WO2010072265A1

WO2010072265A1 - Network apparatus and method for identifying network resources

Info

Publication number: WO2010072265A1
Application number: PCT/EP2008/068238
Authority: WO
Inventors: Zoltan Nagy; Csaba Vulkan
Original assignee: Nokia Siemens Networks Oy
Priority date: 2008-12-23
Filing date: 2008-12-23
Publication date: 2010-07-01

Abstract

According to an exemplary embodiment of the invention a network apparatus (200) for identifying a network resource in a network (100) may be provided. The network apparatus may comprise a receiving device (210) and an evaluating device (220). The receiving device (210) may be adapted for receiving a first network time value. Furthermore, the receiving device (210) may be adapted for receiving a second network time value. In addition, the evaluating device may be adapted for classifying the received first network time value and the received second network time value using at least a first class and a second class and wherein at least one class may indicate a network resource (250, 300, 350, 400).

Description

Network apparatus and method for identifying network resources

Technical field of the invention

The present invention relates to telecommunication networks. In particular the present invention relates to a network apparatus, to a method for identifying a network resource, to a communication system and to a computer-readable medium.

Background of the invention

Telecommunication networks may comprise a plurality of network apparatuses, which may send one or a plurality of flows over the network. The network may comprise several subnetworks, which may provide one or a plurality of links between several network apparatuses. Some network apparatuses may send a flow or an information in a flow, other network apparatuses may receive one or more flows and further network apparatuses may receive and send one or more flows.

The document 3GPP TS 25.425 UTRAN Iub Interface User Plane Protocols for Common Transport Channel data streams, Release 8, Version 8.0.0 of September 2008, may describe UTRAN RNS- RNS (Iur) interface user plane protocols for Common Transport Channel data streams.

The document 3GPP TS 25.413 UTRAN Iu Interface: RANAP Signaling, Release 8, Version 8.0.1 of September 2008, may describe radio network layer signalling protocol called Radio Access Network Application Part (RANAP) for the Iu interface. The document 3GPP TS 25.402 Synchronisation in UTRAN Stage 2 Release 6, Version 6.6.0 may describe specifications of different synchronisation mechanisms in UTRAN and on Uu.

The document "Some methods for classification and analysis of multivariate observations" by J. MacQueen, In: Proceedings of 5^th Berkeley Symposium on Mathematical Statistics and Probability, 1967 may describe a process for partitioning an N-dimensional population into k sets on the basis of a sample.

In WO 2008/064983 A2 a method for decoupling congestion control in a cascade of network elements of an UMTS radio access network may be described.

The document "QoS Aware HSDPA Congestion Control Algorithm Networking and Communications" by Laszlό Kδrδssy, Csaba Vulkan, 2008 In: WIMOB '08, IEEE International Conference on Wireless and Mobile Computing, 12-14 October 2008, pages 404- 409, ISBN 978-0-7695-3393-3, may describe an approach for detecting a congestion in a network with one interface.

Summary of the invention

There may be a need to provide a more efficient network apparatus .

According to an exemplary embodiment of the invention a network apparatus for identifying a network resource in a network may be provided. The network apparatus may comprise a receiving device and an evaluating device. The receiving device may be adapted for receiving a first network time value. In addition, the evaluating device may be adapted for classifying the received first network time value into a class, wherein the class may indicate a network resource.

Thus, a network apparatus may be provided which may be suitable for situations in a network in cases when a Drift

Radio Network Controller (DRNC) may have forwarding functionality.

The network apparatus may be installed in a network. The network may be a transport network. In the network an in- sequence-delivery may be provided. An in-sequence delivery may mean that a first information, which may have been sent firstly in time in comparison to a second information, which may have been sent timely after the first information, may be received timely before receiving the second information. In other words the order of information or messages may remain unchanged in the network.

Furthermore, a first network apparatus may send an information in a flow, wherein the sending network apparatus may send a time information, i.e. a time stamp. A second network apparatus may receive the time information, i.e. a network time value originating from the first network apparatus. The network may comprise nodes, which nodes may be synchronized in time to each other, meaning i.e. one second may be the same time span in each clock of a network apparatus. However, the nodes may not be phase aligned, meaning different nodes may comprise different absolute times, i.e. a first network apparatus may have 1 o'clock while a second network apparatus may have according to an internal timer of the network apparatus 1 o'clock and 20 minutes. Phase aligned may mean when the first network apparatus has instead of 1 o'clock afterwards 1 o'clock and 1 minute, then the second network apparatus may have after one minute 1 o'clock and 21 minutes. This time shifts between a plurality of network apparatuses may occur in network based on UMTS technology.

A network resource may be a characteristic of a network, in particular of a telecommunication network. The telecommunication network may be a network based on an UMTS technology, in particular an UMTS network. A network resource may be an apparatus of a network, i.e. a RNC, a SRNC. Furthermore, a network resource may be a link in the network. A link may connect a first network apparatus with a second network apparatus. The link may be an air link, a cable link or a combination thereof. In this context a connection may be understood as a flow. In this context a link may be understood as an interface, an Iur-link, an Iub-link, a Iub or a Iur. Therefore, the term "flow" may be understood as an equivalent term to the term "connection".

A network time value may comprise a time information, in particular a time information of a network apparatus, i.e. a time stamp, meaning an information when a frame may have been sent by a network apparatus. A time value may be an absolute time value, such as a Delay Reference Time (DRT) . A DRT may be included into a data frame. The data frame may comprise a DRT field with a time information. Thus, a DRT may be a part of a data frame. A HS-DSCH frame (High Speed Downlink Shared Channel frame) may comprise a DRT.

The term "network apparatus" may comprise a Radio Network Controller (RNC) , in particular a Serving Radio Network Controller (SRNC) and a Drift Radio Network Controller (DRNC) . Furthermore, the term "network apparatus" may comprise a Base Transceiver Station (BTS) . The term "classifying or clustering" may comprise providing at least one group or class or service class, wherein the at least one group or class or service class may comprise one or a plurality of group members, which may have a common characteristic. A common characteristic may be the same source of a flow or the same transfer paths, i.e. a connection in the network.

The evaluation device may evaluate at least one class, which class may comprise a network resource. In other words, the class may indicate a group, which group may represent a characteristic of a network resource. This characteristic may also change in time. Therefore, the evaluation device of the network apparatus may receive within predetermined or random time intervals further network time values, which may be classified in addition to the former received time values into the existing classes or into one or more further classes not known in the past when receiving the first network time value. The network time value may represent a time when a HS- DSCH frame may have been sent from an RNC. The network time value may originate from a timer in the RNC.

