Universal Mobile Telecommunication System

UNIVERSAL MOBILE TELECOMMUNICATION SYSTEM

The Universal Mobile Telecommunications System (UMTS) is a third generation mobile cellular system for networks based on the GSM standard. Developed and maintained by the 3GPP (3rd Generation Partnership Project), UMTS is a component of the Standard International Union all IMT-2000 telecommunications and compares it with the standard set for CDMA2000 networks based on competition cdma One technology. UMTS uses wideband code division multiple access (W-CDMA) radio access technology to provide greater spectral efficiency and bandwidth mobile network operators.

Network Evolution

An Evolution that Makes Sense

HSUPA : High Speed Uplink Packet Access

HSDPA : High speed downlink packet access

The main idea behind 3G is to prepare a universal infrastructure able to carry existing and also future services. The infrastructure should be so designed that technology changes and evolution can be adapted to the network without causing uncertainties to the existing services using the existing network structure.

WCDMA Technology

The first Multiple Access Third Generation Partnership Project (3GPP) Wideband Code Division networks (WCDMA) were launched in 2002. At the end of 2005, there were 100 WCDMA networks open and a total of more than 150 operators with licenses for frequencies WCDMA operation. Currently, WCDMA networks are deployed in UMTS band of around 2 GHz in Europe and Asia, including Japan and America Korea.

WCDMA is deployed in the 850 and 1900 of the existing frequency allocations and the new 3G band 1700/2100 should be available in the near future. 3GPP has defined WCDMA operation for several additional bands, which are expected to be commissioned in the coming years. As WCDMA mobile penetration increases, it allows WCDMA networks to carry a greater share of voice and data traffic.

WCDMA technology provides some advantages for the operator in that it allows the data, but also improves the voice of base. Voice capacity offered is very high due to interference control mechanisms, including frequency reuse of 1, fast power control, and soft handover. WCDMA can offer a lot more voice minutes to customers. Meanwhile WCDMA can also improve broadband voice service with AMR codec, which clearly provides better voice quality than fixed telephone landline. In short, WCDMA can offer more voice minutes with better quality.

In addition to the high spectral efficiency, third-generation (3G) WCDMA provides even more dramatic change in capacity of the base station and the efficiency of the equipment. The high level of integration in the WCDMA is achieved due to the broadband carrier: a large number of users supported by the carrier, and less radio frequency (RF) carriers are required to provide the same capacity.

With less RF parts and more digital baseband processing, WCDMA can take advantage of the rapid evolution of digital signal processing capability. The level of integration of the high base station enables efficient building high capacity sites since the complexity of RF combiners, additional antennas or power cables can be avoided. WCDMA operators are able to provide useful data services, including navigation, person to person video calls, sports and video and new mobile TV clips.

WCDMA enables simultaneous voice and data which allows, for example, browsing or email when voice conferencing or video sharing in real time during voice calls.

The operators also offer mobile connectivity to the Internet and corporate intranet with maximum bit rate of 384 kbps downlink and both uplink. The first terminals and networks have been limited to 64 to 128 kbps uplink while the latter products provide 384 kbps uplink.

HSPA Standardization

High-speed downlink packet access (HSDPA) was standardized as part of 3GPP Release 5 with the first specification version in March 2002. High-speed uplink packet access (HSUPA) was part of 3GPP Release 6 with the first specification version in December 2004. HSDPA and HSUPA together are called High-Speed Packet Access‘

(HSPA).

The first commercial HSDPA networks were available at the end of 2005 and the commercial HSUPA networks were available on 2007. The HSDPA peak data rate available in the terminals is initially 1.8Mbps and will increase to 3.6 and 7.2 Mbps during 2006 and 2007, and later on 10Mbps and beyond 10Mbps. The HSUPA peak data rate in the initial phase was 1–2 Mbps and the second phase was 3–4Mbps.

