• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    method for performing a measurement on a data stream, and communication network. A method for performing a measurement (namely, a data loss measurement and/or a time measurement) on a data stream to be transmitted over a communication network is described. the method comprises: by transmitting the data stream, during the first block periods that alternate in time with the second block periods, marking each data stream data unit by setting a resource thereof to a first value and updating of a first parameter; upon receipt of the data stream, check the resource for each data unit received and update a second parameter when the resource equals the first value; while transmission and reception are carried out, processing a timer to determine whether a current block period is one of the second block periods and, if so, providing current values of the first and second parameters; and perform a measurement using the current values of the first and second parameters.
    公开号:BR112013010739B1
    申请号:R112013010739-1
    申请日:2010-11-05
    公开日:2021-08-17
    发明作者:Maurizio Cociglio；Luca Castaldelli
    申请人:Telecom Italia S.P.A.；
    IPC主号:
    专利说明:

    Field of Invention
    [0001] The present invention refers to the field of communication networks. In particular, the present invention relates to a method for performing a measurement, in particular, a data loss measurement and/or a time measurement (in particular, an interarrival delay and/or jitter measurement), in a stream. data transmitted over a communication network. Additionally, the present invention relates to a communication network that implements such a method. Fundamentals of the Invention
    [0002] In a packet-switched communication network, data is transmitted in the form of packets that are routed from a source node to a destination node through possible intermediate nodes. Exemplary packet-switched networks are Local Area Networks (eg Ethernet) and Geographical Area Networks (eg Internet).
    [0003] On the other hand, in circuit-switched networks, data is transmitted in the form of continuous bit streams conducted from the source node to the destination node in plesiochronous or synchronous frames. Exemplary circuit-switched networks are PDH, SDH, Sonet, and OTN networks.
    [0004] Next, the expression “data unit” will designate a part of a data stream transmitted over a communication network. In particular, in the case of a packet-switched network, a unit of data can be a packet or a part of a packet. Furthermore, in the case of a circuit switched network, a data unit can be a plesiochronous frame, a part of a plesiochronous frame, a synchronous frame or a part of a synchronous frame.
    [0005] Data transmitted in both a packet-switched network and a circuit-switched network from a source node do not always reach the destination node, that is, they can get lost during transmission through the network.
    [0006] Data loss can be due to different reasons. For example, in a packet-switched network, data in a packet may be dropped by an intermediate node, as the port on which the packet is received or through which the packet needs to be forwarded is congested. Furthermore, in both packet-switched and circuit-switched networks, data can be discarded by an intermediate node or by the destination node, as they contain bit errors.
    [0007] During the provision of a service by data transmission over a packet-switched network or a circuit-switched network, the rate of data lost during transmission affects the quality of service (QoS) of this service.
    [0008] In addition, a data unit is transmitted in one transmission time by the source node and is received in one reception time by the destination node. The time that elapses between transmission time and reception time is typically called “one-way delay” (or, in short, “delay”). The delay of a data unit is given by the following equation:
    where Si is the transmission time and Ri is the data unit reception time.
    [0009] The delay of a data unit depends mainly on the number of possible intermediate nodes crossed by the data unit from source to destination and the dwell time of the data unit at the source node and at each possible intermediate node. In packet-switched networks, where data units are routed hop-by-hop through each node, both the number of possible intermediate nodes crossed by data units and the dwell time of data units at each node are unpredictable. In this way, the delay of a data unit is almost unpredictable. Furthermore, data units from the same data stream can have different delays.
    [00010] In a packet-switched communication network, the difference in delays of two data units (ie packets) of the same data stream is called “inter-arrival instability”. In particular, if Si and Sj are the transmission times for a first packet i and a second packet j, and Ri and Rj are the receive times for the first packet i and second packet j, the interarrival instability can be expressed as:

    [00011] When a communication service (in particular, a real-time voice or data service, such as calls, conference calls, videoconferences, etc.), is provided through a communication network, the delay and the Interarrival instability of the data flows that carry the service strongly affect the quality of service (QoS) perceived by the end users of the service.
    [00012] Therefore, the measurement of data loss and interarrival delay/instability of data flows that carry services is of particular interest to network operators.
