2. Requirements from UMTS

      2.1. General

      This chapter introduces the UMTS as a future mobile system and presents the requirements on integration with a broadband network imposed by the mobile network. UMTS must serve two functions: support all those services that customers presently enjoy, and offer the potentiality to accommodate yet undefined broadband multimedia services and applications. UMTS exploits the 2 GHz radio band and is a multi-function, multi-service, multi-application digital system for multi-operator environments that will use end-of-the-century technology to support universal roaming and offer broadband multimedia services requiring up to 2Mb/s throughput. [DaS]

      At the time UMTS is in operation, ATM will be an established transmission technique. Therefore the UMTS environment must support ATM cell transmission to the user. This will enable service providers to offer broadband services to the users regardless whether they are mobile or fixed subscribers. The compatibility between UMTS and the fixed network functionality and the multi-service capability at the radio interface must be considered carefully. UMTS concept merges paging, cordless phone, mobile terrestrial and mobile satellite standards into a single system. The services supplied will be comparable in quality and integrity to those provided by the fixed B-ISDN. [Cos]

      2.2. Generic Integration Aspects

      An integration between mobile and fixed communication systems will make use of the functional commonality. Nevertheless some mobile network specific functions like handover, macro diversity, location update and paging need to be introduced to the mobile users but are useless for fixed users. By integrating mobile and fixed communication into one network, duplication of common functionality can be avoided, which leads to a transparency between mobile and fixed services as well as to a reduction of the costs. An integrated UMTS/B-ISDN system will cater for the requirements of both fixed and mobile users and will differ from one designed for either fixed or mobile access only. Overhead introduced to the B-ISDN by UMTS functionality should be kept minimal. The key integration aspects are:

      • Possibility to add mobility to networks in a modular way.
      • Minimum impact on the backbone network.
      • Exploit the capabilities of UMTS underlying networks to the full.

      MONET identified four areas of integration:

      • Service integration. The services offered to the user should be as much as possible the same for mobile and fixed communication.
      • Infrastructure integration. Transport and switching infrastructure between UMTS and B-ISDN communication should be shared.
      • Functional integration. Functional entities in the fixed network should be shared with UMTS to a maximal extent.
      • Protocol integration. Fixed protocols should be selected to support UMTS information transfer and signalling at the interfaces. [SSD]

      These four areas of integration will be studied in the following chapters and integration solutions to these areas will be presented.

      2.3. UMTS Services

      Though UMTS and B-ISDN can have common services, UMTS will not be able to provide the full set of B-ISDN services. The limited bandwidth on the radio interface will restrict the UMTS services to a data rate of 2 Mbits/s if not even lower. With a common service architecture the bandwidth limitations of the radio link can be understood as a restriction caused by the terminal capabilities.

      The UMTS Service Provision Model is the basis for the design of an open system. It is a layered model consisting of four different objects and three interfaces between them. These objects are:

      • User Object represents the person involved as an end user.
      • Application Object makes use of the services offered by UMTS and enhances these or constructs end-user application upon them.
      • Service Object is a collection of functions to enhance the network services and to offer the UMTS services to users and applications.
      • Network Object provides the network services built on the infrastructure.

      The model is depicted in figure 3 with interfaces between the objects.

      Figure 3. UMTS Service Provision Model. [MON101]

      The UMTS Service Components are defined at the Application Service Interface (ASI, see figure 3 above). They represent service-user information to the system below the ASI as well as to the applications above ASI. The UMTS Service Components are:

      • Audio provides the presentation capabilities for speech and other audio information.
      • Video service components are defined with qualities corresponding to general standards.
      • Data provides the presentation e.g. text, pictures, vector graphics and program executables or object code.

      The Network Services in turn are defined at the Service Network Interface (SNI, see figure 3 above) and they provide information transfer capabilities between two Network Service Access Points. They are defined along the specific information transfer they are serving. The Network Services are:

      • Constant Bit Rate Isochronous Transfer.
      • Variable Bit Rate Isochronous Transfer.
      • Connection-Oriented Packet Transfer.
      • Connectionless Packet Transfer.
      • User Service Management.

      The Network Services require that B-ISDN/ATM support them at the transport level. This will be studied later with the exclusion of User Service Management since management aspects are out of the scope of this study.

