Connection identifier system and method

Abstract

A method and system are proposed for establishing a requested connection between a source node and a destination node in a telecommunications network. The system and method are described in relation to a 3GPP network, but are applicable to other types of networks. The method includes generating a source application identifier for the connection within the source node, retrieving a source node identifier for the source node and transmitting the source application identifier and the source node identifier to the destination to provide a source connection identifier for the requested connection between the source node and the destination node.

Claims

1. A MME (Mobility Management Entity) comprising: a transmission circuit configured to transmit, to a base station, a handover request message which includes a Mobility Management Entity User Equipment S1 Application Protocol (MME UE S1AP) identity and a global MME identity; and a receiving circuit configured to receive, from the base station, a handover request acknowledgement message, wherein the MME UE S1AP identity uniquely identifies the UE over the S1 interface within the MME.

2. A method performed by a MME (Mobility Management Entity), the method comprising: transmitting, to a base station, a handover request message which includes a Management Entity User Equipment S1 Application Protocol (MME UE S1AP) identity and a global MME identity; and receiving, from the base station, a handover request acknowledgement message, wherein the MME UE S1AP identity uniquely identifies the UE over the S1 interface within the MME.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) An embodiment of the invention will now be described, by way of example, with reference to the accompanying drawings in which:

(2) FIG. 1 schematically illustrates a mobile telecommunication system of a type to which the embodiment is applicable;

(3) FIG. 2 schematically illustrates a base station forming part of the system shown in FIG. 1;

(4) FIG. 3 schematically illustrates a gateway device forming part of the system shown in FIG. 1;

(5) FIG. 4a schematically illustrates the establishment of a connection between an eNodeB and an MME component;

(6) FIG. 4b schematically illustrates the establishment of a connection between two eNodeBs;

(7) FIG. 5 schematically illustrates a potential error situation that may arise in a prior art system;

(8) FIG. 6 schematically illustrates a second potential error situation that may arise in a prior art system;

(9) FIG. 7 schematically illustrates a third potential error situation that may arise in a prior art system;

(10) FIG. 8a is a schematic illustration of a first message from an eNodeB to an MME according to one embodiment; and

(11) FIG. 8b is a schematic illustration of a first message from an MME to an eNodeB according to one embodiment.

BEST MODE FOR CARRYING OUT THE INVENTION

Overview

(12) The following description sets out a number of specific embodiments of the method and system claimed herein. It will be clear to one skilled in the art that variations of the features and method steps may be provided and that many of the features described are not essential to the invention, the scope of which is defined by the claims.

(13) FIG. 1 schematically illustrates a mobile (cellular) telecommunication system 1 in which users of mobile (or cellular) telephones (MT) 3-0, 3-1, and 3-2 can communicate with other users (not shown) via one of the base stations 5-1 or 5-2 and a telephone network 7. The telephone network 7 includes a plurality of components including gateway components 9-1, 9-2. It will be appreciated by the skilled person that the each base station 5-1, 5-2 may connect to the telephone network 7 via either gateway 9-1, 9-2 and that both base stations 5-1, 5-2 may connect via the same gateway 9-1, 9-2. Similarly, each mobile telephone 3 may connect to the telephone network 7 via either base station 5 and that both mobile telephones 3 may connect via the same base station 5.

(14) In this embodiment, the base stations 5 use an orthogonal frequency division multiple access (OFDMA) technique in which the data to be transmitted to the mobile telephones 3 is modulated onto a plurality of sub-carriers. Other well known data transmission techniques may also be used. When a mobile telephone 3 enters the network 7, for example by being switched on, a connection is established between the mobile telephone 3 and a base station 5 and between the base station 5 and a gateway device 9. This enables communication between the mobile telephone 3 and other components in the network 7.