    [00013] WO 2010/072251 (on behalf of the same Applicant) describes a method for measuring the data loss of a data stream transmitted through a communication network from a transmitting node to a receiving node. Before transmitting the data units of the data stream, the transmitting node marks each data unit to divide the data stream into blocks. In particular, the transmitting node marks each data unit by setting one bit of its header to “1” or “0”. Marking results in a sequence of blocks, in which blocks of data units marked with “1” alternate temporally with blocks of data units marked with “0”. Blocks can have the same duration called “block period” Tb. Additionally, when marking data units, the transmitting node increases by a first counter C1 every time a data unit is marked by "1" and a second counter C0 every time a data unit is marked by "0" . Then the tagged data units are received at the receiving node. Every time the receiving node receives a data unit, it checks its tag, increases a third counter C’1 if the tag is “1” and increases a fourth counter C’0 if the tag is “0”.
    [00014] According to WO 2010/072251, while the transmitting and receiving nodes are operating as stated, a management server that cooperates with the transmitting and receiving nodes periodically detects the values of counters C1, C0, C'1 and C' 0, and uses them to calculate data loss. The detection period Td has a maximum value that depends on the block period Tb according to the following formula: Tb > 2 * Td. In other words, the values of counters C1, C0, C’1 and C’0 must be detected at least twice in each block period. Since Tb equals 5 minutes, Td can equal 2 minutes.
    [00015] At each detection time Td, the management server compares the current detected values of C1, C0, C'1 and C'0 with the corresponding previously detected values to determine which of the C1, C0 counters and which of the C counters '1, C'0 have a fixed value during the current block period. In fact, from the above-described marking procedure, it follows that, in each block period, one of the C1, C0 counters has a fixed value, while the other is increasing, and one of the C'1, C'0 counters has a fixed value while the other is increasing. Whereas the management server determines that, during the current block period, counters C0 and C'0 have a fixed value, the management server calculates data loss as a difference between the currently fixed values of counters C0 and C '0. Invention Summary
    [00016] The need to improve the solution described by WO 2010/072251 was perceived.
    [00017] In fact, the computational effort and bandwidth usage required on the management server to support the exposed solution, in some cases, can be very high.
    [00018] Additionally, in case the transmitting node does not transmit any data unit to the receiving node for a block period (this can happen when the traffic between the transmitting node and the receiving node is very low), all C1 counters , C0, C'1 and C'0 have a fixed value. In this way, the management server may encounter some difficulties in determining which pair of counters (ie, the C1-C'1 pair or the C0-C'0 pair) should be used to properly calculate the data loss in relation to to the preceding block period.
    [00019] In view of the above, the problem of providing a method to perform a measurement on a data stream transmitted from a first communication device to a second communication device of a communication network that is more efficient in terms of from a computational point of view and from the point of view of the use of network bandwidth and which allows the measurement to be carried out even after no data unit has been transmitted during a block period.
    [00020] In the following description and in the claims, the expression "perform a measurement on a data stream" shall designate an operation for measuring the loss of data on a data stream and/or an operation for performing a time measurement on a data flow.
    [00021] Additionally, in the following description and in the claims, the expression "data loss measurement" shall designate an operation of measuring a difference between numerous data units (i.e., packets, packet parts, plesiochronous frames, frame parts plesiochronous frames, synchronous frames, synchronous frame parts) transmitted by a first communication device and numerous data units (i.e., packets, packet parts, plesiochronous frames, plesiochronous frame parts, synchronous frames, synchronous frame parts) received in a second communication device, this difference corresponding to the number of data units lost in transmission from the first communication device to the second communication device.
    [00022] Additionally, in the following description and in the claims, the expression "performing a time measurement on a data stream" shall designate a measurement operation: - a delay induced in a data unit of the data stream by the transmission between the first communication device and second communication device; and/or - an interarrival instability induced in a pair of data units of the data stream by transmission between the first communication device and the second communication device.
    [00023] Additionally, in the present description and in the claims, the expression "communication device" will designate a port or interface of a node of a communication network.
