for E-GPRS
High quality
Voice (enhancement compared to GSM)
Data (multimedia)
Convergence
Fixed and mobile networks
Data and telecommunication networks (mobile phone and computer may merge)
Services
New, personalized, ubiquitous (but yet to be invented!)
Depend on the location
countryside and big cities
high bit rate services will be offered when standing close to the base station
Depend on the terminal
different classes of terminals according to the services the user will have
To access services from everywhere in the world,
but radio interfaces should be adapted to the environment
2 Mbps small cells (due to interference level)
144 kbps large cells
Transmission in TDD is discontinuous. This implies a reduced average transmission power and leads to smaller cells for TDD (pico and micro).
What about UMTS deployment?
UMTS will be compatible with GSM networks (Handover between the two systems should be applied)
There will be UMTS islands in a sea of GSM (at least at the beginning)
What about the satellite component?
The MSS (Mobile Satellite Service) is also called Satellite Component.
It aims to fill the gap coverage, especially maritime coverage, and to provide global roaming (niche market of global roamers)
But it cannot penetrate the core of modern buildings.
It is likely to come by 2007.
It is the same well-known architecture as the 2nd generation mobile system, but
Reconfiguration of the AN, or changes in the AN domain functionality shall have minimal impact on Core Network functions, and vice-versa.
A given Access Network (e.g., the UTRAN) may provide access to different type of Core Networks via the Iu reference point and vice versa (UTRAN, BRAN, Satellite)
That’s why we speak about Iu reference point, not about Iu interface (an interface differs from a reference point in that an interface is defined where specific information is exchanged and needs to be fully recognised)
In the following material we will not speak about MSS and BRAN, only about UTRAN.
FDD (Frequency Division Duplex): use of DS-CDMA (or W-CDMA)
Frequency Bands1920/1980 MHz (UL) / 2110/2170 MHz (DL): Region 1
Channel Spacing5 MHz
Channel Raster200 kHz
Carrier chip rate3.84 Mchip/s
Radio Frame length10 ms with 15 TS
FEC codesConvolutional codes, Turbo-codes
ModulationQPSK
Bearer Capabilityup to 2 Mbps
Inter RNS synchronot needed
TDD (Time Division Duplex): use of TD-CDMA
Frequency Band1900/1920 MHz and 2010/2025 MHz (UL&DL)
Idem FDD
Inter RNS synchroneeded
The variable rates are achieved by the used of codes (& multi-slot allocation for the specific case of TDD)
FDD (Frequency Division Duplex) shall provides a continuous 3G coverage.
TDD (Time Division Duplex) mode provides specific solutions for asymmetric traffic and dedicated indoor systems, in line with the market requirements
In the following material we will focus on UMTS FDD
Note : FEC = Forward Error Correction
The FDD band is split into 6 licenses in Germany, into 4 in France.
MSS not allocated yet.
No band guards between operators and between TDD and FDD:
it may cause problems!
Need for cooperation between operators
Basic telecommunication services are divided in two broad categories:
- bearer services: provide the capability of transmission of signals between access points. They are related to lower layers.
- tele-services: provide the complete capability, including terminal equipment functions, for communication between users. They are related to higher layers.
Examples:
- Bearer services: transmission at 9,6 kbps with a max BER of 10-3. This service can not be used alone, it needs protocols of upper layers to be controlled and relayed.
-Tele-services: file transfer (the bit rate transfer depends not only on the bearer service but also on the application)
See 3GPP23.107
Whereas 2G mobile systems offer mainly speech services (the content is provided by the user), UMTS has to support a wide range of applications with different quality of services.
New Services: we can also imagine that the customer himself will be able to create its own new services (easy access ways to create services)
UMTS bearer services shall provide the necessary capabilities to support multimedia services and to enable the user of a single terminal to establish and maintain several connections simultaneously.
3GPP shall standardise service capabilities (bearer services) and not the services (teleservices) themselves.
Existing systems have largely standardised the complete sets of tele-services, applications and supplementary services which they provide. As a consequence, substantial re-engineering is often required to enable new services to be provided. In addition, the market for services is largely determined by operators and standardization. This makes it more difficult for operators to differentiate their services.
This is the reason why tele-services should not be standardized : to motivate competition between new actors of the telecommunication market, i.e content providers.
Today, it is hardly possible to predict the nature and the usage of most applications, as UMTS ought to be generic by nature to allow good support of existing applications and to ease the evolution of new applications.
The VHE is defined as a system concept for personalized service portability across network boundaries and between terminals.
The exact configuration available to the user at any instant will be dependent upon the capabilities of the USIM, terminal equipment and network currently being used, on behalf of subscription restrictions.
The VHE can be considered as a distributed user profile, owned by the service provider, distributed at any moment between the terminal equipment, the USIM, the network operator and the service provider.
A user can reasonably expect the service to be the same in any network (home and visited). In fact this is not likely to be the case:- emergency numbers change from one country to another- announcements are preferably made in the local language- value-added services, such as traffic news, are not localized, but refer back to the home area
The VHE is the framework for configuring the state of the terminal and the services accessible to it.
The Personal Service Environment describes how the user wishes to manage and interact with its communications services. The PSE is a combination of a list of subscriptions (detailing provisioned services), preferences associated with those services, terminal interface preferences and other information related to the user's experience of the system. Within the PSE the user can manage multiple subscriptions e.g. both business and personal, multiple terminal types and express location and temporal preferences. The Personal Service Environment is defined in terms of one or more User Profiles.
See 3GPP 22.121
VHE defines Service Capability Servers and standardises the features.
Services capabilities:
Service capabilities are based on functionality and mechanisms /toolkits such as provided by SAT, MExE, IN and CAMEL. These service capabilities can be made visible to the applications through an application interface.
Service Capability Servers:
GSM/GPRS/UMTS bearer services: they offer mechanisms for applications to access basic bearer capabilities.
MExE (Mobile Execution Environment) servers: Value added services are offered through a client/server relationship between the MExE server in the network and the Mobile Execution Environment (e.g. Java Virtual Machine or WAP browser) in the terminal (TS 22.057)
SAT (SIM Application Toolkit) servers: mechanisms that offer additional capabilities to the communication protocol between smart card and mobile station (TS 22.004)
CAMEL (Customised Application for Mobile networks Enhanced Logic) servers: CAMEL extends the scope of IN services provisioned to the mobile environment (TS 23.078)
Service Capability Features
Functionality offered by service capabilities that are accessible via the standardised application interface. Examples: Call Control, Location/Positioning, PLMN Information & Notifications
Bearer Services:
The service characteristics as they apply at a given reference point where the user accesses the bearer service.
Other examples of (tele)services built from service capabilities features:
Call Barring : to prevent outgoing calls to certain sets of destinations, based on the number dialled and on a wider range of parameters (time of day, day of week, location, roaming, type of call requested, cost of the service and/or destination).
Call Filtering/Forwarding: this service allows the control of whether incoming calls are accepted, forwarded or terminated
Hold: this service allows an established call to be maintained, whilst suspending use of the bearer from the incoming access point of the network. This saves on both air interface and network traffic resources when a call is temporarily suspended.
Transfer: this service allows either an established or held call to be redirected to another destination.
Call-back When Free: this service allows to be informed when the destination is next able to accept the call, allowing a new call to be originated.
See 3GPP 22.105 (Annex A)
See 3GPP TS 22.105
QoS: Quality of Service
PS and CS domains provide a specific set of bearer capabilities.
The bit rate target have been specified according to the Integrated Services Digital Network (ISDN):
- the 144 Kbps data rate provides the ISDN 2B+D channel
- the 384 Kbps provides the ISDN H0 Channel
- the 1920 Kbps provides the ISDN H12 Channel
(even though 2Mbps is generally used as the upper limit for IMT-2000 services, the exact service is specified to be 1.92 or 2.048 Mbps)
Several backward compatibility requirements influence the technology applied to 3G systems.
See 3GPP TS 22.105
Teleservices provide the full capabilities for communications by means of terminal equipment, network functions and possibly functions provided by dedicated centres.
Multimedia teleservices support the transfer of several types of information.
M-commerce :
Non-physical = electronic goods (e-banking, e-flight ticketing, ...)
Physical = electronic payment of physical goods (food, supplies, hardware, ...)
Conversational (real time user to user)
Adaptive Multi-Rate (AMR) speech service (see “Appendix” for more details):
a multi-rate speech coder is used with 8 source rates: 12.2 (GSM-EFR), 10.2, 7.95, 7.40 (IS-41), 6.70 (PDC-EFR), 5.90, 5.15 and 4.75 Kbps.The AMR bit rates are controlled by the radio access network and do not depend on the voice activity. The AMR coder is able to switch its bit rate every 20ms.
Video telephony (H324, H323, IETF multimedia architecture)
H324 (originally specified for PSTN) should be used for video in CS connections
H323 and IETF architecture (IETF SIP Session Initiation Protocol) are candidates for PS connections.
Streaming (real time user to server)
the data transfer has to be processed as a continuous stream. With streaming the browser can start displaying the data before the entire file has been transmitted These applications are typically unidirectional.