(15) Also, when a mobile telephone 3 moves from the cell of a source base station (e.g. base station 5-1) to a target base station (e.g. base station 5-2), a handover procedure (protocol) is carried out in the source and target base stations 5 and in the mobile telephone 3, to control the handover process. The handover is enabled by the establishment of a connection between the source and target base stations 5. As part of the handover process, the gateway device 9-1, 9-2 via which communications from a mobile telephone 3 are transmitted to the telephone network may change. Alternatively, the gateway device 9-1, 9-2 through which communications are transmitted may remain the same, but the base station 5-1, 5-2 to which the mobile device is connection may change. These transfers are also enabled by the establishment of connections between the base stations 5 and the gateways 9.

(16) Base Station

(17) FIG. 2 is a block diagram illustrating the main components of each of the base stations 5 used in this embodiment. As shown, each base station 5 includes a transceiver circuit 21 which is operable to transmit signals to and to receive signals from the mobile telephones 3 via one or more antennae 23 and which is operable to transmit signals to and to receive signals from the telephone network 7 via a network interface 25. The network interface 25 includes an S1 network interface for communicating with network components, such as gateway nodes, using the S1 protocol. The network interface 25 also includes an X2 Interface for communicating with other base station components using the X2 protocol. A controller 27 controls the operation of the transceiver circuit 21 in accordance with software stored in memory 29. The software includes, among other things, an operating system 31, an application identifier generator 33 and a connection identifier generator 35 which may also be referred to as a context identifier generator. The memory stores a node identifier 34 and in some embodiments also stores a network identifier for the network with which the base station is associated. The operation of the application identifier generator 33 and the connection identifier generator 35 are described below.

(18) Gateway

(19) FIG. 3 is a block diagram illustrating the main components of each of the gateway components 9 used in this embodiment. As shown, each gateway 9 includes a transceiver circuit 41, which is operable to transmit signals to and to receive signals from at least one base station 5 via a base station interface 43 and which is operable to transmit signals to and receive signals from the rest of the telephone network 7 via a network interface 45. A controller 47 controls the operation of the transceiver circuit 41 in accordance with software stored in memory 49. The software includes, among other things, an operating system 51, an application identifier generator 53 and a connection identifier generator 55, which may also be referred to as a context identifier generator. The memory stores a node identifier 54 and in some embodiments also stores a network identifier for the network with which the base station is associated. The operation of the application identifier generator 53 and the connection identifier generator 55 are described below.

(20) In the above description, both the base stations 5 and the gateways 9 are described for ease of understanding as having respective discrete modules which operate according to the methods described herein. Whilst the features may be provided in this way for certain applications, for example where an existing system has been modified to implement the invention, in other applications, for example in systems designed with the inventive features in mind from the outset, these features may be built into the overall operating system or code and so the modules described above may not be discernable as discrete entities.

(21) The following description will use the nomenclature used in the Long Term Evolution (LTE) of UTRAN. Therefore, the mobile telephone 3 will be referred to as a UE, each base station 5 will be referred to as an eNodeB (or eNB) and each gateway component will be referred to as an MME. The protocol entitles used in LTE have the same names as those used in UMTS (Universal Mobile Telecommunication System) except for the Radio Unk Control (RLC) entities which, under LTE, are called the Outer ARQ (Automatic Repeat Request) entities. The Outer ARQ entities of LTE have substantially the same (although not identical) functionality to the RLC entities of UMTS.

(22) Error Scenarios

(23) As set out above, in prior art systems application identifiers sent with the context establishment request messages do not necessarily uniquely identify the context at the receiving component. Details of identifiers that are sent within context set-up messages in this embodiment are provided below. However, examples of situations in which a problem may arise are first discussed in more detail.

(24) A transport protocol connection, such as a Stream Control Transmission Protocol (SCTP) connection, may carry signals relating to different UEs. In order to forward the signalling internally to the correct UE context manager in the receiving node, application routing information is included in the S1 or X2 application (AP) messages. This will be further explained and illustrated with reference to FIGS. 4a and 4b.