HSPA is deployed over the WCDMA network on the same carrier or - for high capacity and high speed solution - using another carrier. In both cases, WCDMA and HSPA can share all the network elements in the core network and the radio network comprising base stations, radio network controller (RNC), Serving GPRS Support Node (SGSN) and the Gateway GPRS Support Node (GGSN). WCDMA and HSPA also share the site base station antennas and antenna cables.

The upgrade WCDMA HSPA requires new software and potentially new equipment in the base station and RNC to support the rate and higher data capacity. Because of the shared infrastructure between WCDMA and HSPA, the cost of the upgrade WCDMA HSPA is very low compared to the construction of a new stand-alone data network.

UMTS - Radio Interface and Radio Network Aspects

After the introduction of UMTS the amount of wide area data transmission by mobile users had picked up. But for the local wireless transmissions such as WLAN and DSL, technology has increased at a much higher rate. Hence, it was important to consider the data transmission rates equal to the category of fixed line broadband, when WIMAX has already set high targets for transmission rates. It was clear that the new 3GPP radio technology Evolved UTRA (E-UTRA, synonymous with the LTE radio interface) had to become strongly competitive in all respect and for that following target transmission rates were defined:

·      Downlink: 100 Mb/s

·      Uplink: 50 Mb/s

Above numbers are only valid for a reference configuration of two antennas for reception and one transmit antenna in the terminal, and within a 20 MHz spectrum allocation.

UMTS – All IP Vision

A very general principle was set forth for the Evolved 3GPP system. It should ―all IP‖, means that the IP connectivity is the basic service which is provided to the users. All other layer services like voice, video, messaging, etc. are built on that. Looking at the protocol stacks for interfaces between the network nodes, it is clear that simple model of IP is not applicable to a mobile network.

There are virtual layers in between, which is not applicable to a mobile network. There are virtual layer in between, in the form of ―tunnels‖, providing the three aspects - mobility, security, and quality of service. Resulting, IP based protocols appear both on the transport layer (between network nodes) and on higher layers.

UMTS – Requirements of the New Architecture

There is a new architecture that covers good scalability, separately for user plane and control plane. There is a need for different types of terminal mobility support that are: fixed, nomadic, and mobile terminals. The minimum transmission and signalling overhead especially in air, in an idle mode of the dual mode UE signalling should be minimized, in the radio channel multicast capability.

It is required to be reused or extended, as roaming and network sharing restrictions, compatible with traditional principles established roaming concept, quite naturally, the maximum transmission delay required is equivalent to the fixed network, specifically less than 5 milliseconds, set to control plane is less than 200 milliseconds delay target.

Looking at the evolution of the 3GPP system in full, it may not seem less complex than traditional 3GPP system, but this is due to the huge increase in functionality. Another strong desire is to arrive at a flat structure, reducing CAPEX/OPEX for operators in the 3GPP architecture carriers.

Powerful control functions should also be maintained with the new 3GPP systems, both real-time seamless operation (for example, VoIP) and non-real-time applications and services. The system should perform well for VoIP services in both the scenarios. Special attention is also paid to the seamless continuity with legacy systems (3GPP and 3GPP2), supports the visited network traffic local breakout of voice communications.

UMTS – Security and Privacy

Visitor Location Register (VLR) and SNB are used to keep track of all the mobile stations that are currently connected to the network. Each subscriber can be identified by its International Mobile Subscriber Identity (IMSI). To protect against profiling attacks, the permanent identifier is sent over the air interface as infrequently as possible. Instead, local identities Temporary Mobile Subscriber force (TMSI) is used to identify a subscriber whenever possible.

Each UMTS subscriber has a dedicated home network with which it shares a secret key K_i long term. The Home Location Register (HLR) keeps track of the current location of all the home network subscribers. Mutual authentication between a mobile station and a visited network is carried out with the support of the current GSN (SGSN) and the MSC / VLR, respectively. UMTS supports encryption of the radio interface and the integrity protection of signalling messages.