Interactive (non real time user to server with delay requirements)
Web browsing
location based services
computer games (sometimes classified as conversational class due to end-to end delay)
Background (non real time user to server with fewer delay requirements, from a few seconds to a few minutes):
e-mail delivery
Short Message Service (SMS)
Real-time services have higher priority than non-real time services.
See 3GPP 23.107
Conversational speech
Audio transfer delay requirements depends on the level of interactivity of the end users. To preclude difficulties related to the dynamics of voice communications, ITU-T Recommendation G.114 recommends the following general limits for one-way transmission time (assuming echo control already taken care of):
0 to 150 mspreferred range
150 to 400 msacceptable range (but with quality decreasing)
above 400 msunacceptable range
Interactive games
Requirements for interactive games are obviously very dependent on the specific game, but it is clear that demanding applications will require very short delays, and a value of 250 ms is proposed, consistent with demanding interactive applications.
Web-browsing
In this category we will refer to retrieving and viewing the HTML component of a Web page, other components like images, audio/video clips are related to separate QoS Classes. From the user point of view, the main performance factor is how fast a page appears after it has been requested. A value of 2-4 seconds per page is proposed, however improvement on these figures to a target figure of 0.5 seconds would be desirable.
Delay values represent one -way delay (i.e. from originating entity to terminating entity).
See 3GPP TS 22.105 Annex B
From 3GPP TS 22.115
At the moment UMTS specifies that it will provide location information to an accuracy of 50m. Different positioning methods are specified in R’99 such as:
the cell coverage-based positioning method
Observed Time Difference Of Arrival-Idle Period Down-Link (OTDOA-IPDL)
network-assisted GPS methods
3GPP TS 22.071, TS 24.030
Functionally speaking, the User Equipment (UE) is composed of the Mobile Equipment (ME) and the UMTS Subscriber Identity Module (USIM).
The Role of USIM is very similar to that of the SIM in GSM:
- it is used to store subscriber identity, subscription data, authentication and ciphering keys, authentication algorithms
- its security is improved compared to GSM with a mutual authentication between the card and the network.
The interface between ME and USIM is the Cu interface, the importance of which is crucial for compatibility: even if full multi-mode Terminals will not be developed (in a first period at least), USIM-roaming will allow the subscriber to use different IMT2000 terminals with the same card.
The UICC (UMTS integrated Circuit Card) is similar to SIM card in GSM with the same size (either ISO or plug-in).It may contain one or several USIM for different applications and also the SIM module in order to be used in a GSM terminal .Another possibility is to include additional mechanisms in the USIM part in order to provide the GSM access and be usable in a multi-mode UMTS/GSM terminal.
TS 21.111: USIM and IC card requirements
Bluetooth (See http://www.bluetooth.com)
The idea was born in 1994. Ericsson initiated a study to investigate the feasibility of a low-power, low-cost radio interface between mobile phones and their accessories. The aim was to eliminate cables between mobile phones and PC cards, headsets and desk top devices… In February 1998, 5 companies (Ericsson, Nokia, IBM, Toshiba and Intel) ventured into a Special Interest Group (SIG)
The Bluetooth system is operating in the 2.4 GHz ISM (Industrial Scientific Medicine) band. In a vast majority of countries around the world the range of this frequency band is 2400 - 2483.5 MHz. The equipment is classified into 3 power classes (class1 = 100mW=20dBm, class 2 = 2.5 mW=4dBm, class 3 = 1mW=0dBm
WAP (Wireless Application Protocol)
WAP is a technology designed to provide users of mobile terminals with rapid and efficient access to the Internet.
Today, most people access the Internet from a desktop or home PC, which has a large screen and comprehensive keyboard. The mobile phone, on the other hand, has limited display capabilities and a simple keyboard. WAP helps overcome these limitations. A special "micro browser" takes the information from the Web and pares it down so that only the key information required by the user is displayed.
Solution:
E: 1-T; 2-F; 3-T; 4-F; 5-T
F:- necessity to cope with a new access technology WCDMA
- necessity of backward compatibility towards GSM/GPRS/EDGE
- necessity to design a very efficient battery
These 3 visions of a UMTS network are developed in the 3 sections of this chapter.
CN
2 separated domains: Circuit Switched (CS) and Packet Switched (PS) which reuse the infrastructure of GSM and GPRS respectively.
UTRAN
- new radio interface: CDMA
- new transmission technology: ATM
CN independent of AN
The specificity of the access network due to mobile system should be transparent to the core network, which may potentially use any access technique.
Radio specificity of the access network is hidden to the core network.
UE radio mobility is fully controlled by UTRAN.
Some correspondences with GSM:
CNNSSUuUm
UTRANBSSIubA-bis
RNCBSCIurno equivalent
Node-BBTSIu-CSA
UEMSIu-PSGb
Principle: to capitalize on existing GSM/GPRS Core Network
MSC (Mobile-services Switching Center) / VLR (Visitor Location Register): switch (MSC) and database (VLR) that serves the user in its current location in CS domain
GMSC (Gateway MSC): switch with a gateway function with the PSTN
SGSN (Serving GPRS Support Node): similar to MSC/VLR in PS domain
GGSN (Gateway GPRS Support Node): similar to GMSC in PS domain
EIR (Equipment Identity Register): database that stores the identities of mobile terminals (can be used for example to set a list of stolen terminals)
HLR (Home Location Register): database of the user’s home system that stores the master copy of the user’s service profile.
Auc (Authentication Center): database for secret keys and algorithms used for authentication and security procedures
VHE (Virtual Home Environment): it is not a physical entity, but a system concept for Personal Service Environment (PSE) portability across network and terminal boundaries. For Release 99, e.g. CAMEL, MExE and SAT are considered the mechanisms supporting the VHE concept. (See subchapter 2.1)
An RNS (Radio Network Subsystem) contains one RNC (Radio Network Controller) and at least one Node-B.
The RNC takes a more important place in UTRAN than the BSC in the GSM BSS. Indeed RNC can perform soft HO, while in GSM there is no connection between BSCs and only hard HO can be applied.
A Node-B is also more complex than the GSM BTS, because it handles softer HO.
Controlling RNC (CRNC): a role an RNC can take with respect to a specific set of Node-Bs (ie those Node-Bs belonging to the same RNS). There is only one CRNC for any Node-B. The CRNC has the overall control of the logical resources of its Node-Bs
1°) UE makes a call
2°) Softer HO
3°) Soft HO
4°) Serving RNC (SRNC1): on UL it collects information from the Drift RNC and from its own Node-B and performs selection of the signal on a best frame quality basis. On DL it duplicates Iu-information to Drift RNC and to its own Node-B and recombination of the signal is performed by the UE. There may be only one Serving RNC per UE.
Drift RNC (DRNC2): it performs the routing of information from/to the Serving RNC. There may be up to 4 Drift RNC(s) per UE.
5°) Temporary state
6°) SRNS relocation: change of Iu interface.
Former DRNC (DRNC2) becomes SRNC (SRNC2) for that UE, former SRNC (SRNC1) no longer plays a role in the call
Hard HO (in the GSM sense) can also be performed, but it is not recommended.
However hard HO must be performed to switch between two different frequencies : it is the case between two different CDMA carriers or between UMTS and GSM.
Note:
The RNC can take different roles in the UTRAN:
- Controlling RNC (CRNC): a role an RNC can take with respect to a specific set of Node-Bs (ie those Node-Bs belonging to the same RNS). There is only one CRNC for any Node-B. The CRNC has the overall control of the logical resources of its Node-Bs.
- Serving RNC (SRNC): a role an RNC can take with respect to a specific connection between UE and UTRAN. There is only one SRNC for each UE that has a connection to UTRAN. The SRNC is in charge of the radio connection between a UE and the UTRAN. SRNC terminates the Iu for this UE.
- Drift RNC (DRNC): a role an RNC can take with respect to a specific connection between UE and UTRAN. A RNC that supports a SRNC with radio resources when the connection between UE and UTRAN needs to use cell(s) controlled by this RNC is referred as DRNC.
A manufacturer can produce only the Node-B (and not the RNC). This is not possible in GSM (A-bis is a proprietary interface)
The Iur physical connection can go through the CN using common physical links with Iu-CS and Iu-PS. However there is a direct logical connection between the 2 RNCs: the Iur information is not handled by the CN.
In the OSI (Open system Interconnection) reference model:
- NAS refers to higher layers (3 to 7). Entities of this part will exchange tele-services and bearer services.
e.g. CC (Call Control), MM (Mobility Management) and applications
- AS refers to lower layers (1 to 3). Entities of this part will exchange bearer services only.
e.g radio protocols and Iu protocols
Notes:
(1) The radio interface protocols are defined in documents TS 25.2xx and TS 25.3xx from 3GPP.
(2) The Iu interface Protocols are defined in documents TS 25.41x from 3GPP.
NAS messages between the CN (CS and PS domains) and the UE are transmitted on a transparent way through the UTRAN using Iu protocols and Radio protocols.
For CS domain:
- CM: Connection Management
- MM: Mobility Management (network mobility management)
- CS traffic
For PS domain:
- SM: Session Management
- GMM: GPRS Mobility Management (network mobility management)
- PS traffic
The radio protocols are responsible for exchanges of signalling and user data between the UE and the UTRAN over the Uu interface:
- User plane protocolsThese are the protocols implementing the actual Radio Access Bearer (RAB) service, i.e. carrying user data through the access stratum (EXAMPLES 1,2 and 4).