      The bandwidth limitations require bandwidth efficient source coding. For calls between fixed and mobile terminals transcoding of bearer components will be needed. Even if the coding of the bearer components is different, the services and service architecture can still be the same. If ATM is used in the radio access, the content of the ATM cells in the Core Network will be different from those in the Access Network for the same bearer connection. This results in service requirement on B-ISDN:

      Transcoding should be adjusted to the ATM transport mechanism, and possibly integrated with the B-ISDN coding operations. [SSD]

      2.4. UMTS Architectures

      The UMTS architectures are defined using a top down approach. The functional architecture represents the most generic UMTS configuration and it is independent of implementation considerations. This gives network operators and equipment manufacturers the freedom to design their own better suiting network architectures and equipment taking into consideration service provision, fixed network integration, minimisation of the number of interfaces, multivendor implementations and performance. Network architecture and entities are depicted in figure 4 which represents the most generic set of functional entities and functional interfaces that can exist in a full network implementation. One or more entities can be implemented in a single physical entity (e.g. MSCP and MSDP).

      Figure 4. UMTS network architecture.

      Within the UMTS network architecture framework, there is a vertical division between the Access Network, which is UMTS specific, and the Core Network, which is common with the fixed communication. There is also a horizontal division between basic transport and switching functions, and mobility and service control functions. Data level is separated in this figure 4 but most likely it will be physically merged to mobility and service control level.

      Mobile Terminal (MT) makes an access to a communication network available. Base Transceiver Station (BTS) sets up, maintains, and releases a radio link in co-operation with MT. Cell Site Switch (CSS) is a basic switch in the Access Network. Local Exchange (LE) switches both fixed and mobile connections. It supports the UMTS services but is not UMTS specific switch like the CSS. [SSD]

      The Mobility and Service Control Point (MSCP) includes the functionalities needed to control mobility procedures and service operation within a certain area. It can access, modify and delete information at the Mobility Service Data Point (MSDP). The MSDP can be considered as a node in the UMTS Distributed Data Base (DDB). It stores information concerning terminal, subscriber and location data. Functional entities are separated by functional interfaces. [Cos]

      The UMTS network is composed of four sub-networks (see figure 4).

      • Access Network provides the basic transmission and local switching functions needed for the mobile terminal to access the fixed network resources over the radio interface. It contains also interfaces toward the Core Network and the Service and Mobility Control Network. Access Network comprises Mobile Terminal on the one side of radio interface and Radio Access System on the other side. [Cos] [Bui]
      • Core Network supports UMTS with fixed network switching and network resources. It consists of B-ISDN local and transit level switches and the transmission systems. [Cos] [Bui]
      • Service and Mobility Control Network contains e.g. mobility related higher level functions like handover decision functions and it stores the user related data to support access to the network and mobility. The Service and Mobility Control Network consists of MSCPs and MSDPs both on the Core Network and the Access Network side.
      • Telecommunication management network (not depicted in figure 4.) will be probably based on IN concept and is left out of study in this master's Thesis.

      The UMTS network can also be divided into Backbone and Mobile Network. The Backbone Network concept comprises both Core Network and core side part of the Service and Mobility Control Network. Mobile Network comprises all the UMTS specific elements.

      Some infrastructure originated requirements are to be seen on B-ISDN. These are e.g.:

      • Network planning and management should permit sufficient transport and switching capacity to UMTS, and optimise the use of the infrastructure respect to fixed and mobile communication.
      • Integrated design transport: The use of ATM should not set restrictions on the transport mode and the capacity of the air interface.
      • Handover and marco diversity support: The B-ISDN switching and transport infrastructure has to support terminal mobility concerning interconnections in networks and capacity. [SSD]

      2.5. UMTS Mobility Functions

      The UMTS mobility function concepts are understood in this context as a transport function which performs UMTS specific operations due to terminal mobility. These are e.g. location management, paging, security functions, handover and macro diversity to mention a few. In this study it is not possible to take all of these functions into consideration but only those having strongest impact on B-ISDN, like handover and macro diversity. These functions are taken into further study since they also depend on the chosen radio access techniques.

      2.5.1 Handover

      Mobile Terminals (MT) roam to the radio access environment and gain access to the UMTS network through BTSs. To maintain a call while moving, a mobile terminal must handover a call to a new BTS and respectively the network must route the connection to that BTS.