(25) When a sending node requests, in some cases implicitly, the establishment of a dedicated S1/X2 signalling connection for a certain UE, e.g. by sending an S1-AP Initial UE message, it informs the peer node on the S1/X2 interface the application (AP) identifier it has assigned for that UE. In the case of a connection between an eNodeB 61 and an MME 63, this message takes the form “eNB S1-AP” 65 as shown in FIG. 4a, which may also be written as “eNB S1-AP UE ID”. For a transfer from one eNodeB 69 to a peer eNodeB 71, the message is of the form “Source eNB S1-AP” 73 as shown in FIG. 4b. The response message of the peer node then contains the S1/X2 Application Identifier of both the originating and the peer node, for example “eNB S1 AP. MME S1-AP” 67 as in FIG. 4a, which may also be written as “eNB S1-AP UE ID. MME S1-AP UE ID” or, in the case shown in FIG. 4b “Source eNB S1-AP, Target eNB S1-AP” 75.

(26) Further details of the application identifiers used in the S1 and X2 interfaces are provided below:

(27) eNodeB S1-AP UE Identity

(28) The eNodeB S1-AP UE identity 65 is allocated to uniquely identify the UE over the S1 interface within the eNodeB 61. When an MME 63 receives the eNodeB S1-AP UE Identity 65, it stores it for the duration of the UE-associated logical S1 connection for this UE. Once known to the MME 63, this Identity (IE) may be included in all UE associated S1-AP signalling (uplink (UL) as well as downlink (DL)).

(29) MME S1-AP UE Identity

(30) The MME S1-AP UE Identity 687 is allocated to uniquely identify the UE over the S1 interface within the MME 63. When the eNodeB 61 receives the MME S1-AP UE identity it stores it for the duration of the UE-associated logical S1 connection for this UE. Once known to the eNodeB, 61 the IE may be included in all UE associated S1-AP signalling (UL as well as DL).

(31) Source eNodeB UE Context ID

(32) The source eNodeB UE Context ID (Source eNB S1-AP) 73 is allocated to uniquely identify the UE over the X2 interface with the source eNodeB 69. When the target eNodeB 71 receives the source eNodeB context ID, it stores it for the duration of the context for that UE. Once known to the target eNodeB 71, the IE may be included in all UE associated X2-AP signalling.

(33) Target eNodeB UE Context ID

(34) The target eNodeB UE Context ID (Target eNB S1-AP) 75 is allocated to uniquely identify the UE over the X2 interface with the target eNodeB 71. When the source eNodeB 69 receives the target eNodeB context ID, it stores it for the duration of the context for that UE. Once known to the source eNodeB 69, the IE may be included in all UE associated X2-AP signalling.

(35) Hence, as set out above, in previous embodiments, the Application Identifiers (AP IDs) are unique within the generating node. As a consequence, they may not be uniquely identified at the receiving peer node when it receives the first message (i.e. before the signalling connection exists). There are described below some scenarios depicting potential error situations in cases in which the application identifier is not unique in the receiving node.

(36) In some situations, errors occur due to a node receiving two messages for triggering the establishment of an S1 or X2 connection containing the same S1/X2 application identifier from the originating node.

(37) A first error situation will be described with reference to FIG. 5. In FIG. 5, two eNodeBs 77, 79 are connected to the same MME 81. The MME 81 receives two S1-AP initial UE messages, which should trigger the establishment of two S1 connections. However, the two eNodeBs 77, 79 have by chance assigned the same eNB S1-AP UE identity to the messages. The application layer in the MME 81 produces an error since it cannot make a distinction between the two messages. Error handling is then undertaken by the system, which may result in a delay in connections being established or a failure to establish at least one connection.

(38) A second potential error situation is illustrated in FIG. 6, which illustrates schematically a flex scenario. The eNodeB 83 is the target for both UE1 89 and UE2 91, served respectively by MME1 85 and MME2 87, during an intra LTE Handover with MME involvement. During the Handover process, the MME 85 sends to the target eNodeB 83 the S1-AP Handover Request message. The target eNodeB 83 may receive two S1-AP Handover Request messages from the two MMEs 85, 87 containing the same MME S1-AP UE Identity. The application layer in the target eNodeB 83 produces and error since it cannot make a distinction between the two messages. Error handling is then performed, which may cause delays and communication problems as above.