- Control plane protocolsThese are the protocols for controlling the radio access bearers and the connection between the UE and the network from different aspects including requesting the service (EXAMPLE 5), controlling different transmission resources, handover & streamlining etc... Also a mechanism for transparent transfer of Non Access Stratum (NAS) messages is included (EXAMPLE 3).
Some principles:
The Radio Protocols are independent of the applied transport layer technology (ATM in R99): that may be changed in the future while the Radio Protocols remain intact.
The main part of radio protocols are located in the RNC (and in the UE). The Node-B is mainly a relay between UE and RNC.
Note: Transport Network Layer and Radio Network Layer are included in the layers 1 to 3 of ISO/OSI model.
The Iu protocols are responsible for exchanges of signalling and user data between two endpoints of an Iu interface (e.g. Node-B and RNC over the Iub interface):
Radio Network Layer: contains all UTRAN-related issues.
- Application Protocol: the messages of this protocol are used for carrying signalling. The Application Protocol also triggers the establishment and release of Data Bearers (for Data Streams) directly or indirectly (via ALCAP protocol).
- Data Streams: carry user information and can also carry signalling of higher layers.
Transport Network Layer: contains the standard transport technology (ATM will be applied in Release 99). It may be changed without any UTRAN-specific changes.
- ALCAP:- is a generic name for the signalling protocols of the Transport Network
Control Plane used to establish/release Data Bearers.
- makes establishment/release of Data Bearers on request of the
Application Protocol.
- Bearers: - Application Protocol and Data Streams are carried on independent Bearers (Signalling Bearers and Data Bearers respectively)
- ALCAP messages are also conveyed on specific signalling bearers.
Note: for the difference between UMTS bearer service, RAB and Radio Bearer, see chapter 2.1 “What is a service?”
A UMTS bearer service is mapped on one or several RAB(s): each RAB can have its own QoS requirements according to the negotiated UMTS Bearer Service.
A RAB must be flexible enough to support different traffic types, activity levels, throughput rates, transfer delays and bit error rates and its QoS parameters may change during an active connection.
A RAB offers a wide range of QoS for both connection oriented packet-switched services, connectionless (store-and-forward) services, and circuit-switched traffic.
However, a RAB no longer makes a distinction between user data coming from CS-domain and PS-domain, which may have adequate QoS, but are handled by the same entities and the same protocol stacks inside the UTRAN.
In the Multi-call case, one AMR RAB from CS domain and one interactive or background RAB from PS domains can be supported simultaneously.
Radio Bearer: The service provided by the RLC layer for transfer of user data between UE and RNC
Establishment of a call Inside the CN
CS part: identical procedures as the GSM ones
PS part: identical procedures as the GPRS ones
The CN does not require resource dedicated to the UTRAN but some RAB with a given QoS.
The UTRAN has total freedom to reach the required QoS and set RABs attributes in order to allow efficient use of radio resources.
Connection to UTRAN
UE establishes a signaling link with UTRAN: RRC connection establishment
Request for service
UE sends to the RNC its “Connection Management Service Request” through RRC signaling message
RNC forwards the CM service request to the MSC with a RANAP message
Note : If the user originates one or more MO new calls in a multi-call configuration, UE sends a “CM service request” through the existing signaling connection for each new call.
Authentication and ciphering / integrity
Mutual Authentication is performed between Core Network and UE. It is transparent to the UTRAN.
Ciphering is used for confidential data transfer on air interface and Integrity for authentication of signaling messages on air interface (not used in GSM): Ciphering and Integrity keys are calculated during authentication procedure. Both Ciphering and Integrity procedures are implemented in the radio part but activated by the Core Network.
Set-up
UE indicates the bearer capability required for the call. CN translates this bearer capability into a basic service and determines whether an inter-working function is required.
RAB assignment and Radio Bearer Allocation
CN sends an Allocate channel message to UTRAN (“RAB assignment Request”) to trigger UTRAN and UE to set up a traffic channel over the radio interface (Radio Bearer(s) Setup).
Alert and connect
‘Alert’ to indicate to the calling user that the called party is being alerted.
‘Connect’ to instruct the calling equipment to connect the speech path.
See 3GPP 23.018
See http://www.cdg.org for IS-95
In CDMA field, we have experience of IS-95
IS-95 vocabulary:
forward channel=downlink
reverse channel=uplink
handoff=handover
Spectrum efficiency : transmission capacity per spectrum unit (bandwidth), i.e kbit/MHz. This must not be confused with the traffic capacity.
The spectrum efficiency in UMTS is higher than in GSM (25x200kHz carriers in GSM offering 335 kbps** while a 5 MHz UMTS carrier offers 400 kbps). If we factor in densification (frequency reuse pattern), the UMTS traffic capacity is dramatically increased. According to CDMA Development Group:
“Capacity increases by a factor of between 8 to 10 compared to an AMPS analog system and between 4 to 5 times compared to a GSM system”
** calculation details:
GSM = 25 x 13,4kbps * 7 TS / 7 (frequency reuse factor) = 335 kbps
UMTS : frequency reuse factor =1!!
Code synchronization between the transmitter and the receiver is crucial for de-spreading the wideband signal successfully.
What is the spreading factor?
It is the number of chips per bit (=chip rate/bit rate).
The chip rate is linked with the CDMA carrier bandwidth and has a constant value of 3,84 Mcps.
It is quite easy to match the bit rate of the signal with the CDMA chip rate just by choosing the adequate spreading factor.
The higher the spreading factor, the more redundancy you add in the signal and the lower the probability of bit error is by transmitting the signal.
It is also traduced by the processing gain (see below).
Code synchronization?
It is difficult to acquire and to maintain the synchronization of the locally generated code signal and the received signal.
Indeed synchronization has to be kept within a fraction of the chip time.
Note: the operation on the transparency above is N-ExOR
(00 = 1; 01 = 0 ; 10 = 0 ; 11 = 1)
What is processing gain?
After de-spreading, the amplitude of the desired signal is higher than that of any interfering signal by a factor called processing gain.
It means that the amplitude of the desired signal could even be below the thermal noise!
That is the reason why its detection is difficult without knowledge of the spreading sequence (spread spectrum systems have found their origin in military applications: the signal is hidden below the omnipresent thermal noise)
W is constant (3,84 Mcps related to CDMA carrier 5MHz): the higher the bit rate, the lower the Processing Gain.
Note that the processing gain and the spreading factor are mathematically identical, although they refer to different concepts.
We can also say that the processing gain is higher with low bit rate, because the spreading factor is higher, i.e you add more redundancy (more chips per bit).
The rainbows cells mean that the whole bandwidth (5 MHz) is reused in each cell.
In GSM there is also intra-cell interference when there are 2 (or more) TRXs in the same cell. But it is a small problem (as each TRX runs on a different frequency)
In CDMA intra-cell interference is an important problem.
Quasi-orthogonal: it is not necessary to have primary colors at the receiver to separate the user. Red and orange for example can also be distinguished.
Orthogonality between the codes is impossible to maintain after transfer over the radio interface (multi-path on DL, UEs not synchronized on UL )
CDMA is instable by nature:
one user may jam a whole cell by transmitting with too high power
need for accurate and fast power control
too many users in one cell would have the same effect
need for congestion control
A CDMA resource has 2 dimensions: the codes and the power. Obviously the power is the limiting factor ; the better we can control the power usage, the more capacity (users) we can allocate.
Spreading consists of two steps:
The channelization code (also called spreading code) transforms every data symbol into a number of chips, thus increasing the bandwidth of the signal. The narrowband signal is spread into a wideband signal with a chip rate of 3.84 Mchips/s.
The system must choose the adequate spreading factor to match the bit rate of the narrowband signal.
The spreading factor is directly linked with the length of the channelization code.
The scrambling code does not affect the signal bandwidth: it is only a chip-by-chip operation.
The scrambling code is cell-specific on the downlink and terminal-specific on the uplink.
What is a channelization code?
OVSF (Orthogonal Variable Spreading Factor)
Length: 4-256 chips according to the spreading factor
(in downlink also 512 chips is possible to match very low bit rate)
Number of codes:
The channelization codes can be defined in a code tree, which is shared by several users.
If one code is used by a physical channel, the codes of underlying branches may not be used.
The number of codes is consequently variable: the minimum is 4 codes of length 4, the maximum is 256 codes of length 256.
The channelization code (and consequently the spreading factor) may change on a frame-by-frame basis
How is Code Allocation managed?
The codes within each cell are managed by the RNC.
No need to coordinate code tree resource between different base stations or terminals.
Usually one code tree per cell. If two code trees are used, it is necessary to use the secondary scrambling code.
In fact, there are two types of scrambling codes:
Long codes:
Gold codes constructed from a position wise modulo 2 sum of 38400 chip segments of two binary sequences (generated by means of 2 generators polynomials of degree 25)
used with Rake Receiver : the PRACH is constructed from the long scrambling sequences. There are 8192 PRACH preamble scrambling codes in total, divided into 512 groups of 16 each.