      Handover Types

      Handover must maintain the call at a reasonable quality which depends on the Quality of Service (QoS) agreed during a call set-up. Handovers might be avoided or delayed if a quality reduce during the handover might exceed a preset limit. If an MT disappears abruptly from the old RASs (Radio Access Systems) coverage area, the data in transit and buffered along the old path, depending on the used handover type, needs to be recovered from old buffers or retransmitted. Handover types are:

      • Seamless handover is unnoticeable at the user service level. This means that higher protocol layers have to maintain the QoS for the connection. Soft handover and hard handover can be seamless.
      • Soft handover is a technique in which MT is contacted to several BTSs at the same time. In this way macro diversity is achieved which means that the probability of losing data is smaller than in hard handover
      • Hard handover is a technique in which connections are just switched over from one path to another without taking care of data lost on the old path. If a transmission is discrete (i.e. delays between packet data) and switching can be made fast enough, seamless handover is possible here.

      Interleaving in Handover

      The problem of handover in relation to interleaving is that after a handover an interleaving matrix will be left in the old BTS. The options for handover in relation to interleaving are:

      • 1) Carry out a handover to the new BTS and declare the old interleaving matrix lost. The mobile terminal restarts the interleaving process with the new BTS.
      • 2) Stop interleaving on the connection between the MT and old BTS just before a handover. The connection is switched over without an interleaving process. After the handover an interleaving process is started again.
      • 3) If the de-interleaving point is above the handover point in the network, a connection can be switched over to a new path without impact to an interleaving matrix.
      • 4) Provide a make-before-break connection on the radio interface and in the fixed network so that previously sent service frames continue to be transmitted to and from the old BTS whilst new service frames are transmitted to and from the new BTS.

      In CODIT, the block interleaving is used which means that handover between these blocks eliminates the impact of interleaving on the handover process, i.e. all interleaved data is sent to the same BTS where it can be deinterleaved without any problem. [SSD]

      2.5.2. Macro Diversity

      The goal of macro diversity is to enhance radio performance in both upward and downward direction by transmitting user and signalling data streams over several separate paths to and from the MT. Macro diversity is essential for CDMA/CODIT performance but is not of great importance to ATDMA.

      Macro Diversity Combining

      In macro diversity a connection is conveyed over a number of distinct physical paths. For the down link direction the information is multicast onto the different paths at a multicasting point in the network and combined in the MT. In the up link direction multicasting is performed in the MT, and combining in a certain point in the network. The combining function plays with stringent timing and synchronisation restrictions.

      Combining at the Physical Level (Down Link Only)

      In the down link direction the BTSs in the macro diversity set could transmit the same RF signal. MT just receives one radio signal similar to situation of receiving multipath propagated signal from one BTS. This technique is used in CDMA/CODIT where BTS transmissions are synchronised within 0.1 ms and differences are aligned by the RAKE receiver in the MT. Each BTS monitors MT transmissions to align its transmissions as seen at the MT.

      Combining at the Bearer Level (Mainly Up Link)

      In the up link direction, combining is most useful at the bearer level or above. Each frame or cell carries a quality parameter of the radio link which predicts an error probability for that data. These quality indications can be used to weight the data streams in the combining process. Synchronisation information needs to be included in the information transported from the BTSs to the combining point to allow data streams from different BTSs to be aligned there. Combining at the bearer level in the upstream direction is used in CODIT. For the downstream direction the bearer level is less suitable for combining than the physical level. [SSD]

      UMTS Functions in General

      For the functional integration it would be beneficial to allocate the UMTS functions, which have the most influence to the implementation of ATM functionalities in the RAS, to the two pairs of categories, namely

      • radio dependent/radio independent functions
      • bearer related/bearer unrelated functions

      This would help to evaluate the implementation of the ATM transport technique to RAS intraconnections and control functionalities. The mapping gives possibility to consider which functions could or should be carried out in UMTS and which ones in the B-ISDN network. Those functions that are bearer related can be implemented to the B-ISDN network and carried out there. Bearer unrelated could be located to UMTS but some of them could as well be placed to the B-ISDN network. They do not have much impact to the transport technique (e.g. ATM) but all the more to the B-ISDN control plane.

      Radio dependent functions, as they are understood here, are functions customised for each radio interface separately (TDMA/CDMA). This division separates the Core UMTS Network functions (radio independent) from the Access Network specific functions. Depending on the level of integration some Core Network functions could be carried out by B-ISDN, for instance in a Mobility Server etc. Radio dependent functions have to be located in each radio access system independent of the Core Network.