(39) A further problem in the situation illustrated in FIG. 6 arises if an S1 Handover Cancel message is sent before the S1 Handover Request Acknowledgement is received. In this situation, the target eNodeB 83 will not know which resources to release.

(40) A third potential error situation is illustrated in FIG. 7 which illustrates intra LTE mobility without MME involvement. In this situation, eNodeB2 99 is the target eNodeB for an intra LTE Handover triggered by eNodeB1 97 for an intra LTE Handover triggered by eNodeB3 93. As a consequence, eNodeB2 99 receives two X2 Handover Request messages from eNodeB1 97 and eNodeB3 93 which by chance use the same Source eNodeB UE Context ID. The application layer in the target eNodeB2 99 may get confused and cannot make a distinction between the two messages. Error handling is then performed to resolve the situation.

(41) A further problem in the situation illustrated in FIG. 7 also arises if a Handover Cancel Request is received from eNodeB1 97 or eNodeB3 93 before eNodeB2 99 can reply to the original Handover Request. In this case, the eNodeB2 99 will not know which resources should be released.

(42) Operation

(43) To avoid situations such as those set out above, it has been appreciated that further identification of connection request messages would be helpful. This is implemented by enabling the S1 and X2 interfaces to use distinct application identifiers. This may be achieved in a number of different ways and these are summarised and described in more detail below.

(44) In one embodiment, a new Node Identifier is added to the relevant S1/X2-AP messages. The Node Identifier may identify the source eNodeB or MME. In this embodiment, the Node Identifier is used only in the first message, in which the connection establishment is requested. The Node Identifier is not needed for the reply message and subsequent messages, which revert to exchanging only the application identifier. This reduces the amount of information (bits) that needs to be carried in the subsequent messages. This implementation is illustrated schematically in FIGS. 8a and 8b. FIG. 8a illustrates a first message from an eNodeB to an MME which includes both an eNB S1-AP UE id and an eNB Global id. FIG. 8b illustrates a first message from an MME to an eNodeB which includes both an MME S1-AP UE id and an MME Global id. Hence a predetermined number of bits are reserved in the initial S1-AP message to represent an identifier for the eNodeB or MME.

(45) In an alternative embodiment, a Node Identifier is embedded inside the eNB/MME S1-AP identifier IE. That is, the connection request message includes a single identifier, but the identifier incorporates both information sufficient to identify the source node and information to identify the connection within the source node.

(46) In a further embodiment, a new IE is added representing a logical identifier that is related by means of a well known function to a transport layer (TNL) identifier for the node, such as the IP address or SCTP port for the node. For example each eNodeB will have a different IP address and this may be used to derive a source node identifier. It is further noted that, in a typical implementation, the TNL does not inform the application protocol of the source IP address, thus the application protocol is agnostic of the sender ID unless it is indicated in the application message itself. This arises from the UMTS concept that the TNL and RNL should be implemented independently.

(47) In a further implementation, the range of S1/X2-AP identifiers that one eNodeB or MME can assign is restricted so that the identifier ranges for two nodes do not overlap. This may be done, for example, by configuring a different range of identifiers for each eNodeB or MME during the pre-operational stage. Each node then cycles through its predefined range of identifiers as it establishes connections. This solution enables a unique application identifier to be provided from each node within one network. A further network identifier (such as the PLMN ID) may also be needed in some implementations where multiple networks are used to identify the network. This may be provided either embedded in the application identifier or as a separate identifier.

(48) The implementation of a system according to any of the embodiments described above resolves the error situations set out above.

(49) In the first error situation that was illustrated in FIG. 5, the addition or incorporation of an eNodeB identifier into each of the S1-AP Initial UE messages from eNodeB1 77 and eNodeB2 79 provides a unique identifier for each connection request received at the MME and hence enables the MME to provide a one-to-one mapping between the eNB S1-AP UE id and the MME S1-AP UE id.

(50) Similarly. In the second error situation described in relation to FIG. 6, the addition or incorporation of an MME identifier into each of the MME S1-AP UE identifiers from MME1 85 and MME2 87 would uniquely identify the MME messages (and hence the originating MMEs) at the eNodeB 83. Further, if the handover request was cancelled, since the MMEs are uniquely identified in the initial message, the eNodeB 83 would know which resources to release to enable a connection to be cancelled without affecting establishment of the other connection.