    [00024] The first communication device and the second communication device can be composed in different nodes of the communication network. For example, the first communication device can be an output port of a first node that transmits the data stream, while the second communication device can be an input port of a second node that receives the data stream. The first node and second node can be the source node and target node of the data flow, or alternatively they can be intermediate nodes placed between the source node and target node. The first node and the second node can either be physically adjacent (that is, they are connected by a physical link, such as, for example, an optical fiber) or they can be connected through other nodes.
    [00025] Alternatively, the first communication device and the second communication device can be composed in the same node. For example, the first communication device can be an input port of a node that receives the data stream from an upstream node, while the second communication device can be an output port of the same node that forwards the data flow to a downstream node. Data loss between the first communication device and the second communication device, in this case, indicates the number of data units lost at the node (for example, discarded due to congestion). Additionally, the delay between the first communication device and the second communication device indicates the residence time of the lost data units in the node.
    [00026] In addition, in the following description and in the claims, the expression "marking a data unit" shall designate an operation of setting a data unit resource to a value suitable for distinguishing the data unit from other data units of the same data stream. For example, the operation of marking a data unit may comprise the operation of setting one or more bits of the data unit (for example, a bit or a sequence of bits of its header) to a predefined value, a operation of setting its frequency or its phase to a predefined value and the like.
    [00027] According to a first aspect, the present invention provides a method for performing a measurement on a data stream to be transmitted from a first communication device to a second communication device of a communication network, the method comprising: a) upon transmission of the data stream: - during the first block periods that alternate in time with the second block periods, mark each data unit of the data stream by setting a data unit resource to a first value and updating a first parameter with respect to data units; b) upon receipt of the data stream: - check the resource for each data unit received and update a second parameter in relation to the data units when the resource is equal to the first value; c) while steps a) and b) are carried out, processing a timer to determine whether a current block period is one of the second block periods and, if so, providing a current value of the first parameter and a current value of the second parameter; and d) performing a measurement on the data stream using the current value of the first parameter and the current value of the second parameter.
    [00028] According to first advantageous modalities: - in step a), the update comprises increasing a first transmission counter by transmitting each data unit with the resource defined in the first value; - in step b), the update comprises increasing a first reception counter upon receipt of each data unit with the resource defined in the first value; - in step c), the provision comprises providing a current value of the first transmission counter and a current value of the first reception counter; and - in step d), performing a measurement comprises calculating a data loss as a difference between the current value of the first transmission counter and the current value of the first reception counter.
    [00029] Furthermore or alternatively: - in step a), the updating comprises updating a first transmission time record by transmitting a predetermined data unit with the resource defined in the first value; - in step b), the updating comprises defining a first reception time record upon receipt of the predetermined data unit; - in step c), the provision comprises providing a current value of the first transmission time register and a current value of the first reception time register; and - in step d), performing a measurement comprises performing a time measurement using the current value of the first transmission time record and the current value of the first reception time record.
    [00030] Preferably, performing a time measurement comprises measuring at least one of an interarrived delay and instability of the packet flow.
    [00031] Preferably, all the first block periods and the second block periods have the same duration Tb.
    [00032] Preferably, in step c), the timer counts a marking period Tm equal to an even integer multiple of the duration Tb.
    [00033] According to first advantageous variants, step c) comprises comparing a current time indicated by the timer with a marking timing table comprising information on start time and/or end time of at least one of the first periods of block and the second block periods.
    [00034] Alternatively, if the duration Tb is equal to an odd number of time measurement units, step c) comprises determining whether a current time indicated by the timer and expressed in the time measurement units is a multiple of the duration Tb and is an even number.
    [00035] Alternatively, if duration Tb is equal to an even number of time measurement units, step c) comprises determining whether a current time indicated by the timer and expressed in time measurement units is not an integer multiple of the duration Tb multiplied by 2.
    [00036] Preferably: - step a) further comprises, during the second block periods, marking each data unit of the data stream by setting the data unit resource to a second value and updating a third parameter in relation to the units of data; - step b) further comprises updating a fourth parameter with respect to data units when the resource is equal to the second value; - step c) comprises, if negative, providing a current value of the third parameter and a current value of the fourth parameter; and - step d) comprises performing the measurement using the current value of the third parameter and the current value of the fourth parameter.