      This consideration should be done for every integration scenario separately because they differ from each other by the level of integration and exploitation of B-ISDN. This might be of use when determining the practicable level of integration. [MON110]

      The goal of integrating Call Control (CC) and Bearer Control (BC) functions is to be able to use the same exchanges for UMTS and B-ISDN. Not only the hardware equipment should be reusable, but also software that implements the CC and BC functions. This results in functional requirements on B-ISDN:

      • The separation of call and connection control avoids bandwidth wastage in the radio interface.
      • The broadband Call Control process should be able to identify UMTS destined calls for adequate treatment of mobile calls.
      • Local view of the B-ISDN LE should be maintained to be able to pass a call to another exchange while still retaining an acceptable QoS and without the need to release the call in an end-to-end fashion. The local view contains relevant bearer and call status data, charging information, and so on. [SSD]

      2.6. Radio Access

      When integrating UMTS and B-ISDN, radio access characteristics and performance differences such as bit rates, transport formats and error protection mechanisms have a strong impact on B-ISDN.

      The UMTS radio spectrum is a limited resource and service specific coding is used for information transfer over the radio interface. It is used to minimise the bit rate over the radio interface whilst the bit rate is increased by error correction coding. The radio interfaces Advanced Time Division Multiple Access System (ATDMA) and Code Division Multiple Access System (CDMA) from the respective ATDMA and CODIT projects have been used as example access systems for transport interworking. [SSD]

      Radio Interface Characteristics

      Each user has one or more time slots within a radio frame in ATDMA. CDMA/CODIT uses a spreading code. Since the MTs with a poor signal propagation can not increase the power, it is essential that MTs can operate in macro diversity mode.

      The bit error rate over the radio interface for data was targeted around 10-6 for both ATDMA and CDMA but it is not sure this will be achieved in a real UMTS radio environment. For voice the radio interface error rate is 10-3. Error rates on the radio interface will be constantly changing with changing radio environments. As the MT moves interference may occur from other radio sources and there may be instances when the MT moves into a blind spot and the connection is lost.[SSD]

      2.6.1. ATDMA Radio Access

      Typical user bit rates for ATDMA are 9.6 kbit/s for data and 8 kbit/s for voice. Frequency hopping, interleaving and ARQ (Automatic Repeat Request) can be used to reduce bit errors. Radio interface delays can be in the range of 30 ms to 300 ms.

      UMTS must support a wide range of services from speech to high bit rate data, many of which will be asymmetric, i.e. based on different bit rates for up link and down link. The radio access system needs only offer a set of independent uni-directional radio connection elements i.e. bearers [Uri]. Four different bearers are provided for speech, low delay data, high delay data and unconstrained delay data. The first three use Forward Error Correction (FEC) plus interleaving whilst unconstrained delay includes ARQ. ATDMA services are shown in table 1.

      Table 1. The ATDMA services with corresponding radio access characteristics and performances. [SSD]

      Service       Service   Slots/each  Interleave Link          
              bit       5 ms frame   depth     performance   
              rates                                          

Voice         6.4       1 or 2      Not less   Delay <30 ms  
              kbit/s                than       BER <10-3     
              to                    4 frames                 
              9.6                                            
              kbit/s                                         

Low delay     9.6       Not less    Not less   Delay <30 ms  
data          kbit/s    than 1.5    than       BER <10-6     
              to        for         4 frames                 
              2 Mbit/s  9.6 kbit/s                           

High delay    9.6       Not less    Not less   Delay <300    
data          kbit/s    than 1.5    than       ms            
              to        for         36 frames  BER <10-6     
              2 Mbit/s  9.6 kbit/s                           

Unconstrained 8 to 53   Not less    -          Packet loss   
 delay data   byte      than 1 (10             10-6          
                        slots for                            
                        53 byte)

      A speech codec uses speech characteristics to map voice from 64 kbit/s in the fixed network to a net bit rate in the range 6.4 kbit/s to 9.6 kbit/s. If radio interface error rates increase, then the codec reduces its bit rate so extra bits on each radio frame can be used for error protection. Alternatively additional time slots can be allocated to the connection. The delay of 30 ms is obtained from each speech frame capture, transmission over 4 radio frames and processing. If these blocks are not received correctly, then a Negative-Acknowledge is sent to the transmitter which then returns a new message containing additional parity-check bits to correct these errors.