The network time value, which may be a reference time value, may be compared to a time of a further network apparatus, i.e. to the internal time of a BTS. Thus, the network time value may be compared to a time, which may be a receiving time of the BTS. The BTS time or receiving time may indicate a time value when the network time value may have been received in the BTS. In other words, the network time of RNC and the receiving time of the BTS may be compared. The difference between a first network time value and the receiving time at the BTS may be a delay of the HS-DSCH frame

(HS-DSCH: High Speed Downlink Shared Channel) . This delay or network time value may be utilized in order to determine a characteristic of the network and to perform a classification. Thus, the delay or network time value may be utilized in order to classify a network resource. Since difference values may be utilized, i.e. due to a comparison, the absolute time of the BTS and the absolute time of the RNC may be substantially irrelevant and it may be not known by the network apparatus in order to identify a RNC, a SRNC, a DRNC a link, or a characteristic of a link, i.e. a congestion in a link.

A network time value may be an absolute time value, which may indicate a delay from a further network apparatus to the network apparatus when receiving the network time value. The term "absolute" may be understood in this context that the value may remain unchanged.

When a network apparatus may forward a network time value, this network time value may remain unchanged. A network time value may be a Delay Reference Time (DRT) . A DRT may be utilized in an UMTS network and may be sent in a HS-DSCH data frame, which data frame may be received by a BTS.

A network apparatus which may receive a flow may influence or control the further network flows upon an information in relation to network resources, i.e. from which network resource the flow may originate and/or over which link in the network the flow may have been transferred. Therefore, the network apparatus may provide a controlling in the network. Thus, the BTS may control the flows in the network after having received a network time value, in particular a DRT.

According to another exemplary embodiment of the invention, the network apparatus may be at least one apparatus selected from the group of apparatuses consisting of a base station, a base transceiver station, a Node B, an eNode B, a mobile station, a radio network controller, a drift radio network controller, a serving radio network controller, a gateway, an anchor, a switch, a hub, a server and a satellite.

In an UMTS network a processing may occur on a HS-DSCH framing protocol level, which may be utilized in a RNC and a BTS (Base Transceiver Station) . The term "Node B" may be utilized equivalently to the term "BTS".

According to a further exemplary embodiment of the invention, the network resource may be at least one network resource selected from the group of a source of a connection, a source of a congestion and a location of a congestion.

Thus, the network apparatus may identify a source of a connection or a location of a congestion or a source of a congestion. It may be also possible, that the network apparatus may identify both a source of a connection and a location of a congestion or a source of a congestion. The identification of a source of a connection and the identification of a source of a congestion or location of a congestion may be utilized in order to control the network, i.e. the network resources. The evaluating device may therefore control in a direct or indirect manner a network apparatus or a link.

The evaluating device may be adapted for classifying a source of a connection. A source of connection may be a SRNC in a network. A congestion may be detected by the network apparatus, in particular by a BTS. After determining source of a connection a source of congestion may be determined and a controlling of congestion may be provided. Such a controlling may reduce the flow of the source or stop the source sending the flow which may have caused the detected or which may cause an increased congestion of the already detected congestion.

A congestion may occur in a network. In particular, a congestion may occur in an Iur and/or in an Iub, wherein connections may utilize one or a plurality of links in the network, respectively. After determining a location of congestion a controlling of congestion, i.e. congestion control, may be provided. Such a controlling may comprise a reduction of flows in the sub-network or links, where the congestion may have been identified and localized.

A source of a connection may be a class or service class and may be classified by utilizing a received network time value, i.e. a DRT. Furthermore, a location of a congestion may be a further class or further service class and may be classified by utilizing a received network time value, i.e. a DRT.

A connection or a flow may be an Iub-connection or a Iur- connection. An Iub-connection may be a class or sub-class of a connection. An Iur-connection may be a class or a sub-class of a connection. Thus a network resource may be an Iub- connection. Furthermore a network resource may be an Iur- connection .

An Iub connection may interconnect the RNC to Node B, i.e. to the BTS. In other words, an Iub-connection may utilize an Iub-link or an Iub between a RNC and a BTS. Thus, an Iub- connection may be provided between a RNC and a BTS, in particular between a SRNC and a BTS. An Iur link may be a link between a first RNC and a second RNC, in particular between a SRNC and a DRNC. The term "Iur- connection" may be understood as a connection, which connects a first RNC and a second RNC and a BTS, in particular a SRNC and a DRNC and a BTS. In other words, an Iur-connection may connect a first RNC with a BTS over a second RNC. Therefore an Iur-connection may utilize an Iur-link and an Iub-link.

The term "SRNC" and "DRNC" may be utilized from a connection point of view. Therefore a RNC may be a DRNC of a connection and the same RNC may be a SRNC at the same time.

The evaluating device of a network apparatus, i.e. of the BTS, may be adapted for classifying an Iub-connection and an Iur-connection, i.e. providing a first group with Iub- connections and providing a second group with Iur- connections. With other words, a separation of the flows transmitted over both an Iur-link and an Iub-link as an Iur- connection and the flows transmitted only over an Iub-link as an Iub-connection may be separated.

A source of a flow may be an origin of a flow, which flow may be transmitted over one or more sub-networks or links to the network apparatus, i.e. to a BTS. A classification may provide an indication of different sources in the network, from which sources a certain flow may originate. In particular, a source may be a node, such as a RNC or a SRNC.

The term "connection" may be understood as a path of a flow, in particular of a flow over an Iur-connection using an Iur- link and an Iub-link or a flow over an Iub-connection using an Iub-link. Thus, the evaluation device of the network apparatus or the evaluating device of the network apparatus may receive different flows and may provide a classification with a first class, i.e. of Iur-connections and a second class, i.e. of Iub-connections . In this context "first class" and "second class" may distinguish two classes, wherein the terms "first" and "second" may not indicate a ranging or a preference.

According to a further exemplary embodiment of the invention the receiving device may be adapted for receiving a second network time value, wherein the received second network time value was received before the received first network time value and the second network time value may be a member of the class.

In order to classify a first network time value and a second network time value the first network time value and the second network time value may be compared. This comparison may comprise a directly comparing, i.e. comparing a first time value of the first network time value with a second time value of the second network time value or vice versa. The comparison may also be an indirect comparing, i.e. a first time value of the first network time value may be compared to a third time value and the second network time value may be compared to the same third time value or a fourth time value. A third time value and a fourth time value may originate from the network apparatus in which the evaluation device may be present, i.e. the BTS time as an time value. The evaluating device may classify the received time values of one or a plurality of flows, i.e. of downstream flows. For example it may be foreseen, that the network apparatus may receive a first network time value in a first flow and timely afterwards or at substantially the same time may receive a second network time value in a second flow. In other words, the first network time value may arrive timely before the second network time value within a same flow or within a different flow. It may also be possible that these both time values may arrive at almost substantially the same time, i.e. from different network apparatuses.