(51) Finally, in the third error situation described in relation to FIG. 7, eNodeB1 and eNodeB3 would send X2-AP Handover Requests including source eNodeB identifiers. Hence eNodeB2 could uniquely identify each of the two source nodes and would know which resources to release if the handover request was cancelled.

(52) In one embodiment, in the third situation described in relation to FIG. 7, the Handover Request message also contains an MME S1-AP UE identifier, which includes an identifier of the MME component for the S1 signalling connection. This enables the target eNodeB2 99 to contact the MME by looking to the MME S1-AP UE identifier.

(53) The node identifiers incorporated in the connection setup messages may be globally unique or may be unique only within the telephone network in which the node operates, for example within the node's PLMN network. In one embodiment, a global node identifier may be provided by making use of both a node ID (which is unique within a PLMN) and a PLMN identifier. This would enable each S1/X2 AP to be unique even in a RAN sharing scenario.

(54) The application identifiers for the connection are transmitted with the initial connection set up messages. To reduce the overhead in subsequent messages, node identifiers may be omitted from at least some of the messages subsequently sent within the connection.

(55) As mentioned above, the establishment of a connection in a mobile telecommunications network may occur when a new mobile device (UE) enters the network, for example when the mobile device is switched on. New connections may also be established to enable handover as a mobile device moves within a network. Additionally, connections may be established due to configuration changes in the network nodes, for example if a new base station is added or an MME fails.

(56) The generation of an application identifier may be achieved by providing a generator module that uses an algorithm to generate an application identifier dynamically. In an alternative embodiment, the application identifier may be retrieved from a memory or buffer and the node may cycle through using a sequence of different application identifiers for sequential connections.

(57) To determine a node identifier, the node may simply retrieve a fixed identifier from storage in memory. Alternatively, the node may obtain its identifier from a generator module that dynamically generates a node identifier. As set out above, the node identifier may be based on a transport layer identifier for the node, such as an IP address for the node.

(58) In one embodiment, the connection identifier is generated using the application identifier and the node identifier. In a further embodiment, the connection identifier is generated using the application identifier and the network identifier. In a third embodiment, the connection identifier is generated using the application identifier, the node identifier and the network identifier. The identifiers may be concatenated to form a single connection identifier or a function may be used to combine the identifiers. The use of a function to combine the identifiers may make the connection identifier shorter than it would be if the identifiers were concatenated. However, concatenation of the identifiers may enable individual identifiers to be removed from subsequent messages once the connection has been established.

Glossary of 3GPP Terms

(59) LTE—Long Term Evolution (of UTRAN)

(60) eNodeB—E-UTRAN Node B

(61) AGW—Access Gateway

(62) UE—User Equipment—mobile communication device

(63) DL—downlink—link from base to mobile

(64) UL—uplink—link from mobile to base

(65) AM—Acknowledge Mode

(66) UM—Unacknowledge Mode

(67) MME—Mobility Management Entity

(68) UPE—User Plane Entity

(69) HO—Handover

(70) RLC—Radio Link Control

(71) RRC—Radio Resource Control

(72) RRM—Radio Resource Management

(73) SDU—Service Data Unit

(74) PDU—Protocol Data Unit

(75) NAS—Non Access Stratum

(76) ROHC—Robust Header Compression

(77) TA—Tracking Area

(78) U-plane—User Plane

(79) TNL—Transport Network Layer

(80) S1 Interface—Interface between Access Gateway and eNodeB

(81) X2 Interface—Interface between two eNodeB

(82) MMEs/SAE Gateway—New name for Access Gateway having both MME and UPE entities

(83) The following is a detailed description of the way in which the present inventions may be implemented in the currently proposed 3GPP LTE standard. Whilst various features are described as being essential or necessary, this may only be the case for the proposed 3GPP LTE standard, for example due to other requirements imposed by the standard. These statements should not, therefore, be construed as limiting the present invention in any way.