Short codes:
Length : 256 chips
used with advanced multi-user detector
likely to be used later
Refer to Technical Specification 3GPP TS 25.213
“A single carrier”: in fact each operator may use several carriers of 5MHz each (2 in Germany, 3 in France)
The rake receiver can only be used with signals on the same carrier.
Rake fingers are allocated to the peaks at which significant energy arrives. Update rate: tens of ms
Each finger tracks the fast-changing phase and amplitude values due to fast fading and removes them
Rake Receiver resides in both UE and Node-B.
The numbers of fingers for a Rake Receiver is implementation dependant.
* we will see later that it is also possible to multiplex several services on the same code!
Indeed on a dedicated physical channel (which is identified by its spreading code) a user can multiplex several services as long as the total bit rate of the services does not exceed the bit rate of the physical channel.
See subchapter 5.UTRAN/ Physical Layer (Transport Channel Multiplexing)
Which codes make possible to separate the two signals at the receiver?
Scrambling codes (the two signals may have the same channelization code)
What is multipath propagation?
The signal travels from transmitter to receiver over different paths, due to reflections, diffractions or scattering. Consequently the same signal arrives at the receiver with a little delay.
The chip rate can be considered as the resolution of the CDMA system. It is linked with the 5 MHz carrier.
Multi-path propagation usually reduces the quality of the signal.
But in most cases a Rake Receiver can take advantage of multi-path to improve the quality of the signal. Indeed the dispersion is often greater than the chip duration.
Note: with IS-95 (cdmaOne), the carrier bandwidth is about 1 MHz and the chip duration is consequently longer: 1 µs (300 m). Multi-path components can not be separated in urban areas with IS-95.
Near-Far problem:
MS1 at the cell edge suffers a path loss, say 70 dB above that of MS2 which is near the base station.
MS1 and MS2 operate within the same frequency, separable at the base station by their respective spreading codes.
If MS2 is not power-controlled, it can easily over-shout MS1 and a large part of the cell.
Cocktail party effect!
We can also refer to the so called cocktail party effect to explain the necessity of power control: if someone speaks too loud, all the people in the room will be disturbed. Without PC a single overpowered mobile could block a whole cell as well.
Power control is crucial on UL (not only a enhancement feature such as in GSM), not on DL, although it is also used for different motivation (no near-far problem due to one-to-many scenario here)
Power classes of the mobile:
Class421 dBm (126 mW) – UMTS phone
Class324 dBm (251 mW) – UMTS phone
Class230 dBm (1W) – GSM/GPRS phone
Class133 dBm (2W) – GSM/GPRS phone
Basic mechanism:
PC is intended to reduce the interference level in the system by maintaining the quality if the UE-UTRAN radio link as close as possible to the minimum quality required for the type of service requested.
How is Power Control performed ?
- Open loop power control (also called slow power control):
it consists for the mobile station of making a rough estimate of path loss by means of a DL beacon signal and adding the interference level of the Node-B and a constant value.
It’s far too inaccurate and only used to provide a coarse initial power setting of the mobile station at the beginning of a connection
- Closed-loop power control:
See next slide
Inner Loop (Fast Loop Power Control)
In UL, the serving cells should estimate signal-to-interference ratio SIRest of the received uplink DPCH. The serving cells should then generate TPC commands and transmit the commands once per slot according to the following rule: if SIRest > SIRtarget then the TPC command to transmit is "0" , while if SIRest < SIRtarget then the TPC command to transmit is "1".
Upon reception of one or more TPC commands in a slot, the UE shall derive a single TPC command, TPC_cmd, for each slot, combining multiple TPC commands if more than one is received in a slot. TPC_cmd values = +1(power up), -1 (power down), 0
The step size TPC is under the control of the UTRAN (value = 1 dB or 2 dB)
UE shall adjust the transmit power of the uplink DPCCH with a step of DPCCH (in dB) which is given by DPCCH = TPC TPC_cmd.
The command rate of 1500Hz is faster than any significant change of path loss.
Outer Loop
The RNC checks the quality of the signal using for example a CRC-based approach (Cyclic Redundancy Check) and uses this result to adjust SIR target for the inner loop.
The big issue is to meet constantly the required quality: no worse and also no better, because it would be a waste of capacity.
The required quality may change with the multi-path profile (related to the environment) and with the UE speed.
The outer loop management is handled by the CRNC because a soft HO may be performed.
Frequency of the outer loop: 10-100 Hz typically
Note: in GSM only slow power control is employed (about 2 Hz)
Soft HO: UE is in the overlapping area of two adjacent sectors belonging to two different Node-Bs.
The communication between UE and UTRAN takes place concurrently via two different radio links from each Node-B separately:
- In DL: the Node-Bs make use of two separate scrambling codes. In the UE the two signals are received by means of Rake Processing, which selects the better frame between the two candidates.
- In UL: the UE transmits one signal coded by its scrambling code and this signal will be decoded by the two Node-Bs. In each Node-B the signal are routed to the RNC, which selects the better frame between the two candidates.
Softer HO: UE is in the overlapping area of two adjacent sectors of the same Node-B.
The communication between UE and UTRAN takes place concurrently via two different radio links from each sector separately:
- In DL: the Node-B make use of two separate scrambling codes (one for each sector). In the UE the two signals are received by means of Rake Processing, which selects the better frame between the two candidates.
- In UL: the UE transmits one signal coded by its scrambling code and this signal will be decoded separately by each sector of the Node-B. The 2 signals received are combined using the Maximum Ratio Combining.
IS-95: about 30% to 40% of mobile phones are in soft HO
The number of NodeBs to which UE is connected is called “Active Set”. The increase of paths in UL or DL leads to diversity gain. The gain is higher in Soft HO situation than that of the softer HO (because of macro diversity, ie, space between 2 Node Bs is more important than between 2 antennas).
The selection of new link establishment is based on CPICH measurement of the neighbouring cells reported to the SRNC.
HO decision process:
UE receives from the network all the necessary information for measurement reports: Ec/I0, Path Loss, ...
Possible causes that could be used for trigger a HO process (non-exhaustive list): Uplink quality, Uplink signal, measurements; Downlink quality, Downlink signal measurements, Distance, Change of service, Better cell, O&M intervention, Directed retry, Traffic, Pre-emption, …
Other types of HO:
Inter-frequency hard HO (from one CDMA carrier to another)
Inter-mode hard HO (from FDD to TDD)
Inter-system hard HO (e.g between UTRAN FDD and GSM)
The Soft Handover procedure is composed of a number of single functions:
Measurements; Filtering of Measurements; Reporting of Measurement results;
The Soft Handover Algorithm; Execution of Handover.
At the start of diversity handover, the reverse link dedicated physical channel
transmitted by the UE, and the forward link dedicated physical channel transmitted
by the diversity handover source Node-B will have their radio frame number and
scrambling code phase counted up continuously as usual, and they will not change at
all. Naturally, the continuity of the user information mounted on them will also be
guaranteed, and will not cause any interruption.
UMTS radio dimensioning process is very complex, especially because of the interdependence between coverage, capacity and quality.
No frequency planning required (all users use the same carrier)
Why will the cell radius decrease when bit rate increases?
The higher the bit rate, the lower the processing gain, the lower the uplink range.
The cell radius values are only indicative. It belongs to:
thermal noise density,
required Eb/N0,
NodeB noise,
NodeB sensitivity,
NodeB antenna gain,
Rx diversity gain,
service bit rate,
UE speed,...
Those value are estimated for a given BER/BLER, an expected Eb/N0, as well as a reasonable speed.
BER speech in CS : 10-3
BER data in CS : 10-6
BLER date in PS : 0,1
BLER = Block Error Rate
Note:
GSM terminal transmit power is usually 2 W (= 33 dBm).
AMR codec: see “Appendix” for more details
What are additional gains from these sources of diversity?
Each type of improvement can provide an additional gain of 1-4 dB according to the environment (a few hundred meters additional uplink range).
But there are no a priori values for any diversity gains, because each gain depends on the degree of the other diversity sources.
In the uplink the codes originate from different points: they can not be synchronised.
The number of orthogonal codes is not a hard-blocking limitation. A second scrambling code can be taken if necessary, which gives a second set of orthogonal codes (second code tree).
Downlink interference level:
The above DL interference level has to be understood as intra-cell (between the channelization codes). Multi-path propagation and lack of DL synchronization between the DCH lead to a loss of orthogonality on the receiver side.
Note:
The capacity figures above are the results of a calculation taken from “WCDMA for UMTS” (see “Related documentation”). Many assumptions have been made, e.g.:
- all terminals are equally distributed at the edge of the cell
- the amount of inter-cell interference is assumed to be lower in micro cells where streets corners isolate the cells more strictly than in macro cells.
These capacity figures give some typical values, but they are very dependant on the radio environment and may be much changed with other assumptions.
“Breathing cells” means that the coverage of the cell can increase (decrease) when the load of the cell decreases (increases): the cell can breath!
Some ways of capacity improvements:
- soft handover
- better network planning (e.g hot spots)
- more carriers (support of inter-frequency handover)
- transmit diversity
- smart antennas (a smart antenna is an array of antenna elements connected to a digital signal processor. Such a configuration dramatically enhances the capacity of a wireless link through a combination of diversity gain, array gain, and interference suppression. Increased capacity translates to higher data rates for a given number of users or more users for a given data rate per user).