UMTS – WCDMA Technology

The first Multiple Access Third Generation Partnership Project (3GPP) Wideband Code Division networks (WCDMA) were launched in 2002. At the end of 2005, there were 100 WCDMA networks open and a total of more than 150 operators with licenses for frequencies WCDMA operation. Currently, WCDMA networks are deployed in UMTS band of around 2 GHz in Europe and Asia, including Japan and America Korea. WCDMA is deployed in the 850 and 1900 of the existing frequency allocations and the new 3G band 1700/2100 should be available in the near future. 3GPP has defined WCDMA operation for several additional bands, which are expected to be commissioned in the coming years.

As WCDMA mobile penetration increases, it allows WCDMA networks to carry a greater share of voice and data traffic. WCDMA technology provides some advantages for the operator in that it allows the data, but also improves the voice of base. Voice capacity offered is very high due to interference control mechanisms, including frequency reuse of 1, fast power control, and soft handover.

WCDMA can offer a lot more voice minutes to customers. Meanwhile WCDMA can also improve broadband voice service with AMR codec, which clearly provides better voice quality than fixed telephone landline. In short, WCDMA can offer more voice minutes with better quality.

In addition to the high spectral efficiency, third-generation (3G) WCDMA provides even more dramatic change in capacity of the base station and the efficiency of the equipment. The high level of integration in the WCDMA is achieved due to the broadband carrier: a large number of users supported by the carrier, and less radio frequency (RF) carriers are required to provide the same capacity.

With less RF parts and more digital baseband processing, WCDMA can take advantage of the rapid evolution of digital signal processing capability. The level of integration of the high base station enables efficient building high capacity sites since the complexity of RF combiners, additional antennas or power cables can be avoided. WCDMA operators are able to provide useful data services, including navigation, person to person video calls, sports and video and new mobile TV clips.

WCDMA enables simultaneous voice and data which allows, for example, browsing or email when voice conferencing or video sharing in real time during voice calls.

The operators also offer mobile connectivity to the Internet and corporate intranet with maximum bit rate of 384 kbps downlink and both uplink. The first terminals and networks have been limited to 64 to 128 kbps uplink while the latter products provide 384 kbps uplink.

WCDMA-3G

3G wireless service has been designed to provide high data speeds, always-on data access, and greater voice capacity. Listed below are a few notable points: The high data speeds, measured in Mbps, enable full motion video, high-speed internet access and video-conferencing. 3G technology standards include UMTS, based on WCDMA technology (quite often the two terms are used interchangeably) and CDMA2000, which is the outgrowth of the earlier CDMA 2G technology.

UMTS standard is generally preferred by countries that use GSM network. CDMA2000 has various types, including 1xRTT, 1xEV-DO and 1xEV-DV. The data rates they offer range from 144 kbps to more than 2 mbps.

Sub-systems of 3G Network

A GSM system is basically designed as a combination of three major subsystems:

·        Network Subsystem (NSS): MSC/VLR, HLR, AuC, SMSC, EIR, MGW. Common for both 2G & 3G Network.

·        UTRAN: RNC & RBS.

·        Operation and maintenance Support Subsystem (OSS).

There are three dominant interfaces, namely,

·        IuCS: Between RNC and MSC for speech & Circuit data;

·        IuPS: Between RNC & SGSN for packet data;

·        Uu interface: Between the RNC and MS.

UMTS – 3GPP

3^rd Generation Partnership Project or 3GPP, is the standardization group for mobile networks and is in existence since 1998. 3GPP specification come in bundles called

―Release‖.

3rd Generation Partnership Project (3GPP)

3GPP releases are from Release 99 to Release 7.

3GPP2 is the corresponding part of 3GPP market. 3GPP2 standards body has also developed a large set of specifications describing own mobile network technology, the current generation being labelled as CDMA2000 ©. 3GPP2 is 3GPP concepts and solutions, but is chosen selectively different. Regarding LTE, there has been a growing interest of 3GPP2 operators in recent years to allow between flexible and efficient. The inheritance 3GPP2 technology includes a component called 1xRTT CS and PS component (EVDO vs eHRPD). 3GPP2 consider their (eHRPD) high-speed packet data network as equivalent to 3GPP old system, the right to transfer procedures optimized specially designed.