The term "flow" may be understood as equivalent term to connection .

According to an exemplary embodiment of the invention the first network time value and the second network time value may belong to a same service class.

The evaluating device may be adapted to utilize a service class for classifying.

A service class may comprise one or more flows of a certain service in the network. The service may comprise transmitting data, transmitting voice or a mixture of data and voice. In the present context relating to flows a service class may refer to a differentiation between data flows. With other words, a "service class" in this context may be understood as a subset of Iub-connections and/or Iur-connection, which may be a first service class. Furthermore another subset of Iub- connections and/or Iur-connections may be understood as a second service class. According to an aspect of the invention it may be provided to determine a delay caused by an Iur-link by subtracting the delay from Iub-link from measurement of other services in the same service class.

According to a further exemplary embodiment of the invention, classifying may comprise utilizing an interval.

The evaluating device may be adapted to utilize an interval for classifying. An interval may contain possible valid network time values and/or delay values indicating a certain network resource, i.e. a source of a flow. An interval may be calculated by the evaluation device of the network apparatus. In order to determine a minimum value and a maximum value of an interval, a network time value, such as a delay reference time (DRT) and/or a reception time (BFN) of a received data frame, received by the network apparatus, may be utilized. The interval may be a time interval.

According to a further exemplary embodiment of the invention, classifying may comprise utilizing a predetermined threshold.

A threshold may be received by the receiving device of the network apparatus. It may also be possible that the threshold may be predetermined by the network apparatus and may be stored in a storage device of the network apparatus.

According to a further exemplary embodiment of the invention, classifying may comprise utilizing a clustering algorithm.

Thus, the evaluating device may be adapted to utilize clustering for determine a location of a congestion. The evaluating device may be adapted to utilize a clustering algorithm for classifying the first network time value and/or a delay value. Furthermore the evaluating device may be adapted to utilize a clustering algorithm for classifying subsequent network time values and/or delay values, i.e. a second network time value and/or a second delay value and a third network time value and/or a third delay value. The clustering algorithm may cluster of classify the received network time values and/or the delay values. A BTS may perform such clustering algorithm.

An algorithm, in particular a clustering algorithm, may utilize a value of a delay reference time (DRT) and a reception time of a received data frame or a frame number, received by the network apparatus, i.e. by a BTS. Furthermore, a clustering algorithm in particular a k-means algorithm may be utilized. A k-means algorithm may be utilized for partitioning an N-dimensional population into k sets on the basis of a sample. When a clustering may be performed with a cluster of two groups, the k-means algorithm may be a 2-means algorithm, wherein k=2. The document "Some methods for classification and analysis of multivariate observations by J. MacQueen, In: Proceedings of 5^th Berkeley Symposium on Mathematical Statistics and Probability, 1967 may describe a process for partitioning an N-dimensional population into k sets on the basis of a sample.

According to a further exemplary embodiment of the invention, the evaluating device may be adapted for determining a time delay value.

A time delay value may be a calculated value, i.e. calculated from one or a plurality of network time values, i.e. comparing a first network time value and a second network time value and receiving a time delay value. This time delay value may be calculated by the network apparatus when receiving a flow or a connection.

According to a further exemplary embodiment of the invention, the evaluating device may be adapted for classifying the time delay value.

A delay value may be provided by comparing a first network time value and a second network time value. In other words, a first absolute value may be compared with a second absolute value and the result may be a delay time as a relative value. The delay value may be understood as a calculated time delay, i.e. by comparing two network time values or by comparing a network time value with a receiving time value. Thus, the time delay value may be a relative time value, wherein a DRT may be an absolute time value originating from a sending network apparatus. A classification of a time delay value may provide an information concerning a congestion in the network, i.e. a location of a congestion.

According to an exemplary embodiment of the invention, a method is provided for identifying a network resource. The method may comprise receiving a first network time value. The method may further comprise classifying the received first network time value into a class, wherein the class may indicate a network resource.

Thus, the method may comprise classifying at least one network resource.

A network time value may be received when a HS-DSCH data frame may arrive at the network apparatus. This HS-DSCH data frame may comprise a delay reference time set by a further network apparatus, in particular a SRNC. The network time value may be a DRT.

According to a further embodiment of the invention, the method may further comprise identifying a source of a connection in the network.

According to a further embodiment of the invention, the method may further comprise identifying a location of congestion in the network.

In an example a source of a congestion may comprise a location of a congestion in the network. In an example a source of a congestion or a location of a congestion may be a Iur-link or a Iub-link.

In order to identify a source of a congestion or a location of a congestion, stored values of already performed measurements or evaluations may be utilized. Furthermore, a source of congestion may be identified by observations from the past measurements or evaluations. A source of a congestion may be a narrow Iub link between an RNC and a BTS or a narrow Iur link between a first RNC and a second RNC. In this context "narrow" may mean that the capacity of the Iur link and/or of the Iub link may be not sufficient in order to transmit the present information or messages over the Iur- link and/or over the Iub-link, respectively.

According to an exemplary embodiment of the invention a computer readable medium may be provided, which computer readable medium may comprise program code, which program code when executed on a processor may be adapted to carry out: receiving a first network time value and classifying the received first network time value in a class, wherein the class may indicate a network resource.

A computer readable medium may be a floppy disk, a hard disk, an USB (Universal Serial Bus) storage device, a RAM (Random Access Memory) , a ROM (Read Only Memory) or an EPROM (Erasable Programmable Read Only Memory) . A computer readable medium may also be a data communication network, e.g. the Internet, which may allow downloading a program code.

According to an exemplary embodiment of the invention, a program element may be provided. The program element may comprise a program code, which program code may be adapted, when being executed on a processor to carry out: receiving a first network time value and classifying the received first network time value in a class, wherein the class indicates a network resource.

According to a further aspect of the invention a communication system may be provided. The communication system may comprise a network, wherein the network may comprise a first sub-network and a second sub-network. Moreover, the first sub-network and the second sub-network may be connected over a link and the first sub-network may be connected with a base transceiver station. Furthermore, the first sub-network may be connected with a first radio network controller and the second sub-network may be connected with a second radio network controller. Moreover, the base transceiver station may be adapted to identify a resource of the network by classifying a received first network time value into a class or service class.

A resource information may be a source of a flow and/or a connection of a flow or a source of a flow or a location of congestion or a source of a congestion. In the network the first RNS may operate as a SRNC and/or as a DRNC depending on the origin of flows.

It has to be noted that exemplary embodiments of the present invention and aspects of the invention have been described with reference to different subject-matters. In particular, some embodiments have been described with reference to apparatus type claims whereas other embodiments have been described with reference to method type claims. However, a person skilled in the art will gather from the above and the following description that unless other notified in addition to any combination between features belonging to one type of subject-matter also any combination between features relating to different subject-matters in particular between features of the apparatus claims and the features of the method claims may be considered to be disclosed with this application.