Question B:
Coding rate 1/3 means:
for 1 bit of traffic data, 3 bits are sent on the air interface (because a channel coding is added to protect the data on the air interface). Applied before radio modulation.
Solution:
A: 1-F; 2-T; 3-T; 4-F
B1 :SF signal 1= 3,84.106 / [12.103 / (1/3)] = 106,67. SF=64
SF signal 2= 3,84.106 / [384.103 / (1/2)] = 5 . SF=4
B2 : PG signal 1= 10 Log10(64) = 18 dB
PG signal 2= 10 Log10(4) = 6 dB
How is the common handling of PS and CS data performed?
There is a unique UTRAN protocol stack and a common set of radio bearers for both the PS and CS domains of the core network.
There is no notion of CS and PS in UTRAN. A radio bearer can convey either CS data or PS data. Indeed a radio bearer can offer a wide range of QoS and thus be adapted to all type of data. QoS will be typically a zero delay constraint for data from CS and low bit error rate for data from PS.
A radio bearer is always applied in connected mode, but the resources are not reserved for the duration of the call.
What is the difference between logical and physical interface?
A logical interface can be conveyed on any physical interfaces (ATM, IP). Iub, Iur and Iu can even be conveyed on the same physical interface!
In the first release (R99) ATM will be used as the main transport mechanism on the logical Iu interfaces, but other technologies such as IP are planned to be used later.
A Radio Bearer is the service provided by a protocol entity (i.e. RLC protocol) for transfer of data between UE and UTRAN.
Radio bearers are the highest level of bearer services exchanged between UTRAN and UE.
Radio bearers are mapped successively on logical channels, transport channels and physical channels (Radio Physical Bearer Service on the figure)
In the control plane, signalling information (e.g NAS signalling or RRC connection establishment ) is mapped on Signalling Radio Bearers (SRB), which are mapped on Control Logical Channels.
In the user plane, user information (e.g Telephony Speech, Web browsing, SMS Cell Broadcast service) is mapped on User plane Radio Bearers, which are mapped on Traffic Logical Channels.
Logical Channels are mapped on Transport Channels (no distinction between control and user plane at this stage), which are mapped (successively) on:
- Physical Channels on the air interface
- Iub Data Bearers on the Iub interface
- Iur Data Bearers on the Iur interface (in case of soft HO)
Please note that RAB (Radio Access Bearer) are only provided in the user plane.
What is a RRC connection?
When the UE needs to exchange any information with the network, it must first establish a signalling link with the UTRAN: it is made through a procedure with the RRC protocol and it is called “RRC connection establishment”.
During this procedure the UE will send an initial access request on CCCH to establish a signalling link which will be carried on a DCCH.
A given UE can have either zero or one RRC connection.
AMR codec:see “Appendix” for more details
The logical channels are divided into:
Control channels for the transfer of control plane information
Traffic channels for the transfer of user plane information
A transport channel is defined by a Transport Format (TF) which may change every Time Transmission Interval (TTI).
The TF is made of a Transport Block Set. The Transport Block size and the number of Transport Block inside the set are dynamical parameters.
The TTI is a static parameter.
What is TTI (Transmission Time Interval)?
- it is equal to the periodicity at which a Transport Block Set is transferred by the physical layer on the radio interface
- it is always a multiple of the minimum interleaving period (e.g. 10ms, the length of one Radio Frame)
- MAC delivers one Transport Block Set to the physical layer every TTI.
Solution:
1. Table
TTI = 40 ms
The TTI is always a multiple of the length of the radio frame (=10 ms).
Transport Block Size = 576 bits
2. The system will deliver the transport blocks to the physical layer at a bit rate of
576 bits * 4 /40 ms = 57,6 kbps
This transport channel can be applied to a FAX.
3. The Transport Format Set (TFS) contains 5 TF(s) (B=0,1,2,3,4)
4. The transfer may have been reduced so as to give more capacity to a service with a higher priority (e.g real time services)
The transport channels are divided into:
Common channels: they are divided between all or a group of UEs in a cell. They require in-band identification of the UEs when addressing particular UEs.
Dedicated channels: it is reserved for a single UE only. In-band identification is not necessary, a given UE is identified by the physical channel (code and frequency in FDD mode)
BCH
>high power to reach all the user and low fixed bit rate so that all terminals can decode the data rate whatever its ability: only one Transport Format because there is no need for flexibility (fixed bit rate)
PCH
>only two transport channels can NOT carry user information: BCH and PCH.
Note: Beam-forming is also called “Inherent addressing of users”: it is the possibility of transmission to a certain part of the cell.
RACH and FACH are mainly used to carry signalling (e.g at the initial access), but they can also carry small amounts of data.
When a UE sends information on the RACH, it will receive information on FACH.
DSCH and CCPH seem to be symmetrical, but:
- DSCH is on the DL, so that different user data are synchronised with each other (the information on whether the UE should receive the DSCH or not is conveyed on the associated DCH)
- CPCH is on the UL, so that different user data can NOT be synchronised (the mobile phones are not synchronised). It may cause big problem of collisions!
DCH
> It is different from GSM where TCH carries user data (e.g speech frames) and ACCH carries higher layer signalling (e.g HO commands)
User data and signalling are therefore treated in the same way from the physical layer (although set of parameters may be different between data and signalling)
> wide range of Transport Format Set permits to be very flexible concerning the bit rate, the interleaving...
> Fast Power Control and soft HO are only applied on this transport channel.
According to the slide above and the previous one, we can say state that :
Except BCH and PCH, each type of transport channel can be used for the transfer of either control or traffic logical channels.
A. Web-browsing at very low bit rate
B. Web-browsing at very high bit rate
C. UE measurement reporting (UL)
D. Telephony speech
E. SMS Cell Broadcast service (DL)
F. Audio, Video streaming
G. Fax
H. Paging of UE when the network doesn’t know UE location (DL)
I. System Information Broadcasting (DL)
J. Initial access (RRC Connection Establishment)
The radio protocols are responsible for exchanges of signalling and user data between the UE and the UTRAN over the Uu interface
The radio protocols are layered into:
- the RRC protocol located in RNC* and UE
- the RLC protocol located in RNC* and UE
- the MAC protocol located in RNC* and UE
- the physical layer (on the air interface) located in Node-B and UE
Two additional service-dependent protocols exists in the user plane in the layer 2: PDCP and BMC.
Each layer provides services to upper layers at Service Access Points (SAP) on a peer-to-peer communication basis. The SAP are marked with circles. A service is defined by a set of service primitives.
Radio Interface Protocol Architecture is described in 3GPP 25.301.
(*except a part of protocol used for BCH which is terminated in Node-B)
RRC is a protocol which belongs to control plane.
The RRC functions are:
Call management
RRC connection establishment/release (initial access)
Radio Bearer establishment/release/reconfiguration (in the control plane and in the user plane)
Transport and Physical Channels reconfiguration
Radio mobility management
Handover (soft and hard)
Cell and URA update (see “5.UTRAN/ Mobility Management”)
Paging procedure
Measurements control (UTRAN side) and reporting (UE side)
Outer Loop Power Control
Control of radio channel ciphering and deciphering
RRC can control locally the configuration of the lower layers (RLC, MAC...) through Control SAP. These Control services are not requiring peer-to-peer communication, one or more sub-layers can be bypassed.
See 3GPP 25.331 RRC protocol (over 500 pages!)
See 3 GPP 25.323 (PDCP protocol) and 25.324 (BMC protocol)
There is no difference between RLC instances in Control and User planes. There is a single RLC connection per Radio Bearer.
RLC main functions:
RLC Connection Establishment/Release in 3 configuration modes:
- transparent data transfer (TM): without adding any protocol information
- unacknowledged data transfer (UM): without guaranteeing delivery to the peer entity (but can detect transmission errors)
- acknowledged data transfer (AM): with guaranteeing delivery to the peer entity. The AM mode provides reliable link (error detection and recovery, in-sequence delivery, duplicate detection, flow Control, ARQ mechanisms)
ARQ=Automatic Repeat Request (it manages retransmissions)
Transmission/Reception buffer
Segmentation and reassembly (to adjust the radio bearer size to the actual set of transport formats)
Mapping between Radio Bearers and Logical Channels (one to one)
Ciphering for non-transparent RLC data (if not performed in MAC), using the UEA1, Kasumi algorithm specified in R’99
Encryption is performed in accordance with TS 33.102 (radio interface), 25.413, 25.331(RRC signaling messages) and supports the settings of integrity with CN (CS-domain/PS-domain)
3GPP 25.322 RLC protocol
MAC belongs to control plane and to user plane.
MAC main functions:
Data transfer: MAC provides unacknowledged data transfer without segmentation
Multiplexing of logical channels (possible only if they require the same QoS)
Mapping between Logical Channels and Transport Channels
Selection of appropriate Transport Format for each Transport Channel depending on instantaneous source rate.
Priority handling/Scheduling according to priorities given by upper layers:
- between data flows of one UE
- between different UEs
Priority handling/Scheduling is done through Transport Format Combination (TFC) selection
Reporting of monitoring to RRC
Ciphering for RLC transparent data (if not performed in RLC)
3GPP 25.321 MAC protocol
Note: CCTrCH = Coded Composite Transport Channel
MAC can re-select another TFC (inside TFCS) every TTI.