Architecture of the 3GPP System

The overall architecture of the 3GPP, evolved system as well as the core and access networks already existing 3GPP defined are called "legacy 3GPP system".

The access networks which are not defined by the 3GPP, but may be used in conjunction with the evolved 3GPP system are called "non-3GPP access networks".

The area of service must be understood as the multitude of IP services, so in general they are represented and implemented by packet data networks (PDN). IP service can simply offer a raw IP connectivity (i.e. allowing an internet connection), providing a connection to a corporate network, or an advanced IP-based control functionality such as telephony and instant messaging via IMS.

It is called "Evolved UTRAN" (EUTRAN). GERAN and UTRAN are the existing radio access networks and are connected to the legacy PS domain. Evolved Packet Core (EPC) in addition to the basic functions to manage packet routing and forwarding (for the transport of user data) contains all the features necessary to control especially for mobility, session handling, safety and load.

For interworking with legacy CS domain, the CS core network should be considered as well and interfaced with the backend IMS. The dotted arrow indicates an optional interconnection between legacy CS core networks and the new network Evolved Packet Core, the decline in profit to the CS domain for voice services, if necessary.

UMTS – Radio Access Network

The more general term "Evolved Radio Access Network" (eRAN), can also be used as part of signalling protocols, as the term "access stratum" (AS) can be used. The comparison reveals that E-UTRAN consists of one type of nodes, namely Evolved Node B (eNodeB), and the variety of interconnections is reduced to a minimum. eNodeB is a radio base station and transmits/receives via its antenna in an area (cell), limited by physical factors (signal strength, interference conditions, and conditions of radio wave propagation). It has logical interfaces X2 with neighbouring eNodeB and the EPC via S1.

Both have a control part (that is, say for signalling) and a user plane part (for payload data). Point to the EU reference (which includes radio link interface and a mobile network protocol stack bound) is called "LTE-U u" to indicate that it differs from the legacy counterpart EU X2 connectivity neighbouring eNodeBs. They may be considered for most of the E-UTRAN and is used in most cases of handovers between radio cells.

As the UE moves, long handover preparation is done via signalling, through X2 between the two data eNodeBs and affected users can be transmitted between them for a short period of time. Only in special cases, it may happen that X2 is not configured for eNodeB between two neighbours. In this case transfers are always supported, but the preparation of transfer and the data transmission is then made via the EPC. Accordingly, higher latency and less "homogeneity" must therefore be provided.

In more detail, the functions performed by the eNodeB are:

·        Radio Resource Management: Radio Bearer Control, Radio Admission Control, Connection Control Mobility, dynamic allocation of resources (i.e. scheduling) to UES as uplink and downlink.

·        Header compression of IP and encryption of user data stream.

·        Forwarding the data packets of user plane to the EPC (especially, toward the GW node service).

·        Transport level packet marking in the uplink, for example, DiffServ code point setting, based on the QoS class index (QCI) of the EPS bearer associated.

·        Planning and delivery of paging messages (on request of MS).

·        Planning and transmission of broadcast information (origin of the MME or O & M).

·        Measurement configuration delivering and reporting on the extent of mobility and programming.

UMTS – evolved packet core

By the early architectural work for the system evolved 3GPP, two views on the implementation of mobility with the user plane and control plane protocols were presented. The first was promoted as the good performance of the GPRS Tunnelling Protocol (GTP), while the other pushed for the new (and the so-called "base" of the IETF) protocols.

Both had good arguments on their side:

·        GTP evolution : This protocol has proven its usefulness and capabilities to operators, and was very successful in the large scale operations. It was designed exactly to the needs of the mobile networks PS.