(84) Title: Use of Global Node Id in Support of Application Routing

(85) 1 Scope

(86) The scope of this contribution is to: highlight some error scenarios which are caused by the S1/X2 application identifier not being globally unique propose several solutions to overcame this issue
2 Discussion
2.1 Current Status

(87) An SCTP connection carries signalling related to different UEs. In order to forward internally the signalling to the correct UE context manager, application routing information must be included in the S1/X2 AP messages.

(88) Whenever a sending node requests implicitly the establishment of a dedicated S1/X2 signalling connection for a certain UE, by e.g. sending the S1-AP Initial UE message, it will inform the peer node on the S1/X2 AP identifier it has assigned for that UE.

(89) The response message of the peer node will contain, in a normal scenario, the S1/X2 application id of both originating and peer node (see FIGS. 4a & 4b).

(90) See below definitions of Application identifier in both S1 and X2 interface [1]:

(91) eNB S1-AP UE Identity:

(92) The eNB S1-AP UE Identity shall be allocated so as to uniquely identify the UE over the S1 interface within the eNB. When MME receives eNB S1-AP UE Identity it shall store it for the duration of the UE-associated logical S1-connection for this UE. Once known to MME this IE is included in all UE associated S1-AP signalling (UL as well as DL).

(93) MME S1-AP UE Identity:

(94) The MME S1-AP UE Identity shall be allocated so as to uniquely identify the UE over the S1 interface within the MME. When eNB receives MME S1-AP UE Identity it shall store it for the duration of the UE-associated logical S1-connection for this UE. Once known to eNB this IE is Included in all UE associated S1-AP signalling (UL as well as DL).

(95) Source eNB UE Context ID:

(96) The source eNB UE Context ID shall be allocated so as to uniquely identify the UE over the X2 interface within the source eNB. When target eNB receives source eNB UE Context ID it shall store it for the duration of the context for this UE. Once known to target eNB this IE is included in all UE associated X2-AP signalling.

(97) Target eNB UE Context ID:

(98) The Target eNB UE Context ID shall be allocated so as to uniquely identify the UE over the X2 interface within the target eNB. When target eNB receives target eNB UE Context ID it shall store it for the duration of the context for this UE. Once known to source eNB this IE is included in all UE associated X2-AP signalling.

(99) The exact definitions are not yet agreed but it is well known/agreed among the 3GPP groups that these application identifiers are unique within the generating node.

(100) According to the definitions above these AP ids, are unique within the generating node. As a consequence they are not univocally identified at the receiving peer node when it receives the first message (i.e. signalling connection does not exist) unless some node-specific identifier are also present in the message. In the paragraph 2.3 some scenarios, depicting the potential error situation in case of the application id is not globally unique, are explained.

(101) 2.2 T-Plane Identifiers

(102) Since Control Plane relies on SCTP over IP on both S1 and X2 interface, following TNL identifiers exist:

(103) SCTP Layer:

(104) Source Port Number, Destination Port Number

(105) SCTP stream id

(106) IP Layer:

(107) Source/destination IP address

(108) In LTE there will be one SCTP association for each S1/X2 interface, with one pair of SCTP streams for common procedure and few pairs of SCTP streams to be shared by all S1/X2 dedicated signalling connections.

(109) The SCTP Source port in combination with the source IP address, the SCTP destination port and possibly the destination IP address is used to identify the association to which this packet belongs.

(110) As a consequence, the SCTP association may be used as node identifier (SCTP stream id is not useful for this purpose).

(111) However, in line with the UMTS principle, it is proposed not to use the TNL identifier (i.e. SCTP association) for application protocol purpose in order to keep the RNL and TNL as much independent as possible and to avoid any implementation constraints for e.g. blade/SW entities communication.