TFC selection is based on:
- RLC buffer status
- RB attributes (e.g traffic class)
- Power indication from layer 1
By selecting TFC, MAC can:
- handle priorities between UEs (with common transport channels)
- handle priorities between data flows of one UE
- scheduling the transport blocks
TFS and TFCS are assigned to MAC by RRC.
The physical layer belongs to control plane and to user plane.
Physical layer main functions:
Multiplexing/de-multiplexing of transport channels on CCTrCH (Coded Composite Transport Channel) even if the transport channels require different QoS.
Mapping of CCTrCH on physical channels
Spreading/de-spreading and modulation/demodulation of physical channels
RF processing (3 GPP 25.10x)
Frequency and time (chip, bit, slot, frame) synchronization
Measurements and indication to higher layers (e.g. FER, SIR, interference power, transmit power, etc.)
Open loop and Inner loop power control
Macro-diversity distribution/combining and soft handover execution
3GPP 25.2xx
...
Solution:
1. See Subchapter 5.1
a UE which makes web browsing while downloading e-mails will use 2 DTCHs. A UE which makes a call uses 3 coordinated DTCHs.
2. MAC-b for broadcast transport channel (BCH)
MAC-c/sh for control/shared transport channels
MAC-d for dedicated channels
3. One MAC-d per UE (UTRAN side) and one MAC-d per UE (UE side), located in SRNC and in UE.
Note: One MAC-b per cell (UTRAN side) and per UE (UE side), located in Node-B and in UE.
One MAC-c/sh per cell (UTRAN side) and per UE (UE side), located in CRNC and in UE.
4. Dedicated Traffic logical channels can be mapped on common transport channels (Channel Switching)
5. See Subchapter 5.2
6. true (because MAC multiplexing is performed before channel coding)
7. RNTI is used for inband identification of UEs for common transport channels. For dedicated transport channels one UE is identified by the physical channel.
8. In the physical layer (see further in this chapter for more details)
9. CCTrCH (Coded Composite Transport Channel)
10. TFC selection is performed by MAC and TFC is assigned by RRC
11. yes/yes/no (TFC selection is made either in MAC-c/sh or in MAC-d: it is impossible to multiplex a common and a dedicated transport channel)
12. no (no interaction with RRC when selecting TFC inside TFCS)
13. MAC measurement reports enables RRC to detect a need for transport channel reconfiguration, e.g switch from a common to a dedicated transport channel.
The Iu protocols are responsible for exchanges of signalling and user data between two endpoints of an Iu interface (e.g. Node-B and RNC over the Iub interface)
Note: AAL2 and AAL5 are sub-layers of ATM which provide some adaptation between the application (voice, data, signalling) and the ATM layer.
NBAP
is used to carry signalling (e.g Radio Link Establishment)
Examples of actions of NBAP during Radio Link Establishment:
- signalling exchanges over Iub, which permits the RNC to reserve radio resources of Node-B for the Radio Link
- signalling transaction with ALCAP, which will setup a Iub data bearer (on AAL2) to carry the Radio Link
Frame Protocols
At this stage Data Streams (carrying RABs, NAS signalling, SMS Cell Broadcast service, RRC connection establishment…) have been mapped on transport channels
The Frame Protocols (FP) define the structures of the frame and the basic in-band control procedures for every type of transport channels.
ALCAP
is used to set up AAL2 connections for Data Streams.
Bearers
Data Streams are carried on AAL2, which enables better bandwidth efficiency for user packets but requires its own signalling (ALCAP signalling is used to set up AAL2 connections for Data Streams).
NBAP and ALCAP messages are carried on AAL5.
Note: AAL2 and AAL5 are sub-layers of ATM which provide some adaptation between the application (voice, data, signalling) and the ATM layer.
RNSAP
It is used to carry signalling (e.g Radio Link Establishment)
e.g. actions of RNSAP during Radio Link Establishment:
- signalling exchanges over Iur: the SRNC request the DRNC to reserve radio resources for the Radio Link (the DRNC will afterwards reserve these radio resources in the suitable Node-B)
- signalling transaction with ALCAP, which will setup a Iur data bearer to carry the Radio Link
Frame Protocols
At this stage Data Streams (carrying RABs, NAS signalling, SMS Cell Broadcast service, RRC connection establishment…) have been mapped on transport channels
The Frame Protocols (FP) define the structures of the frame and the basic in-band control procedures for every type of transport channels.
ALCAP
It is used to set up AAL2 connections for Data Streams.
Bearers
Data Streams are carried on AAL2, which enables better bandwidth efficiency for user packets but requires its own signalling (ALCAP signalling is used to set up AAL2 connections for Data Streams).
RNSAP and ALCAP messages are carried on AAL5.
1. What is the path of CS traffic through these protocol stacks?
2. Same question for PS traffic?
3. Same question for NAS signalling?
4. Same question for RRC signalling?
5. Which protocol is responsible for establishing AAL2 bearers? what is the path of this protocol?
6. Which protocol is responsible for the signaling exchange between RNC and Node-B? What is the path of this protocol?
7. Which protocol is responsible for the signaling exchange between SRNC and DRNC?
NAS: Non Access Stratum
“Just after switch on” process contains:
Cell selection (including cell search procedure)
PLMN selection
Attachment procedure (see “Appendix” for more details)
See “5.5 Signaling procedures” for BCCH
See “5.6 Physical Layer” for Cell search Procedure
The UE must enter the connected mode to transmit signalling or traffic data to the network
What is the relationship with the states of the mobile phone in GSM?
The two GSM states, idle mode and connected mode, are similar to idle mode and cell_DCH state in UMTS.
What is the relationship with the states of the mobile phone in GPRS?
There is no correspondence between GPRS states (idle, standby and ready) and UMTS states. Indeed there is no notion of connection on GPRS.
.
There is either zero or one RRC connection between one UE and UTRAN. If there are more than one signalling connections between the UE and the network, they all share the same RRC connection.
A RRC connection can have up to 4 SRBs (Signalling Radio Bearer), ie up to 4 DCCHs. The DCCH(s) is (are) setup during RRC connection establishment, but can be reconfigured during the time of RRC connection.
The DCCH(s) is (are) used to carry any type of signalling dedicated to this UE,
e.g radio bearer set-up (used to establish a Dedicated Traffic Channel DTCH ), radio measurement reports, location update…
See “5.5 Signaling procedures” RRC Connection establishment.
The initial state of the UE is determined by the DCCH established during RRC connection establishment:
- if the DCCH is mapped on a DCH, the UE is in cell_DCH state
- if the DCCH is mapped on RACH/FACH, the UE is in cell_FACH state
The UE can move from one state to another during the time of the RRC connection. Transitions between states are:
- based on traffic volume measurements and network load
- always triggered by UTRAN signalling
Note: in cell_DCH state, the DSCH transport channel can also be used.
URA: UTRAN Registration Area (a small set of cells)
Cell_PCH and URA_PCH states are needed for non real time services to optimise usage of codes and battery consumption. It would not be efficient to allocate permanently a DCH which would be used a very low percentage of time (Web application for example)
What is the difference between idle mode, Cell_PCH and URA_PCH states?
In idle mode the location of the UE is not known by the UTRAN, but only by the CN at a Location Area (LA) or Routing Area (RA) level (LA and RA and sets of cells larger than URA, see Subchapter “5.8 Mobility Management” for more details)
The paging message PCH must hence be sent in a LA or in a RA when the UE is in idle mode, whereas it only needs to be sent in a cell in Cell_PCH state or in an URA when the UE is in URA_PCH state (hence the paging procedure is much faster).
UE StatesCNUTRAN
UE IdentifiersUE LocationUE IdentifierUE Location
idle modeIMSI, TMSILA, RAno identifierunknown
connected mode
cell_DCHIMSI, TMSILA, RA*RNTIcell
cell_FACHIMSI, TMSILA, RA*RNTIcell
cell_PCHIMSI, TMSILA, RA*RNTIcell
URA_PCHIMSI, TMSILA, RA*RNTIURA
* Furthermore the CN knows the SRNC of the UE
Does the UE need to read all the SIBs each time a broadcast is repeated?
Dynamic (i.e frequently changing) parameters are grouped into different SIBs from the more static parameters. The UE reads regularly the SIBs containing dynamic parameters, whereas it reads the SIBs containing static parameters only if the “value tag” of the master information block has changed.
The UTRAN can also inform of the change in system information with Paging messages send on PCH or System Information Change Indication sent on FACH.
CN originated broadcast information : Location Area, Routing Area, mobile network code, mobile country code,...
RNC originated broadcast information : URA, serving and neighbouring cell scrambling codes, RACH info,...
Node B originated broadcast information : interference level,...
.
LA = Location Area
RA = Routing Area
URA = UTRAN Registration Area (group of cells smaller than LA or RA)
(see subchapter “5.8 Mobility Management”)
Note:
PCH can also be used:
- to change the UE state from cell_PCH or URA_PCH to cell_FACH
- to indicate change in the system information
UE is in idle mode:
1. CN initiates the paging of a UE over a LA (RA in PS domain) spanning, for example, two RNCs.
2. Paging of UE with Paging Type 1
LA: Location Area, RA: Routing Area (see subchapter “5.8 Mobility Management”)
A similar procedure applies to UE in cell_PCH or in URA_PCH states.