·        IETF based protocols : IETF is the de facto standards body for the internet. Their mobility protocols have evolved from focusing on mobile IP-based network client to "Proxy Mobile IP (MIP)." PMIP was standardized in 3GPP Evolved parallel system. (But Mobile IP client base is used in EPS in conjunction with non-3GPP access support.)

EPC for 3GPP access in non-roaming

The functions provided by the reference points and the protocols employed are:

LTE-Uu

LTE-Uu is the point of reference for radio interface between EU and eNodeB, encompasses control plane and user plane. The top layer of the control plan is called "Radio Resource Control" (RRC). It is stacked on "Packet Data Convergence Protocol" (PDCP), Radio Link Control and MAC layers.

S1-U

SI-U is the point for user plane traffic between eNodeB and serve GW reference. The main activity via this benchmark is to transfer IP packets encapsulated users arising from traffic or tunnel shape. Encapsulation is needed to realize the virtual IP link between eNodeB and GW service, even during the movement of EU, and thus enable mobility. The protocol used is based on GTP-U.

S1-MME

S1-MME is the point for the control plane between eNodeB and MME reference. All control activities are carried out on it, for example, signalling for attachment, detachment, and the establishment of the support of the change, safety procedures, etc. Note that some of this traffic is transparent to the E-UTRAN and is exchanged directly between EU and MS, it is a part called "non-access stratum" (NAS) signalling.

S5

S5 is the benchmark that includes the control and user plane between GW and PDN GW Service and applies only if both nodes reside in the HPLMN; the corresponding reference point when serving GW is VPLMN is called S8. As explained above, two protocol variants are possible here, an enhanced GPRS Tunnelling Protocol (GTP) and Proxy Mobile IP (PMIP).

S6a

S6a is the reference point for the exchange of information relating to subscriptions equipment (download and purging). It corresponds to Gr and D reference point in the existing system, and is based on the DIAMETER protocol.

SGi

This is the point of exit for DPR, and corresponds to the Gi reference point GPRS and Wi in I-WLAN. IETF protocols are based here for the user plane (i.e. IPv4 and IPv6 packet forwarding) protocols and control plane as DHCP and radius/diameter for configuring IP address/external network protocol are used.

S10

S10 is a reference point for the MME relocation purposes. It is a pure control plane interface and advanced GTP-C protocol is used for this purpose.

S11

S11 is a reference point for the existing control plane between MME and GW service. It employs the advanced GTP-C (GTP-C v2) protocol. The holder(s) of data between eNodeB and serve GW are controlled by the concatenation S1-S11 and MME.

S13

S13 is the reference point for Equipment Identity Register (EIR) and MME, and it is used for identity control (e.g. based on IMEI, if blacklisted). It uses the diameter protocol SCTP.

Gx

Gx is the reference point of the QoS policy filtering policy and control the load between PCRF and PDN GW. It is used to provide filters and pricing rules. The protocol used is the DIAMETER.

Gxc

Gxc is the reference point that exists in over Gx but is located between GW and PCRF and serves only if PMIP is used on S5 or S8.

Rx

Rx is defined as an application function (AF), located in NDS and PCRF for the exchange of policy and billing information; it uses the DIAMETER protocol.

EPC for 3GPP Access in Roaming

In roaming this case the user plane either:

Extends back to the HPLMN (via an interconnection network), which means that all EU user traffic is routed through a PDN GW in the HPLMN, where the DPRs are connected;

or For the sake of a more optimal way of traffic, it leaves a PDN GW in the VPLMN to a local PDN.

The first is called "home routed traffic" and the second is called "local breakout". (Note that the second term is also used in the discussion of traffic optimization for home NBs/eNodeB, but with a different meaning because in the concept of roaming 3GPP, the control plan always involves the HPLMN).

Interworking between EPC and Legacy

From the beginning, it was clear that the 3GPP Evolved system will interoperate seamlessly with existing 2G and 3G systems, 3GPP PS widely deployed or, more precisely, with GERAN and UTRAN GPRS base (For aspects of interworking with the old CS system for the treatment of optimized voice).