These and other aspects of the present invention will become apparent from and elucidated with reference to the embodiments described hereinafter.

Exemplary embodiments of the present invention will be described in the following with reference to the following drawings .

Brief description of the drawings

Fig. 1 shows a network apparatus in a telecommunication network according to an exemplary embodiment of the present invention .

Fig. 2 shows a network apparatus in a telecommunication network with a plurality of connections carried according to an exemplary embodiment of the invention.

Fig. 3 shows a diagram for identifying a network resource according to an exemplary embodiment of the invention.

Fig. 4 shows a chart for a method for identifying a network resource according to a further exemplary embodiment of the invention .

Fig. 5 shows a structure of a program according to an exemplary embodiment in order to perform the method according to the invention. Fig. 6 shows a structure of a program according to an exemplary embodiment in order to perform the method according to the invention.

Fig. 7 shows a structure of a program according to an exemplary embodiment in order to perform the method according to the invention.

Fig. 8 shows a structure of a program according to an exemplary embodiment in order to perform the method according to the invention.

Fig. 9 shows a method for identifying a source according to a further exemplary embodiment of the invention.

Fig. 10 shows a method in relation to a clustering algorithm according to a further exemplary embodiment of the invention.

Fig. 11 shows a method for identifying a network connection according to a further exemplary embodiment of the invention.

Detailed description

The illustration in the drawings is schematic. In different drawings, similar or identical elements are provided with the same reference numerals.

Fig. 1 shows a network 100, in particular a telecommunication network, i.e. an UMTS network, comprising a first network apparatus 200, a second network apparatus 300 and a third network apparatus 400. The first network apparatus 200 may be a BTS, wherein the BTS may comprise a receiving device 210 and an evaluating device 220. The receiving device 210 and the evaluating device 220 may be connected by a device connection 230. The receiving device 210 of the first network apparatus 200 may receive a flow or a plurality of flows. The flows may comprise time information.

The first network apparatus 200 and the second network apparatus 300 may be connected over a first sub-network 250 or an Iub-link 250, which may provide one or a plurality of Iub-connections . Furthermore, the second network apparatus 300 and the third network apparatus 400 may be connected over a second sub-network 350 or Iur-link 350, which may provide one or a plurality of Iur-connections for the first network apparatus. The first network apparatus may be a BTS, the second network apparatus may be a RNC, in particular a SRNC for Iub-connections and a DRNC for Iur-connections. The third network apparatus may be a RNC, in particular a SRNC for Iur- connections .

Fig. 2 shows an exemplary embodiment of the invention with a plurality of connections carried over the first sub-network 250 and the second sub-network 350, respectively. In Fig. 2 the network 100 of Fig. 1 may be given.

The BTS 200 may receive a first flow 260 and a second flow 270 carried over the Iub-link 250, respectively. Moreover the

BTS 200 may receive a third flow 360 which may originate from the second RNC 400, which may be a serving RNC (SRNC) . The third flow may be sent from the second RNC 400 and may be transmitted over the Iur-link 350, the first RNC 300 and the Iub-link 250. In the case of the third flow the first RNC 300 may operate as a DRNC. In the case of the first flow 260 and second flow 270 the first RNC may operate as SRNC. Thus, the BTS 200 may receive flows carried over the Iub-link 250 and may also receive in addition, flows carried over both the Iub-link 250 and the Iur-link 350. Each flow may comprise a network time value, which may be in particular a DRT in a HS-DSCH data frame.

The flows may be MAC flows, especially MAC-d flows, which flows may be carried over an Iub-link or over an Iub-link and an Iur link, e.g. after an inter-RNC handover. Fig. 2 shows an example when two connections 260, 270 are carried over the Iub-link 250 or Iub-transport only. One connection 360 may carried over both Iub-link 250 and Iur-link 350 or over Iur- Iub-transport . If a connection may be carried over both the Iur-link 350 and the Iub-link 250, the DRNC may have only forwarding functionalities.

In the following a situation with a congestion occurred in a part of the network 100 may be described.

It may be possible that both the Iur-link 250 and Iub-link 350 may be congested. The information that may be available for congestion detection at the congestion control (CC) agent located at the BTS, or at the evaluation device of the BTS, i.e. the DRT and FSN fields in the HS-DSCH frame header, may be not enough to localize the source of the congestion, i.e. to decide whether the Iub-transport or the Iur-transport may be congested. Moreover, the CC entity or CC agent or evaluating device 220 in the BTS may be not aware of the source of the connections, i.e. there may be no information about the SRNC where the connections may be originated. As there may be no information regarding neither the source of the MAC-d flows nor the location of the congestion, in case congestion may occur over Iur-transport, the CC algorithm of the evaluating device might not be able to perform congestion control and to achieve resource usage, i.e. an optimized resource usage.

Different situations may occur in relation to a congestion in the network 100. Several Iub/Iur transport congestion cases may be identified:

The situation when there may be no congestion over Iub nor over Iur may be not problematic in relation to congestion control. A congestion control (CC) action may be not needed in such a case.

A further situation may occur when the Iub-link 250 may be congested but the Iur-link 350 may be not congested. In this case already existing CC algorithms may be utilized.

In another situation both the Iub and Iur links may be congested. This situation may occur in exceptional cases and may be regarded as a theoretical case.

Another situation may occur when the Iur-link 360 may be congested but the Iub-link 250 may be not congested. In this case when the congestion may be detected by the evaluation device 220 using a CC in the BTS, the CC may erroneously decrease a shaping rate of all flows, i.e. in Figure 2 flows 260, 270, 280 or connections 260, 270, 280. This decrease may be based on a situation in which no information about the location of the congestion may be available. It may happen that in such a situation the shaping rate of the flows that are not using the Iur-link, in Fig. 2 flows 260 and 270, may be reduced in addition to a reduction of the flows using the Iub, in Fig. 2 flow 360. However, the rate reduction of flows using Iur-link might be sufficient, since a congestion may take place in the Iur-link 350. Therefore, an unnecessary rate reduction of flows not using the Iur-link 350 may occur. This rate reduction may cause an underutilization of Iub transport resources. The possibility of the detection of network resources, in particular of the location of the congestion i.e. in a Iub-link or in a Iur-link, might help a CC algorithm to react in such a situation without decreasing the shaping rate of flows not using the congested link.