UE is in cell_FACH or in cell_DCH states:
1. CN initiates the paging of a UE to Serving RNC
2. Paging of UE with Paging Type 2 (on DCCH) using the existing RRC connection
Note:
There is either zero or one RRC connection between one UE and UTRAN.
If more than one signalling connections between UE and CN exist, they all share the same RRC connection.
A RRC connection can have up to 4 SRBs (Signalling Radio Bearer).
1. UE initiates set-up of an RRC connection
Initial UE identity: e.g TMSI
Establishment cause: e.g traffic class
2. RNC decides which transport channel to setup (RACH/FACH or DCH) and allocates RNTI (Radio Network Temporary Identity) and radio resources (e.g TFS, TFCS, scrambling codes) for this RRC connection.
3. A new radio link must be setup. This is done via a signalling procedure between RNC and Node-B which is managed by NBAP protocol (see “Procedure D” for more detail).
4. Logical, transport and physical channel configuration are sent to the UE.
5. RRC Connection Setup Complete message is sent:
- on RACH in case of RRC connection on RACH/FACH (cell_FACH state)
- on DCH in case of RRC connection on DCH (cell_DCH state)
In this example, the UE is in macro-diversity on two Node-Bs from two different RNCs. Therefore the UE could only be in cell_DCH state (soft HO is only possible on DCH)
1. The CN initiates the release of RRC connection
2. -
3. SRNC initiates release of Iu Bearer using ALCAP protocol
4. -
5. -
6. SRNC initiates release of radio link (for Node-B of SRNC) using NBAP protocol
7. SRNC requires release of radio link (for Node-B of DRNC) to DRNC using RNSAP protocol
8. DRNC initiates release of radio link (for Node-B of DRNC) using NBAP protocol
This procedure is used in many RRC procedures, e.g RRC connection establishment (Procedure C1), Radio Bearer Set-up (Procedure F1), soft HO (Procedure G)…
In this procedure:
- a radio link is set up by the RNC on the Node-B side using the NBAP protocol (a similar task is performed on the UE side using RRC protocol, see e.g. procedure C1)
- a terrestrial link (AAL2 bearer) is setup on Iub interface using ALCAP protocol
UE must be in cell_FACH or in cell_DCH states.
RAB: Radio Access Bearer
RB: Radio Bearer
Note:
The Signalling Radio Bearers (SRB) are normally set-up during the RRC Connection Establishment procedure but can also be controlled with the normal Radio Bearer procedures.
Can the UE send user information (e.g voice call) just after Radio Access Bearer establishment?
YES : At the end of this signaling procedure, a RAB has been assigned to the UE to carry user information. The RAB is mapped on the RB which has been set up. The RB is mapped on DTCH: RACH/FACH or DCH.
In this example, the UE is in macro-diversity on two Node-Bs from two different RNCs. Therefore the UE could only be in cell_DCH state (soft HO is only possible on DCH)
This example corresponds to the downlink scrambling code reconfiguration on the Node-B of the DRNC. This new scrambling code can be used, for example, with beam forming antennas.
On the diagram :
3. Physical Channel Reconfiguration Request
4. Physical Channel Reconfiguration Command
UE could only be in cell_DCH state (soft HO is only possible on DCH).
This example corresponds to a soft HO between two Node-Bs of two different RNCs.
The Active Set is the set of cells to which the UE is connected.
Note: “Just after switch on process” is described in more details in subchapter 5.4“UE identifiers”…
Location update is a procedure used for Mobility Management purposes (more details in subchapter 5.8 “Mobility management” : the UE has entered a new Location Area (LA) and uses this procedure to send the new LA to the CN.
Note: “Just after switch on process” is described in more details in subchapter 5.4”UE identifiers”…
Mobile terminated call occurs typically when a fixed telephone calls a mobile phone (UE).
Note: rate matching, 1st interleaving and 2nd interleaving have not been represented to simplify the figure.
Channel Coding aims at providing protection against transmission errors by inserting some redundancy.
After channel coding each transport block is split into radio frames of 10 ms.
The bit rate may be changed for each frame.
Each radio frame is also split into 15 time slots.
But all time slots belong to the same user (this slot structure has nothing to do with the TDMA structure in GSM).
All time slots of a same TDMA frame have the same bit rate.
Fast power control may be performed for each time slot (1500 Hz).
The number of chips for one bit M is equivalent to the spreading factor. It can easily be computed with knowledge of N:
In fact the spreading factor must be equal to 4, 8, 16…256.
Consequently it may be necessary to add some padding bits to match the adequate value of spreading factor (rate matching).
CCTrCH: Coded Composite Transport Channel
Example:
DCH 1 carries voice and DCH 2 carries E-mail (for the same user).
Transport Channel Multiplexing is performed after Channel Coding and enables these two services to be mapped onto the same physical channel.
It allows share of a spreading code and consequently share of power.
There are some restrictions with transport channel multiplexing (See 3GPP 25.212) :
- only dedicated transport channels for the same user can be multiplexed on one CCTrCH.
- one dedicated and one common channel can not be multiplex into the same CCTrCH.
- for common channel, only FACH and PCH may belong to the same CCTrCH.
- One CCTrCH can be mapped onto one or several physical channels.
- Different CCtrCH cannot be mapped on the same physical channel.
- In uplink, a maximum of one CCTrCH is allowed for one UE.
Physical channels are conveyed only up to Node-B.
A layer 1 connection consists of a single DPCCH and zero, one or several DPDCHs.
Several DPDCHs can be used to provide high bit rate.
The different fields of a DPCCH mean:
- known Pilot bits to support channel estimation for coherent detection
- TFCI (Transport Format Combination Indicator): it indicates the format combination of the transport channels mapped onto the physical channels.
- FBI (Feedback Information): will be used later for TX diversity
- TPC (Transmit Power Control): it is the command (“power up” or “power down”) in the closed loop power control.
There is one and only one uplink DPCCH on each radio link, always under a SF of 256; i.e there are 10 bits per uplink DPCCH slot.
UL DPDCH format
The Random Access Transmission of the PRACH is split into 2 parts:
-Preamble part: 256 repetitions of a 16 chips signature under scrambling code (constructed from the long scrambling code sequences). Detected by the base station and acknowledged, before the mobile can send the
-Message part (see below): contains the data, similar structure as DPDCH
There is 8192 preamble scrambling codes divided into 512 groups of 16 codes each with a 1:1 correspondence with the primary scrambling code (i.e 16 scramble RACH codes per cell).
The preamble detector is capable of processing simultaneously 16 signatures with a cell radius of 7 km
Physical channels are conveyed only up to Node-B.
Why are DPDCH and DPCCH time-multiplexed (and not transmitted simultaneously as in UL)?
Discontinuous transmission can cause audible interference to audio equipment close to the terminal (e.g hearing aids), which is a disturbance for user.
In UL the transmission is always continuous, because there is at least the DPCCH which is transmitted. The user will not be disturbed.
In DL the transmission may be discontinuous, but it is no problem (no user at the base station).
Note: The downlink DPDCH/DPCCH physical channels are called the DPCH physical channel.
A few DL DPCH formats
The P-CCPCH is time multiplexed with the SCH which is transmitted during the first 256 chips.
P-CCPCH timing is identical to that of SCH and CPICH (see 3GPP 25.211).
The P-CCPCH contains no layer 1 information.
Even if the PCCPCH is not transmitted during the 256 first chips of each slot (SCH), the scrambling code is aligned with the PCCPCH frame boundary, i.e the first complex chip of the PCCPCH frame is multiplied with chip number zero of the scrambling code.
The Secondary CCPCH, which is used to carry FACH and PCH information, is scrambled under the Primary scrambling code as well.
The CPICH, or Primary CPICH (because it is used for the cell search procedure), provide the phase and power reference for all other physical channels.
It carries 2 pre-defined symbol sequences used for DL antenna diversity (sequence 1 for antenna 1 and sequence 2 for antenna 2, both under the same channelization and the primary scrambling code.
The SCH is time-multiplexed with the P-CCPCH and consists of two sub channels:
- the Primary SCH: common to all UTRAN cells of all operators
- the Secondary SCH: contains a sequence of 15 codes which identifies the Code Group of the cell scrambling code (the 512 downlink scrambling codes are sorted in 64 Code Groups of 8 codes)
Cell Search Procedure (also called synchronisation procedure)
3GPP does not specified this procedure, but provides an informative description how it is typically done( see 3GPP 25.214 for more details):
Step 1: slot synchronization
The UE uses the Primary SCH to acquire slot synchronization to a cell.
Step 2: frame synchronization and code-group identification
The UE uses the Secondary SCH to find the frame synchronization and identify the Code Group of the found cell (64 possibilities).
Step 3: (downlink) scrambling code identification
The UE determines the (primary) scrambling code used by the found cell through symbol-by-symbol correlation over the CPICH (pilot) with all codes within the Code Group identified in the step 2 (8 possibilities).
Afterwards the P-CCPCH can be detected and the system- and cell specific BCH information can be read.
Note: If the UE has received information about which scrambling codes to search for, steps 2 and 3 above can be simplified.