The question of the basic architectural design to 2G/3G in EPS is the location of the GGSN map. Two versions are available, and both are supported:

·        The GW used : It is the normal case where serving the GW ends the user plane (as seen in the existing GPRS network).The control plan is completed in the MME, according to the distribution of users and control plane in EPC. S3 and S4 reference points are introduced, and they are based on GTP-U and GTP-C, correspondingly. S5/S8 is chained to the PDN GW. The advantage is that interoperability is smooth and optimized. The downside is that for this kind of interoperability SGSN must be upgraded to Rel. 8 (due to the necessary support new features on S3 and S4).

·        The PDN GW : In this case the unchanged benchmark inheritance Gn (when roaming, it would Gp) is reused between SGSN and PDN GW, for both control and user plane. The advantage of this use is that SGSN can be pre-Rel. 8. Furthermore, it carries a certain restriction on IP versions, transfer and S5 / S8 protocol.

Interworking with Legacy 3GPP CS System

During the 3GPP Evolved design phase, it became clear that the legacy CS system, with its most important service "voice" communication, could not be ignored by the new system. The operators were simply too related investments in the field, and so very efficient interworking was requested.

Two solutions have been developed:

Single Radio Voice Call Continuity (SRVCC) for transferring voice calls from LTE (with voice over IMS) to the legacy system.

·        CS fallback: Enabling a temporary move to the legacy CS before a CS incoming or outgoing activity is performed.

Single Radio Voice Call Continuity (SRVCC)

In this solution chosen by 3GPP for SRVCC with GERAN/UTRAN, a specially reinforced MSC is connected via a new interface control plane for MME. Note that the MSC serving the EU can be different than supporting the Sv interface. In the IMS, an application server (AS) for SRVCC is necessary. Sv is based on GTPv2 and helps prepare resources in the target system (access and core network and the interconnection between CS and IMS domain), while being connected to access the source.

Similarly, with SRVCC CDMA 1xRTT requires interworking 1xRTT Server (IWS), which supports the interface and signal relay from / to 1xRTT MSC serving the UE S102 with the same purpose. S102 is a tunnel interface and transmits 1xRTT signaling messages; between MME and UE these are encapsulated.

CS Fallback

Serving GW and PDN GW are not separated (S5/S8 is not exposed) and the VLR is integrated with the MSC server. A new SG interface is introduced between the MSC Server/VLR and MME, allowing combined and coordinated procedures. The concept consists of:

·        Signal relay to end the CS request (incoming calls, handling network triggered additional service or SMS Legacy) from the MSC Server for MS on SG and vice versa;

·        The combined operating procedures between the PS domain and the CS domain.

Interworking with Non-3GPP Access

Interworking with different system of 3GPP access networks (called non-3GPP/access) was an important target for SAE; this should be done under the EPC umbrella. This interoperability can be achieved at different levels (and in fact, this was done on the layer 4 with VCC/SRVCC). But for the generic type of interworking, it seemed necessary to rely on generic mechanisms, so the IP level seemed most appropriate.

In general, complete systems for mobile and fixed networks have an architecture similar to that described above. For the evolved 3GPP system there is normally an access network and a core network. In the interworking architecture scheduled evolved 3GPP system, other access technologies systems connect to the EPC.

In general, complete mobile network system and fixed network systems have a similar architecture as described outlined in Evolved 3GPP system and normally consist of an access network and a core network/ It was also decided to allow two different types of interoperability, based on the property of the access systems. For networks with non-3GPP access confidence, it is assumed that secure communication between them and the EPC is implemented and also robust data protection is sufficiently guaranteed.

UMTS – GPRS Tunnelling Protocol

The generation of GPRS Tunnelling Protocol (GTP) was virtually impossible, but is also not desirable to give it for the new system, but, on the other hand, it is quite understandable that the improvements are also needed in order to be able to interact with the world of legacy PS smoothly and support functions needed for the newest system.