The detection of the location of the congestion and the selection of the flows that are experiencing the congestion, meaning the possibility of separation of the flows carried over the congested Iur, may be an aspect of the present invention, in particular of a performant HSDPA CC solution

(HSPDA CC: High Speed Downlink Packet Access Congestion Control) . Proprietary solutions, meaning solutions provided by the network proprietary, like extending the functionality of the existing solution, i.e. an existing standard solution, may be a non-preferred option, since the Iur may be an open interface, therefore a solution that may be based on the existing functionality may be utilized.

A HSDPA congestion control algorithm, may be located in a Base Transceiver Station (BTS) and may be executed by an evaluating device of the BTS. The HSPDA congestion control algorithm may be responsible to handle the congestion situations that may happen over transport network. The CC algorithm may check the frame sequence number of the HS-DSCH data frames (FSN) , the Delay Reference Time (DRT) set by SRNC, and may measure the incoming data rate of the flows, i.e. the MAC-d flows. Based on these information, the CC algorithm may decide whether the transport network may be congested or not. In case of congestion, the CC algorithm may decrease the shaping rate of the flows and may signal these rates to the SRNC using HS-DSCH Capacity Allocation control frames. Otherwise, the CC algorithm may increase the shaping rate gradually.

One aspect of the invention may be an identification of the source of the connections in the network. A further aspect of the invention may be a determination whether the Iub or the Iur transport may be congested. These aspects may be based on the usage of a Delay Reference Time (DRT) IE (IE: Information Element) in the HS-DSCH data frame. The 16-bit long field of the DRT may contain the time (in lms resolution, 0 - 40959) when the HS-DSCH data frame was sent by the SRNC based on an internal clock of the SRNC (RNC frame number, RFN) . The CC algorithm, i.e. an evaluation device may compare the DRT to the internal timer of the BTS, in particular comparing the DRT to the BTS frame number or BFN, e.g. the time when the given HS-DSCH data frame was received (T) by the BTS. The delay of the HS-DSCH data frame may be calculated with the formula

FP_delay = T - DRT

wherein FP delay may mean frame protocol. It may be assumed that there may be only one service class in the transport network. If a plurality of service classes may be present, the proposed methods and network apparatuses according to the invention may be evaluated on service class basis.

In UTRAN a common timing reference among all the nodes may not required. Different nodes' counters (RFN and BFN, RNF: RNC Reference Number, BNF: Node B frame number), even if frequency-locked to the same network synchronization reference, may be not phased aligned. Therefore, the DRT values of HS-DSCH data frames sent by different RNCs may be in different ranges, so in the BTS the DRT values may be used for identifying the RNC where the given and received HS-DSCH data frame come from.

Assuming that the Iur link may be congested but the Iub may be not congested and the delay of the HS-DSCH data frames may be calculated for each active connection. If a connection may go over the Iub link only, its delay or a delay value FP_delay may be formulated as:

FP_delay(Iub)= D (Iub) + X,

wherein D (Iub) may be the waiting time in the Iub transport buffers and X may be an intrinsic delay component due to discrete MAC-d shaping, media delay, forwarding delay etc. If a connection may be carried over both Iub and Iur transport, delay of the flow over this connection or a delay value FPdeiay may be formulated as:

FP_delay (Iur, Iub) = D (Iur) + D (Iub) + Y,

wherein D (Iur) may be the waiting time in the Iur transport buffers and Y may be an intrinsic delay component due to discrete MAC-d shaping, media delay, forwarding delay etc. When Iur is congested, the D (Iur) may be significantly larger than 0, in fact D (Iur) may be the dominant component of the end-to-end delay of the connections that are carried over both Iur and Iub. Therefore the delay or delay value may be used as a basis of differentiation between connections that are carried over the congested Iur transport and those that are not.

The separation of the connections may be relevant if the Iur links are congested. Therefore, this method may be suitable for identifying the Iur connections when this information may be relevant for the CC and in addition, the location of the congestion may be identified in addition. In other words, being able to separate the connections may indicate a congestion over the Iur-link. The case when the connections may not be separated but a delay increase or a loss may be detected, the delay increase or the loss may indicate congestion over Iub. However, the latter situation may be treated by a CC algorithm, so in application point of view it may not be a problem. According to an aspect of the invention a service class based congestion control may be provided.

The utilization of a DRT and the utilization of a delay measurement may be used together to verify or complement each other. In other words, a detection of a source of connection may be provided and in addition a location of a congestion may be provided. This information may characterize the network, i.e. a characterization of a situation in a network in respect to flows. A DRT may be utilized in order to separate the Iur and Iub flows from each other. Utilizing a delay measurement - in addition of being able to separate the Iur connections - may be also able to localize the congestion (Iur or Iub transport) as when the Iur links are congested the connections may be separated on delay basis.

Fig. 3 shows a method or an operation of the algorithm in order to identify a network resource, in particular a source of a flow. At a bottom line 920 a time in the RNC is shown. At an upper line 921 a time in the BTS is shown. A value of the DRT IE 910, indicated by a circle on a bottom line and the reception time T 911, indicated by a circle on top line, of the latest HS-DSCH data frame may be stored for each RNC detected earlier. The reception time "T" may be provided by the BTS, wherein T may originate from a BFN. A BNF may represent the time in the BTS. When a new HS-DSCH data frame may be received at the time T, indicated as circle 912 on top line, an interval 913 containing possible valid DRT values, indicated as interval on the bottom line, may be calculated.

The valid DRT interval 913 may be greater than the last stored value of the DRT IE 910, because the transport network may provide in-sequence delivery, but not lossless for HS- DSCH data frames. If the DRT value of this new HS-DSCH data frame, shown as a circles 914, 915 on bottom line, may be in the valid DRT interval 913, the HS-DSCH data frame and its connection may belong to this RNC. If the new HS-DSCH data frame may be not in the valid DRT interval, as shown for circle 915, then the step of evaluation may be performed for the next RNC. If there may be no more RNC, it may be a suitable assumption that this HS-DSCH data frame may belong to a new RNC, not know yet by the BTS.

It may be proposed an implementation in order to identify a network resource, in particular a source in the following way:

In Fig. 4 a method in order to identify a network resource may comprise the following steps. In a first step S401 several definitions in relation to parameters and variables may be provided.

"IurDRTDetectionThreshold" may be a parameter. The IurDRTDetectionThreshold parameter may indicate the minimum difference between the internal clocks of different RNCs. A default value of the IurDRTDetectionThreshold parameter may be 500 milliseconds. Furthermore, "T" and "DRT" may be input variables. The parameter T may indicate the current time, which current time may be based on the internal timer of the BTS, The parameter T may be indicated in milliseconds. The DRT parameter may be a value of a DRT field of the given HS-DSCH data frame. The parameter DRT may be indicated in milliseconds.

Moreover, internal variables "N", "lastDRT" and "last" may be utilized. The variable N may indicate the number of different and detected RNCs having at least one active connection to the given BTS.