No surprise: each transport channel has its own physical channel, except FACH and PCH which share the same physical channel (S-CCPCH)
Bit rate of RB =4*640 bits / 40 ms = 64 kbps
(it is assumed that there is no RLC and MAC overhead for this service)
Solution:
Radio Frame Length = 10 ms
Bit Rate (after channel coding) ~ 1971/ 10 ms = 197,1 kbps
SF ~ 3,84 Mcps/ 197,1 kbps ~ 19,5
We deduce SF=16
(We can now compute the exact value of the bit rate:
Bit Rate (after channel coding) = 3,84 Mcps/16 = 240 kbps, ie 2400 bits per radio frame
We need 2400-1971 = 429 bits of Rate matching if this service is sent without multiplexing)
Solution:
Bit rate of SRB = (148 bits - 12 bits) / 40 ms = 3,4 kbps
Solution:
Radio Frame Length = 10 ms
Bit Rate (after channel coding) ~ 129/ 10 ms = 12,9 kbps
SF ~ 3,84 Mcps/ 12,9 kbps ~ 297
We deduce SF=256
(We can now compute the exact value of the bit rate:
Bit Rate (after channel coding) = 3,84 Mcps/256 = 15 kbps, ie 150 bits per radio frame
We need 150 - 129 = 21 bits of Rate matching if this service is sent without multiplexing)
2nd interleaving:
This interlacing process (very similar to the 1st interleaving) is applied on the radio frames. Radio frame bits are originated from 1 or several transport blocks. The interlacing is taking place before the segmentation in Time Slot. The matrix has a fixed number of colons set to 30.
This example can be applied to multiplex ISDN service (carried on DTCH/DCH) and signalling (carried on DCCH/DCH).
UE dedicated functions:
Aim at maintaining the required quality of service for a minimum amount of radio resources
UTRAN dedicated functions:
Aim at managing the usage and sharing of radio resources among all users connected to the different cells
Additional info for CRNC specificity: Dynamic scheduling on S-CCPCH/DSCH whether CRNC is the serving or a drift RNC
Packet scheduler:
- located in RNC
- allocates bit rate for a bearer and may change this bit rate if required during a connection (bandwidth on demand)
- gets information from measurements reports from the Node-Bs
- part of network load control, because it can reduce the bit rates of packet bearers, if the load become too high (especially load of non controllable real-time users).
Node B handles specific functions as RACH handling, inner loop power control for the closed loop and computes the dedicated measurements
RRM Strategy:
RRM function aims at maximizing the traffic load that can be supported in each cell, while avoiding congestion or radio overhead. However the maximum acceptable traffic load per cell heavily belongs to :
Propagation conditions
MS’s speed
MS’s distribution
Traffic distribution
Maximum transmit capacity
RACH and FACH have no feedback channel: they cannot used fast closed loop power control (fast PC).
Characteristics of packet traffic:
- bursty
- tolerates longer delay than real-time services
- can be retransmitted by the radio link control (RLC) to have better bit error ratio (BER)
What actions can be taken in case of congestion?
- “power up” deny in fast loop power control
- handover to another CDMA carrier
- handover to GSM
- decrease AMR speech codec bit rates (see “Appendix” for more details)
reduce the throughput of packet data traffic by physical channel reconfiguration (DCH/DCH to DSCH/CPCH or FACH/RACH)
The mechanisms of the idle mode (cell reselection, LA/RA update) run also when UE is in connected mode.
LA: Location Area, RA: Routing Area
URA: UTRAN Registration Area (it is a set of cells smaller than LA or RA)
See “5.8 Signaling procedures” for RRC Connection establishment.
See “5.5 UE identifiers and UE states” for Just after switch process
Radio criteria refers to:
- cell available/unavailable,
- cell barred/not barred
- received level
- quality
- hysteresis (to avoid too many reselections especially when the mobile is located at the border of two cells)
The UE will use the pilot channels (CPICH) of each candidate cell to perform ranking.
Of course, each CPICH is under the scrambling code of the correspondent cell, but the UE can get knowledge of these codes by monitoring the BCH of its current cell.
LA and RA are managed on an independent way, but a RA must always be included in one LA (and not be divided into several different LAs).
LA update is performed by the NAS layer MM (Mobility Management) located in UE and in MSC. RA update is performed by NAS layer GMM (GPRS Mobility Management) located in UE and in SGSN.
In the Core Network, the location information is stored on databases:
- HLR (Home Location Register)
It stores the master copy of user’s service profile, which consists of information on allowed services, forbidden roaming areas,… and which is created when a new user subscribes to the system.
The HLR also stores the serving system (MSC/VLR and/or SGSN) where the terminal is located.
- VLR (Visitor Location Register)
It serves the terminal in its current location for CS services and holds a copy of the visiting user’s service profile.
It stores the Location Area (LA) where the terminal is located.
- SGSN (Serving GPRS Support Node)
It serves the terminal in its current location for PS services and holds a copy of the visiting user’s service profile.
It stores Routing Area (RA) where the terminal is located.
Cell size < URA size < RA size < LA size
Cell_PCH or URA_PCH?
For a fast moving UE, the URA_PCH state is better than the cell_PCH state, because the URA update frequency will not as high as cell update frequency would be.
Cell update (URA update) is performed after cell reselection (if the new cell does not contain the URA for URA_PCH)or after expiry of periodic cell update (URA update)timer.
The UE must move to the cell_FACH state, execute the cell update (URA update) procedure and reenter the cell_PCH state (URA_PCH state).
What is the difference between soft HO and hard HO?
In soft HO the mobile keeps in the same carrier and can manage two radio links simultaneously. The former radio link may be released a long time after the establishment of the radio link with the target base station.
In hard HO the mobile changes the carrier and consequently can not manage two radio links in parallel. The former radio link has to be released before the establishment of the radio link with the target base station.
Please note that there is a third type of handover called seamless handover, where the former radio link is released during the establishment of the radio link with the target base station. This is applied in systems like DECT, but not in UMTS.
Methods for compressed mode:
DL : puncturing / reduction of SF by 2 / higher layer scheduling
UL : reduction of SF by 2 / higher layer scheduling
This usually leads the MS to transmit with a higher power. In order to make sure that the gap occurs at the same time in UL and DL, it is advised to use the same method in both directions. The method is chosen by the SRNC according to the bearer type (real time or non real time), the code tree utilisation (for SF/2), then the RNC indicates the purpose of each compressed mode : TDD measurements, GSM BSIC initial identification, ...
SF/2 : the same scrambling can be used or an alternative one (with the risk of creating additional intra-cell interference by loss of orthogonality).
Problem: how can the quality be kept when performing a handover from UMTS to GSM?
see 3GPP 22.129
Solution:
1. For the initial cell selection, the UE has no information about the (downlink)scrambling code of the cell: it may be any of the 512 codes used in UTRAN and the UE must perform the entire cell search procedure to find it. For the cell reselection, the UE knows from the broadcast channel (BCH) the scrambling codes of the neighbouring cells. The cell search procedure is hence simplified.
2. Cell reselection consists of selecting a new cell according to radio criteria. Cell update is a signalling procedure used by the UE to inform the UTRAN that a new cell has been selected. Cell update is performed after cell reselection.
3. In case of a mobile terminated call, the network would have to search the UE in the whole network without LA/RA mechanisms, because it would have no information about the location of the UE.
4. A small LA would reduce the amount of signalling messages for paging in case of mobile terminated call, but would increase the amount of signalling messages for updating LA (especially if the UE is moving fast). A large LA would produce the opposite effect. Hence a trade off must be found concerning the size of the LA
5. no soft HO for RACH and FACH transport channels
6. URA_PCH is better than cell_PCH for a fast moving UE.
PLMN selection (if HPLMN not available)
The priority rules takes into account agreement between operators or radio criteria. A manual selection may also be allowed.
The UE must check periodically if the HPLMN is still not available. If the HPLMN becomes available, the UE shall move to it.
The IMSI attach is performed immediately after switch on, while the GPRS attach may be performed immediately after switch on or later as the user decides to use services of PS domain.
The IMSI attach and the GPRS attach procedures may be combined by using the optional Gs interface between the MSC/VLR and the SGSN.
After the attachment procedure, the UE is in idle mode and able to be paged. The cell selection process is going on and there may be some changes, especially if the UE is moving:
- a new cell may be selected (e.g. if there are better radio criteria)
- a new PLMN may be selected (e.g. if the current PLMN is no more available)
- LA update is performed (and also RA update, if the UE is GPRS attached)
The AMR_12_20 will provide the best quality in case of low load network or good radio propagation conditions. The other modes are better suited in case of high loading or bad radio conditions, but provides a slightly lower speech quality.
For each AMR mode, the AMR codec operates on speech frames of 20 ms and delivers 3 classes of bits sorted on their sensitivity to errors : class A (the most sensitive), class B, class C.
A stronger channel coding will be further applied for class A than for class B and C
Each class requires a transport channel (need for 3 DCH for speech)
The AMR codec can have asymmetric AMR mode adaptation (different modes on UL and DL)
In GSM, the codec is controlled by the BTS; this solution is not applicable in UMTS due to soft handover procedure. Thus the AMR mode control function should be part of the RNC (it belongs to RRM).
See 3GPP 26.071