GPRS Tunnelling Protocol (GTP)

GTP protocol is designed for tunnelling and encapsulation of data units and control messages in GPRS. Since its design in the late 1990s, it was put to deploy on a large scale, and solid experience has been gathered. GTP for Evolved 3GPP system is available in two variants, control and user plane. GTP-C manages the control plane signalling, and it is necessary in addition to the data transfer protocol on the purity of the user, GTP-U; it is called user plane. Current versions, suitable for EPS are GTPv1 US and GTPv2-C.

The peculiarity of GTP is that it supports the separation of traffic within its primary GTP tunnel holder, or in other words, the ability to group them together and treat carriers. The ends of GTP tunnels are identified by TEIDs (Tunnel Endpoint identifiers); they are assigned to the local level for the uplink and downlink by peer entities and reported transversely between them. TEIDs are used on different granularity by specific example PDN connection on S5 and S8 and EU on S3 / S4 / S10 / S11 interfaces.

Control Plane of GPRS Tunnelling Protocol

GTPv2-C is used on the EPC signalling interfaces (including SGSNs of at least Rel. 8). For

example:

·        S3 (between SGSN and MME),

·        S4 (between SGSN and Serving GW),

·        S5 and S8 (between Serving GW and PDN GW),

·        S10 (between two MMEs), and

·        S11 (between MME and Serving GW).

Corresponding to this, a typical GTPv2-C protocol data unit like shown in the figure above, the specific part GTP is preceded by IP and UDP headers, it consists of a header GTPv2-C and part containing information GTPv2-C variable in number, length and format, depending on the type of the message. As the echo and the notification of a protocol version is not supported, TEID information is not present. The version is obviously firmly set at 2 in this version of the protocol.

GTP had a complex legacy extension header mechanism; it is not used in most GTPv2-C. The message type is defined in the second byte (so the maximum of 256 messages can be defined for future extensions). Below table provides an overview of messages currently defined GTPv2-C. The length of the message is coded in bytes 3 and 4 (measured in bytes and not containing the first four bytes themselves).

TEID is the ID of the tunnel end point, a single value on the opposite/receiving side; it allows multiplexing and de-multiplexing tunnels at one end in the very frequent cases over a GTP tunnel must be distinguished.

Enhanced GTPv1-U

Only a small but effective improvement was applied to GTP-U, and for that it was not considered necessary to strengthen the number of protocol version. Thus, we still expect GTPv1-U, but at least it‘s most recent Rel. 8.

The protocol stack is essentially the same as for GTPv2-C with only the name of the layers and the protocols substituted accordingly. The extension header mechanism is kept in place; it allows inserting two elements if necessary.

·        UDP source port of the triggering message (two octets);

·        PDCP PDU number: related to the characteristic transfer without loss; in this case, data packets need to be numbered in the EPC (two octets).

The improvement is the ability to transmit an "end market" in the user plane. It is used in the inter-eNodeB handover procedure and gives the indication that the pathway is activated immediately after the data packet, for example, the feature is not necessary to pre-Rel.8 because GTP-U did not end in the radio access node (i.e. not in the BS or NodeB) only a few messages exist. GTPv1-U, and they are listed in the table above.

It is clear that, in fact a very limited kind of signaling is possible via GTPv1-U (echo mechanisms and end labeling). The only message that the transfer of real user data is of type 255, the so-called G-PDU message; the only piece of information it carries, after the header is the original data packet from a user or external PDN equipment.

Not all instances of GTP-U tunnels are listed in the reference architecture (which aimed to capture the associations were no longer living between network nodes); temporary tunnels are possible:

·        Between two Serving GWs, applicable for the transfer based on S1, in the case that the service is moved GW;

·        Between two SGSNs, corresponds to the previous case, but in the legacy PS network;

·        Between two RNCs, applicable for the relocation of the RNC in the 3G PS network (no relation to the EPC, it is mentioned here just for completeness).