      High delay data service offers the best service quality due to a long interleaving depth. Both low delay and high delay data services have been designed for constant bit rate services. They allocate slots on the radio interface in such a way that blocking can not occur during a call. The number of slots allocated for low and high delay data services increases with any increase in error rate over the radio interface. Retransmission is used for high delay data.

      For unconstrained delay data service, a 53 byte block is dedicated to the transport of an ATM cell. One ATM cell will be transmitted over 10 or 12 radio frames. Short blocks are used for bi-directional signalling channels that can tolerate variable throughput. FEC and ARQ techniques are used for both cases. [SSD]

      2.6.2 CDMA Radio Access

      For a CDMA radio interface all MTs transmit in the same frequency band whilst all BTSs transmit in another frequency band. The CDMA/CODIT services are shown in table 2.

      Table 2. The CODIT services with corresponding radio access characteristics and performances. [SSD] 

      Services  Service     Bits in    Interleave Link         Service      
          bit rates   radio       depth     performance  performance  
                      frame                                           

Voice     0.4 kbit/s  8 bit to   1          Delay <30                 
          to          320 bit               ms                        
          16 kbit/s                                                   

Signallin 9.6 kbit/s  248 bit    2          Delay 40 ms               
g                                                                     

          9.6         248 bit    4                                    
Data      kbit/s,     1616 bit   6          Delay 120    BER 10-6     
          64 kbit/s,  3232 bit   6          ms                        
          128 kbit/s

      Speech (from a 0.4 kbit/s to 16 kbit/s codec) is split into 10 ms frames to give between 4 to 160 bits per frame. A bit rate of 8 to 320 bits is transmitted in each 10 ms frame at the physical level including error protection. Voice is block interleaved within each 10 ms radio frame.

      For signalling at 9.6 kbit/s a 20 ms service frame is split and interleaved over two 10 ms radio frames 96 bits each. This is increased from 96 to 248 bits for error protection.

      For data at 9.6 kbit/s an 80 ms service frame is split into two 40 ms frames. Each 40 ms frame is split into 496 bits and interleaved over four 10 ms radio frames.

      For data at 64 kbit/s and 128 kbit/s a 60 ms service frame is split and interleaved over six 10 ms radio frames. At the physical level the 640 bits and the 1280 bits in each 10 ms frame are mapped into 1616 bits and 3232 bits to include error protection. [SSD]

      2.7. UMTS Signalling

      A part of an ideal integration scenario is the complete protocol integration. The minimum integration level, the reuse of B-ISDN functions and protocols, consider B-ISDN functions and protocols as a subset of the UMTS protocols. The UMTS signalling is separated to access and network signalling. In the B-ISDN protocol architecture there is a strict separation between signalling at the UserNetwork Interface (UNI) and signalling within the network (Network-Node Interface, NNI). Signalling at the UNI is called Access signalling, while signalling within the network is called Network signalling.

      UMTS can be understood to consist of a number of Access Networks connected to a global Core Network. The Access Network is the part of UMTS where Access signalling is used. Network signalling is used in the Core Network. The Access Networks are connected to the Core Network using UNI/NNI boundary functionality in the UMTS/B-ISDN LE and the MSCP(Core). This boundary functionality is re-used for UMTS. This signalling separation is illustrated in figure 5. [SSD]

      Figure 5. UMTS Core and Access Networks. [MON65]

      Figure 6 shows the network architecture of UMTS. Between LE and MSCP(Core) there is an NNI for signalling, but there is an UNI as well, for signalling between MT or RAS and MSCP. The signalling network will be based on the Signalling System 7 (SS7). The NNI is defined as the interface between all entities in the signalling network. Therefore, network signalling is simply signalling at the NNI.

      Figure 6. UMTS Network Architecture. [SSD]

      The major difference between access and network signalling is the logical structure of the signalling relation. In the direction terminal-to-network, a terminal application is interested in signalling services offered by the peer application process. From the terminal point of view any node in the network is acceptable. The terminal needs only indicate the required service by so called implicit addressing. The destination entity is addressed only by its type, not by a specific address. The selection of the node is done within the fixed network. For network signalling NNI-entities always need to explicitly address the destination of a signalling message. An MSCP that wants to send a handover message to the new MSCP in an inter-MSCP handover has to know which specific MSCP address should be used. [SSD]

      UMTS Network Signalling is left at this level of study and is taken up in the following chapter 6. in the context of signalling protocol interworking. [SSD]

      UMTS Access Signalling

      UMTS Access Network protocol stacks are all based on a general model, shown in figure 7. The model is based on the concepts of the OSI Reference Model but uses only four layers.