The variable lastDRT (i) with i=l..N, may indicate the value of DRT IE of the last HS-DSCH data frame belonging to RNC with id i (id = identification or index) . The variable lastT(i) with i=l..N may indicate the time when the last HS-DSCH data frame belongs to RNC with id i was received.

Furthermore, as an output the variable "j" may be used. The output "j" may be used in a RNC id (j), where the HS-DSCH data frame come from, i.e. the connection belongs to RNC with id j .

When a HS-DSCH data frame may arrive to the BTS, the following steps as steps S402 and S403 may be performed:

In step S402 a loop may be performed wherein a calculation for a valid DRT interval may be provided. The loop may be performed N times and may comprise several operations or sub- steps, which sub-steps may comprise one or more comparisons and one or more calculations of a time value. The step S403 may be performed in case when the DRT may not be classified within one class, for example in the class of a valid DRT interval. In step S403 a new, not already known RNC may be considered as a source of the received DRT value.

In Fig. 5 an algorithm is shown according one exemplary embodiment for a better understanding of the steps S401, S402 and S403 of Fig. 4.

The DRT IE may be a 16-bit long field (0-40959) . Therefore a wrap around functionality may be provided, in order to be not limited to a maximum number of 40959 possibilities. A transformation may be provided by using modulo 40960 at least one step or for all steps using the basic operations, i.e. addition, differentiation, etc.. The comparison step (beginning with if DRT is greater than a minValidDRT value and the DRT value smaller or equal to a maxValidDRT value) described above in step 403 may be replaced by the following step which may treat the wrap around case separately from the non-wrap around case:

In Fig. 6 a wrap around functionality may be shown as algorithm according to an exemplary embodiment.

The wrap around case may operate as follows: first a check is performed whether the minValidDRT value is smaller than the maxValidDRT. If this is true, it may mean that the valid interval may not be wrapped around, thus the check, whether the DRT value is in the valid interval, may be performed as in the non-wrap around case. (i.e. DRT is greater than a minValidDRT value and the DRT value smaller or equal to a maxValidDRT value) . In the case when the minValidDRT value may be greater than the maxValidDRT, it may mean that the valid interval is wrapped around, i.e. the valid interval may consist of two parts: an interval which begins at minValidDRT and finishes at 40959 and an interval which begins at 0 and which finishes at maxValidDRT . Therefore, the check may be performed for these intervals, which can be simplified to the check when DRT value is smaller than maxValidDRT or DRT value is greater than maxValidDRT.

The above proposed method may depend on the value of IurDRTDetectionThreshold which may be eliminated in the following way. Instead of using the meanValidDRT, a calculation of the minValidDRT and of the maxValidDRT may be performed as follows:

minValidDRT (i) = lastDRT(i) maxValidDRT (i) = infinite

The simplified case may operate as follows: it checks only whether the DRT value is greater than the minValidDRT or not. The value of maxValidDRT may be irrelevant. If it is greater it means DRT value is in the valid interval, otherwise it is not .

In another words, instead of calculating the valid interval, the method may detect the permuted DRTs. For instance, if a later sent HS-DSCH data frame may arrive earlier than another HS-DSCH data frame, then it may indicate that these two HS- DSCH data frames are from different RNCs. lastDRT(i) + 40960 / 2 may be interpreted as infinite in case of wrap around. This simplified proposal may make a mistake with a higher probability compared to the threshold based method, but simulation may show in Fig. 9 that after a learning period the method may work with a high success ratio.

The described implementation may be adapted in several ways, for instance: Using information provided by not only the last but some older HS-DSCH data frames, i.e. meanValidDRT (i) may be an average .

The stored information (lastDRT(i) , lastT(i)) may be inaccurate if there is no data for a longer period. In this case, there may be no reason to compare the newly received HS-DSCH data frame with the old information, thus this RNC may be skipped. This period may be inactivity time of the CC or the HS-DSCH channel down switch time.

If the SRNC of a given connection is identified once, there may be no reason to recalculate it anymore, because it may be not possible that the SRNC may be changed for an active connection. One exception may be the SRNS relocation procedure. In this case, if the HS-DSCH framing protocol entity may be recreated in the BTS, the connection may be treated as a new connection. Hence, the proposed algorithm may be used for new connections only and the resulting RNC may be used for further decisions.

In the proposed implementation in order to identify a network resource, in particular a connection, a clustering algorithm, in particular a clustering algorithm called k-means algorithm may be used. However the k-means algorithm may be replaced by any clustering algorithm.

The latest network time value or time delay value or delay value of each connection may be stored and may be used by a clustering algorithm, wherein each connection may be evaluated upon a reception of a new HS-DSCH data frame. Assuming that the clustering algorithm may have found two delay clusters, it may be not known whether the clusters may represent two different links or not. According to an aspect of the invention there may be utilized a differentiation of the connections using an Iub-link only or an Iub-link and an Iur-link, in particular in a situation when the Iur-link may be congested. In relation to a CC algorithm point it may mean that a frame delay (FP_delay) may be greater than a threshold (Thr) :

FP_dei_ay > Thr,

wherein "Thr" may a predefined threshold. In another words, if the Iur may be congested, then the congestion may increase the delay at least with Thr. Therefore, if the difference between the delay over the Iur-link and the delay over the Iub-link may be utilized as a cluster, respectively. If the two clusters may be larger than Thr, it may mean one of the clusters, having greater average delay, may represent Iur connections. If the two clusters may be smaller than Thr or equal to Thr, it may mean the two clusters may represent the Iub link only, or it may represent both the Iub and the Iur connections, but the Iur may be not congested, so no CC action may be needed. The results of the k-means algorithm may be stored, i.e. as Ri and R₂. Furthermore Ri and R2 may be used for the input of the next run of the k-means algorithm, since only one delay value may be changed between the two runs. This may provide a faster run of the k-means algorithm, however two random numbers between the minimum and the maximum of the measured delay may be also sufficient for Ri and R2.

Further aspects of the k-means algorithm may be the following : The parameters k may indicate a number of clusters, which clusters the algorithm may find.

An input variables may be N, indicating a number of active connections in the BTS and

R₁ with 1=1... k, and R₁ indicating an initial mean points of the algorithm. Furthermore an input variable may be D₁ with J=I...N, and D₁ indicating an estimated delay of the HS-DSCH data frame of the j.-th (active) connection, given in milliseconds.

Internal variables may comprise C₁ with 1=1... k, wherein C₁ may indicate clusters containing the index of connections which is belonging to C₁ . For instance if k may be in C₁₁ = [I₁, ±2 ,...}, it may mean the D_k may belong to cluster C₁.

Furthermore an output may comprise C, wherein C may indicate the new clusters, i.e. C=(Ci, ••• , C_k} .