    [00037] According to particularly preferred embodiments, step c) is performed by the first communication device and also by the second communication device, while step d) is performed by a management server that cooperates with the communication network .
    [00038] In this case, step c) preferably comprises: - in the first communication device, processing a transmission timer to determine whether a current block period is one of the second block periods and, if so, sending the value current from the first parameter to the management server; and - in the second communication device, processing a receive timer synchronized with respect to the transmission counter to determine whether a current block period is one of the second block periods and, if so, sending the current value of the second parameter to the server of management.
    [00039] According to a second aspect, the present invention provides a communication network comprising a first communication device and a second communication device, in which: - the first communication device is configured to transmit a data stream to the second communication device and, upon transmission of the data stream, during the first block periods that alternate in time with the second block periods, mark each data unit of the data stream by defining a data unit resource in a first value and updating a first parameter with respect to data units; - the second communication device is configured to receive the data stream from the first communication device and, upon receipt of the data stream, check the resource for each data unit received and update a second parameter in relation to the data units when the resource equals the first value, wherein the first communication device and the second communication device are further configured to process a timer to determine whether a current block period is one of the second block periods and, if so, to provide a current value of the first parameter and a current value of the second parameter to perform a measurement on the data stream. Brief Description of Drawings
    [00040] The present invention will become clearer from the following detailed description, given by way of example and not limitation, to be read in relation to the attached drawings, in which: - Figure 1 schematically shows an exemplary packet-switched network ; figure 2 schematically shows a package structure according to an embodiment of the present invention; figure 3 schematically shows a dial timing table according to an embodiment of the present invention; figure 4 is a flowchart of the operation of the first communication device according to a first embodiment of the present invention; - figure 5 shows three timelines in relation to the operation of the first communication device; - figures 6a and 6b are flowcharts of the operation of the second communication device according to the first embodiment of the present invention; - figure 7 shows three timelines in relation to the operation of the second communication device; figure 8 is a flowchart of the operation of the first communication device according to a second embodiment of the present invention; and - figures 9a and 9b are flowcharts of the operation of the second communication device according to the second embodiment of the present invention. Detailed Description of Preferred Modalities of the Invention
    [00041] In the following, a first preferred embodiment of the method will be described in detail by referring to the particular exemplary case of measuring data loss in a packet-switched network by measuring the number of packets (i.e., the data unit is a packet) lost in transmission from a transmitting node to a receiving node.
    [00042] Figure 1 schematically shows an exemplary packet-switched communication network CN comprising five nodes N1, N2, ..., N5. The nodes N1, N2, ..., N5 of the communication network CN comprise communication devices suitable for transmitting and receiving traffic in the form of packet flows. In particular, node N1 (hereinafter referred to as "transmitting node") comprises a first communication device CD1 and node N2 (hereinafter referred to as "receiving node") comprises a second communication device CD2. For example, the first communication device CD1 can comprise a first Ethernet port, while the second communication device CD2 can comprise a second Ethernet port.
    [00043] Nodes N1, N2, ..., N5 are connected to each other according to a partially meshed topology. In particular, nodes N1 and N2 are adjacent nodes, their communication devices CD1 and CD2 being connected by link L1. The number of nodes and the topology of the CN communication network are merely exemplary. The communication network CN can be, for example, an Ethernet network, an Internet network or any other type of packet-switched communication network.
    [00044] The CN communication network is suitable for cooperating with an MS management server. In Figure 1, the MS management server is connected to node N4. This is merely exemplary, as the MS management server can be connected to the CN communication network via any of its nodes N1, N2, ..., N5. Alternatively, the MS management server can be integrated into any of the nodes N1, N2, ..., N5 of the CN communication network.
    [00045] By way of example, below, only the PF packet stream transmitted from the first communication device CD1 of the transmitting node N1 to the second communication device CD2 of the receiving node N2 through the link L1 will be considered. The transmitting node N1 can be either the source node of the PF packet flow or an intermediate node on the path from the source node to the destination node. Similarly, receiving node N2 can be either the destination node of the PF packet flow or an intermediate node on the path from the source node to the destination node.