(112) 2.3 Potential Error Situation Scenarios

(113) This paragraph lists some possible scenarios which may lead to an error situation in a peer node receiving two messages (which trigger the establishment of S1/X2 connection) containing the same S1/X2 application identifier the originating node. 1) In FIG. 5, two eNBs are connected to the same MME: the MME receives two S1-AP Initial UE messages (which should trigger the establishment of the S1 connection), from 2 different eNBs which by chance have assigned the same eNB S1-AP UE Identity. The application layer in the MME will get confused and, since the message does not contain any information related to the originating node, it cannot make a distinction between the two messages. Error handling will then be performed according to the implementation. This could be avoided by using an S1-AP identifier which is globally unique. 2) In FIG. 6, an S1 flex scenario is depicted: the eNB is the target eNB for both UE1 and UE2, served respectively by MME1 and MME2, during an Intra LTE Handover with the MME involvement. During the Handover preparation, the MME sends to the target eNB the S1-AP: Handover Request message. The target eNB may receive two S1-AP: Handover Request messages from the two different MMEs containing the same MME S1-AP UE Identity. The application layer in the target eNB will get confused and, if the message does not contain any information related to the originating node it cannot make a distinction between the two messages. Error handling will then be performed according to the implementation. This could be avoided by using an S1-AP identifier which is globally unique. 3) In FIG. 7, eNode B2 is the target eNB for the Intra LTE Handover triggered by eNB1 and the Intra LTE Handover triggered by eNB3. As a consequence the eNB2 may receive two X2: Handover Request messages from eNB1 and eNB3 which by chance have assigned the same Source eNB UE Context ID The application layer in the target eNB2 will get confused and, since the message does not contain any information related to the originating node, it cannot make a distinction between the two messages. Error handling will then be performed according to the implementation. This could be avoided by using a Source eNB UE Context ID (i.e. X2 AP Identifier) which is globally unique.
2.4 Proposals

(114) In order to avoid these misleading situations, it is proposed to use a globally unique application id on both S1 and X2 interfaces. This may be reached in different ways:

(115) Solution One: Add a new IE, i.e. Global Node id, in the relevant S1/X2-AP messages

(116) Solution Two: Embedding the Global Node id inside the eNB/MME S1-AP Identifier IE

(117) Solution 3: Add a new IE representing a logical identifier which somehow is related (by mean of a well known function) to TNL identifier (IP address/SCTP port)

(118) Solution 4: Restrict the range of S1/X2AP identifiers that one eNB can assign so that the range of the two different nodes do not overlap (by e.g. configuring a different range for each eNB during the pre-operational state). This solution however will allow having a unique AP-id within one network. As a consequence a network identifier (i.e. PLMN id) is needed either embedded in the AP-id IE or as a separate IE. In case of a separate IE, this will be needed only in the message triggering the connection establishment.

(119) Solution one and two rely on the concept of Global node identifier which make use of:

(120) Node id (unique within one PLMN) and PLMN identifier (this guarantee that the S1/X2 AP are unique even in a RAN sharing scenario).

(121) The first three solutions differ in terms of implementation complexity: NEC preference is Solution 1.

(122) In case of solution one, the Global Node id will be needed only for the first message. Global Node id is not needed for the reply message neither for subsequent messages.

(123) Therefore, amount/bits of information that needs to be carried in the subsequent messages can be reduced by applying solution one.

(124) The fourth solution requires some configuration to be done in the nodes in a pre-operational state.

(125) 3 Conclusion

(126) This contribution has pointed out some potential error scenarios due to the S1/X2-AP ids not being globally unique.

(127) Four proposals have been made to overcome this issue i.e.: Introduction of Global Node id IE in relevant S1/X2 messages Embedding of the Global node id in the S1-AP id IE Add a new IE representing a logical identifier related to TNL identifiers S1-AP range configuration in each eNB in combination with a network identifier (either as a separate IE or embedded in the AP id IE).

(128) It is proposed to discuss the consequences of not having a unique application id and to agree on one of the proposals listed in paragraph 2.4.

(129) If the principle is accepted, NEC is available in writing the related CR.

(130) 4 Reference

(131) R3-071344 “Discussion and proposal for the AP ID handling”, NEC

(132) This application is based upon and claims the benefit of priority from United Kingdom Patent Application No. 0715940.3, filed on Aug. 15, 2007, the disclosure of which is incorporated herein in its entirety by reference.