Fig. 7 shows an algorithm with the parameters and variables k, N, Ri, Di, Ci, C and Index i and j . In Fig. 7 the following steps may be performed:

The method may be performed as follows: In first step the evaluating device may calculate the average of each cluster. Then for each connection the evaluating device may search that cluster which has the closest average to the given connection, i.e. the differences between the delay of the connection and the average of the cluster is less (or equals) compared to the difference from other cluster's average. Then, the algorithm may recalculate the averages (R₁) . If the averages do not changed (i.e. the clusters do not changed) compared to the previous averages then the algorithm returns with the clusters (C=(Ci, ••• , C_k}) . Otherwise, the algorithm may start the whole process or method from the beginning.

The expression "argmin" may be understood in a mathematical sense and may stand for the argument of the minimum, that is to the value of the given argument for which the value of the given expression may attain its minimum value.

In order to identify a network resource, in particular a connection the following steps may be performed:

Parameters may comprise "Thr" , wherein Thr may indicate a redefined delay threshold indicating delay based congestion for CC algorithm. A default value of Thr may be 100 milliseconds.

An input variable may comprise N, wherein N may indicate a number of active connections in the BTS. A further input variable may comprise R₁ with 1=1...2, and R₁ may indicate mean points of the algorithm. A further input variable may comprise D₁ with j=l...N, and D₁ may indicate an estimated delay of the HS-DSCH data frame of the j.-th active connection, and D₁ given in milliseconds.

A internal variable may comprise C₁ with 1=1...2 and C₁ indicating clusters containing the connections number of different and detected RNCs having active at least one connections to the given BTS. For instance if k may be in C₁₁ = [I₁, I2 ,...}, it may mean the D_k may belong to cluster C₁.

An output may comprise a parameter "possiblelurConnections" and may indicate the connection id-s using Iur-link, i.e. possiblelurConnections ={ii, i₂ ,...}. The following steps may be performed or executed when a HS- DSCH frame may arrive to a BTS, which may be also shown in Fig. 8.

In a first step the parameter "possiblelurConnections" may be indicated. In the first step, a check may be performed in order to determine how many active connections may be available. If the number of active connections is greater or equal than two, the next step may be evaluated. Otherwise, when it is assumed that only one connection may be present and the clustering may not be successful. Therefore the procedure may be finished.

Otherwise, if there may be two or more active connections a further step may be performed as a second step. In other words, if N is equal or greater than 2, than a clustering may be performed. In the second step a loop or a repetition may be performed of several sub-steps.

In the second step, the 2-means algorithm may be used with the given parameters. The output may be two clusters which might represent an Iur-cluster or Iur-service class and an Iub clusters or Iub-service class.

The Iur-cluster and the Iub-cluster may be checked in a third step. If the difference between the average delay of the clusters is larger than the threshold, the two clusters may represent an Iub cluster and an Iur cluster. Otherwise, the output clusters of the 2-means algorithm may not represent two clusters but only one cluster.

It may be possible that a connection may change its cluster, because the differentiation may be possible when the Iur may be congested. Therefore, when the Iur starts to be congested, the algorithm may classify some connections into the Iur classes or Iur cluster, which may have been treated as Iub only connections earlier. After a CC action may be performed, the congestion situation on Iub may disappear, so the algorithm may result in one cluster. In this case the Iur may be still used. This may be an inaccuracy, but may be not a problem in relation to the CC algorithm point of view.

In general, there may be some connections which may be carried over a Iub-link only, i.e. Iub-connections, and there may be some other connections which may be carried over an Iur-link and over an Iub-link, i.e. Iur-connections .

It may occur, that in this situation the method may provide only one cluster, although there might be two clusters in fact. This result could occur when the difference between the average delays of the two clusters may be less than the threshold. However, this situation may be not a problem, because if the Iur link is congested the delay would be higher. Therefore, the Iur may be not congested, so no shaping rate decrease, may be needed, which shaping rate decrease may be performed by the CC agent. Thus, from the point of view of the CC agent, this result of the method may be not an issue and no further steps may be taken in order to control the network.

There may be provided an inactivity check. In the inactivity check a parameter, called i.e. thresinactivity may be utilized. The inactivity check may be performed if there may be no HS-DSCH data frame received from a connection for a long period, e.g. 160 milliseconds (thresinactivity) . The period of non receiving a data frame, i.e. no HS-DSCH data frame, may mislead the clustering algorithm. Since the maximum possible value of the HS-DSCH credit interval may be 80 milliseconds, the latest delay value might be invalid. Therefore, the old delay values and the connections related to them may be excluded from the further calculation. The excluded connections might be considered when a new HS-DSCH data frame belonging to the excluded connections may be received.

Furthermore, there may be provided a delay measurement correction. It may be assumed, that a delay of the HS-DSCH data frames (D₁ , ... ,D_N) may be calculated for each active connection. However, if the timers in the SRNCs are shifted means not phased aligned, it may be not straightforward how to obtain a delay of the HS-DSCH data frames. It may be foreseen to store the minimum measured HS-DSCH data frame delay per RNC, which might have been earlier detected. Furthermore, it may be foreseen to decrease the actual measured HS-DSCH data frame delay of all connection belonging to the given RNC with the minimum measured HS-DSCH data frame delay per RNC. This stored minimum delay may tend to the time offset discrepancy between the SRNC and the RBS internal timers plus the intrinsic delay component. An intrinsic component may caused by a propagation delay, a media delay and/or a forwarding delay, etc.. Decreasing the actual measured HS-DSCH data frame delay with this minimum estimation may allow to the CC algorithm to compare the delays of the connections from different SRNCs.

In addition, a generalization in order to find more than two clusters may be provided. Using the proposed implementation the result may be in the most cases two clusters. However, real topologies could be more complex, i.e. more than one RNC may be connected to the DRNC via the Iur interface. It may be foreseen that when the algorithm may have classified the connections into two clusters, a further run of the clustering algorithm may be provided for these two clusters (for C₁ and C2 if I -Ri - -R₂ I > Thr) separately with a lower threshold (e.g. Thr / 2) .

There may be i.e. three possible results:

The further run of clustering algorithm may not have found any sub-clusters of the original ones, so only two clusters (Ci and C₂) may be present.

The further run of clustering algorithm may classify the input clusters into two-two sub-clusters, so four different clusters (Cn, Ci₂ and C21, C22) may be provided. It is also a possible case when one further run of clustering algorithm may find two sub-clusters but the other further run may have not found any sub-cluster. Then there may be three clusters (Cn, Ci2 and C₂, or Ci and C21, C₂₂) •

In the second case and the third case the clustering algorithm may be run again for the newly found sub-clusters with an exponentially decreasing threshold as long as any further sub-clusters may be found. After (and in the first case) no further action may be needed. The found of sub (and original) clusters may represent the different Iur-links and Iub-links .