      Figure 7. Basic protocol model of the UMTS Access Network. [MON65]

      Introducing the Application Layer allows several protocols to be used in the Access Network without modifications to the existing protocols. This is important as UMTS is expected to use Intelligent Network Application Protocol (INAP) and possibly other application protocols in the Access Network parallel to the Call and Bearer Control protocol (Q.2931).

      Two protocol stacks are found in the most UMTS implementations: the stack at the UMTS air interface and the stack at the B-ISDN UNI. A number of other stacks can exist but they depend on the architecture and topology of the Access Network and they are out of this study.

      A simplified Access Network architecture is shown in figure 8 illustrating Physical Entities (PEs), Physical Interfaces (PIs) and Functional Interfaces (FIs) as defined by MONET. FI17 is not a user Access Network interface, but a Signalling Network interface and therefore is considered later in the chapter 6.

      Figure 8. Simplified Functional Architecture of the UMTS Access Network. [MON73]

      Functional Interfaces can span a number of Physical Interfaces and several Functional Interfaces can be implemented over the same Physical Interface. A protocol stack of a PI shows how FIs are separated. Three example protocol stacks are defined here for PI1, PI3 and PI6 corresponding to the UMTS air interface, the interface between the UMTS RAS and the B-ISDN LE, and the interface between the B-ISDN LE and the MSCP. The Protocol stacks are based on the following assumptions:

      • The RAS is seen as a single entity. Interfaces and protocols concerning RAS intraconnections are studied later in chapter 6.
      • The RAS is connected to a B-ISDN LE using a standard BUNI interface (see chapter 3.).
      • Interworking functions between UMTS and B-ISDN are assumed to be located in the RAS.

      The protocol stack for PI1, the UMTS air interface, is shown in figure 9. It contains all the FIs terminating in the MT and the corresponding protocols including also the user plane.

      Figure 9. Protocol Stack for PI1, the radio interface. [MON65]

      The Protocol Stack for PI1 shows that all FIs and the user plane use the same Physical Layer and Data Link Layer. Radio Interface Protocols that require fast interaction between MT and RAS can directly use the lower layer protocols. For all other signalling, the Network Layer is used. The Network Layer provides the routing of the Application Layer Messages. The UMTS Adaptation Layer (UAL) corresponds to the ATM Adaptation Layer (AAL) at the B-UNI. The interworking functions are performed at this boundary.

      Figure 10 shows the Protocol Stack for PI3, the interface between the RAS and the B-ISDN LE. The B-UNI is a subset of the overall stack.

      Figure 10. Protocol Stack for PI3, the interface between RAS and LE. [MON65]

      The User Plane and the FIs to the MSCP and the FI to the LE are separated using different ATM Virtual Channels (VCs). VCs will be presented later in chapter 3. as well as Q.2931* and the Signalling AAL (SAAL). The FIs to the MSCP make use of the Network Layer to distinguish between MT-MSCP signalling and RAS-MSCP signalling.

      Figure 11 shows the Protocol Stack for PI6, the interface between the LE and MSCP. The stack shows both the Access Network and the Core Network interfaces between LE and MSCP (see figure 6). Signalling at the FI17, between LE and MSCP, uses the BISDN SS7 network while the other two FIs use the UMTS Access Network stack over ATM.

      Figure 11. Protocol Stack for PI6, the interface between LE and MSCP. [MON65]

      Some protocol requirements on B-ISDN have been recognised:

      • Address mechanisms at the B-UNI: In order to correctly route signalling messages to the right CSS, BTS, MT, and user.
      • B-ISDN UNI should support non-call related signalling (e.g. paging, session handling, location updating) as a virtual circuit routing capability.
      • Different signalling CC and BC: The required separation of call and connection should result in different CC and BC messages in the signalling protocols.
      • Signalling Network Layer at the B-UNI must support UMTS signalling requirements at the B-UNI, the support of a Signalling Network Layer should be provided by the B-ISDN LE. This may be accomplished by means of (1) permanent/semi-permanent virtual circuits allocated by means of management procedures, and (2) point-to-point signalling circuit allocated by means of meta-signalling procedures.
      • Q.2931 extension: For the support of the UMTS signalling requirements at the B-UNI, the capabilities provided by Q.2931 should be extended in order to allow the exchange of signalling messages between entities of the Access Network and the Core Network. [SSD]