Fig. 9 shows a method in order to classify or cluster a source as network resource. In a first step S501 definitions may be performed. A parameter Iur DRT detection threshold may be defined as a minimum difference between the signal clocks of different RNCs wherein a default value may be 500 milliseconds. Furthermore input variables T and DRT may be defined wherein T may be a current time based on an internal timer of the BTS and measured in milliseconds. The input J

variable DRT may be a value of a DRT field of the given HS- DSCH data frame, measured in milliseconds. Furthermore, internal variables may be defined which may comprise N, last DRT and last T. An N may be a number of different and detected RNCs, having at least one active connection to the given BTS. According to an aspect of the invention a DRT value may be used to estimate the RNC by identifying DRT clusters .

The last DRT may be an internal variable depending on i, wherein i may be equal to one N, which may be a number of different and detected RNCs, having at least one active connection to the given BTS. The last DRT variable may indicate a value of DRT IE of the last HS-DSCH data frame belonging to RNC with ID i. The internal variable last T, in particular last(i) may indicate the time when the last HS- DSCH data frame may belong to RNC with ID i may have been received, wherein i = 1...N. Furthermore an output may be defined as j, wherein the RNC ID(j) may be utilized. The RNC ID(j) may indicate where the HS-DSCH data frame may come from, i.e. the connection may belong to RNC with ID j.

In a second step S502 a calculation of a valid DRT interval may be performed by utilizing a loop procedure, with i = 1 to N. The DRT interval may be calculated by using a current DRT, a last received DRT and a receiving time of the BTS. Furthermore a DRT threshold may be utilized in order to calculate or to determine a minimum valid DRT and a maximum valid DRT. If the current received DRT is higher than the minimum valid DRT and the current received DRT is smaller or equal the valid DRT then the received frame may be belong to the RNC with ID i. The internal variables N, last DRT and last T may be updated, meaning last DRT (i) may be DRT and last T(i) may be equal to T. The loop may be performed until a valid RNC ID may be found. If no valid RNC ID may be found, then a new RNC may be needed. This new RNC may be determined in a step S503.

In Fig. 10 a further exemplary embodiment of the method is given. In a step S601 the parameter k may be defined, as well as the output variables N, R₁, and D₁. Furthermore, in step S601 the internal variable C₁ and an output C may be defined. In a step S602 a clustering may be performed.

In Fig. 11 a method is shown in order to provide a network resource connection. In a step S701 parameters, input variables and internal variables may be defined. Furthermore an output may be defined in step S701, wherein the output may be a "possible Iur-Connections" parameter. In a step S702 possible Iur-Connections may be determined wherein the input variable Rl and the input variable R2 may be calculated as a function of D and Cl, respectively. In a comparison of Rl and R2 in a step S703 a possible Iur-Connection may be determined. If Rl is greater than R2 then the parameter "possible Iur-Connections" is equal to Cl, meaning that the connection may be a connection using an Iur-Connection. Otherwise if Rl ≤ R2, meaning the value of Rl is smaller or equal the value of R2, then the "possible Iur-Connections" parameter may be set equal to C2, meaning that the connection may be an over Iub-connection . It should be noted that the term "comprising" does not exclude other elements or steps and the "a" or "an" does not exclude a plurality. Also elements described in association with different embodiments may be combined.

It should also be noted that reference signs in the claims shall not be construed as limiting the scope of the claims.

Acronyms and Terminology

BFN BTS Frame Number

BTS Base Transceiver Station

CA Capacity Allocation

CC Congestion Control

DRNC Drift RNC

DRT Delay Reference Time

FSN Frame Sequence Number

FP Frame Protocol

HS-DSCH High Speed Downlink Shared Channel

HSDPA High Speed Downlink Packet Access

IE Information Element

MAC Medium Access Control

MAC-d dedicated MAC

RFN RNC reference number

BFN Node B frame number

RNC Radio Network Controller

SRNC Serving RNC

UMTS Universal Mobile Telecommunications System

UTRAN UMTS Terrestrial Radio Access Network,

Claims

Claims :

1. Network apparatus (200) for identifying a network resource (250, 300, 350, 400) in a network (100), the network apparatus comprises: a receiving device (210); an evaluating device (220); wherein the receiving device (210) is adapted for receiving a first network time value; wherein the evaluating device (220) is adapted for classifying the received first network time value into a class; wherein the class indicates a network resource (250, 300,

350, 400) .

2. Network apparatus according to claim 1, wherein the network apparatus (100) is at least one apparatus selected from the group of apparatuses consisting of a base station, a base transceiver station, a Node B, an eNode B, a mobile station, a radio network controller, a drift radio network controller, a serving radio network controller, a gateway, an anchor, a switch, a hub, a server and a satellite.

3. Network apparatus according to claim 1 or 2, wherein the network resource (250, 300, 350, 400) is at least one network resource selected from the group of network resources consisting of a source of a connection, a source of a congestion and a location of a congestion.

4. Network apparatus according to one of the claims 1 to 3, wherein the receiving device (210) is adapted for receiving a second network time value; wherein the received second network time value is received before the first network time value; and wherein the second network time value is a member of the class .

5. Network apparatus according to claim 4, wherein the first network time value and the second network time value belong to a same service class.

6. Network apparatus according to one of the claims 1 to 5, wherein classifying comprises utilizing an interval.

7. Network apparatus according to one of the claims 1 to 6, wherein classifying comprises utilizing a predetermined threshold.

8. Network apparatus according to one of the claims 1 to 7, wherein classifying comprises utilizing a clustering algorithm.

9. Network apparatus according to one of the claims 1 to 8, wherein the evaluating device (220) is adapted for determining a time delay value.

10. Network apparatus according to claim 9, wherein the evaluating device (220) is adapted for classifying the time delay value.

11. Method for identifying a network resource (250, 300, 350, 400) in a network (100), the method comprises: receiving a first network time value; classifying the received first network time value into a class, wherein the class indicates a network resource (250, 300, 350, 400) .

12. Method according to claim 11, wherein the method further comprises : identifying a source of a connection in the network.

13. Method according to claim 11 or 12, wherein the method further comprises: identifying a location of congestion in the network (100) .

14. Computer readable medium, which comprises program code, which program code when executed on a processor is adapted to carry out : receiving a first network time value; classifying the received first network time value into a class, wherein the class indicates a network resource (250, 300, 350, 400) .

15. Program element comprising a program code, which program code is adapted, when being executed on a processor to carry out : receiving a first network time value; classifying the received first network time value into a class, wherein the class indicates a network resource

(250, 300, 350, 400) .