    [00046] The PF packet stream comprises a plurality of Pki packets. As shown in Figure 2, each Pki packet has a Hi header and a Pi payload. The payload Pi comprises a part of the traffic to be transmitted from the source node to the destination node. In addition, preferably, the Hi header comprises information for routing the Pki packet, such as the address of the source node (i.e. the address of the transmitting node N1, if it is the source node) and the address of the destination node (ie, the address of the receiving node N2, if it is the destination node).
    [00047] According to embodiments of the present invention, upon transmission of the Pki packets of the PF packet stream, the first communication device CD1 preferably marks the Pki packets to divide the PF packet stream into blocks, each block comprising numerous Pki packets. Preferably, all blocks have the same length of time, which will be called hereinafter the “block period” Tb.
    [00048] The first communication device CD1 preferably uses a bit bi of the Hi header of each Pki packet to mark the Pki packets. In particular, the first communication device CD1 sets Pki packets by setting the value of bit bi to 1 or 0. Bit bit bi can be, for example, a bit to which the protocol according to which the Pki packet is formatted has not yet assigned a specific role. Alternatively, bit bi can be a bit from a field with other uses, such as, for example, a bit from the priority field in IP packets, or a bit from the label field in MPLS packets.
    [00049] The block period Tb can be set by the network operator according to the desired data loss measurement rate (as will be described in detail below, the block period Tb is also the measurement period). In addition to the block period Tb, the network operator preferably defines a marking period Tm. The marking period Tm and the block period Tb are preferably selected so that the following equation is satisfied:
    where n is an even integer equal to or greater than 2, and Tm and Tb are expressed in the same measurement unit (for example, minutes or seconds). For example, the marking period Tm can be equal to 60 minutes, and the block period Tb can be equal to, for example, 1, 2, 3, 5, 6, 10, 12, 15 or 30 minutes (i.e. , all values satisfying equation [3] when Tm and Tb are expressed in minutes). Then, the marking period Tm comprises an even integer n of block periods T1, T2, T3, T4, ..., Tn-1, Tn.
    [00050] Referring to Figure 4, when the first communication device CD1 determines that a data loss measurement with respect to the packet flow PF needs to be started, preferably it initializes a first counter C1 and a second counter C0 on the value 0 (step 401). Then, preferably, the first communication device CD1 starts a transmission timer cyclically counting from 0 to the marking period Tm (step 402). For example, if Tm equals 60 minutes, the transmission timer can count cyclically from 00:00 to 59:59.
    [00051] When the actual time t* indicated by the transmission timer is equal to kx Tb (k being equal to 0, 1, 2, ..., n-1) (step 403), preferably, the first communication device CD1 determines the current block period from the block periods T1, T2, T3, T4, ..., Tn-1, Tn (step 404). Preferably, such determination is performed using an MTT marking timing table.
    [00052] The structure of the MTT dial timing table is shown in figure 3. Preferably, the MTT dial timing table comprises n lines (i.e., one line for each block period T1, T2, T3, T4, . .., Tn-1, Tn). Each line comprises a start time 0, Tb, 2Tb, 3Tb..., (n-2)Tb, (n-1)Tb, which indicates the time in which the block period T1, T2, T3, T4, ..., Tn-1, Tn starts, and a marking information, which indicates the value (namely, 1 or 0) at which bit bi should be set in Pki packets marked during the block period T1, T2, T3, T4, ., Tn-1, Tn. A local copy of the MTT tag timing table can be stored at transmitting node N1.
    [00053] For example, considering that the marking period Tm is equal to 60 minutes and the block period Tb is equal to 10 minutes (the marking period Tm thus comprising six block periods T1, T2, T3, T4, T5, T6), the resulting MTT timing table is reported below. Table I

    [00054] Start times are expressed in the format “minutes:seconds”.
    [00055] In step 404, preferably, the first communication device CD1 determines the current block period by comparing the current time t* indicated by the transmission timer with the contents of the marking timing table MTT. In particular, the current block period is determined as the block period whose start time is equal to the current read time t* indicated by the transmission timer in step 403.
    [00056] During step 404, preferably, the first communication device CD1 also reads, in the MTT marking timing table, the marking associated with the determined current block period, thereby determining the marking that needs to be applied on packets which will be transmitted during the current block period.
    [00057] For example, by reference to the exemplary Table I set forth, when, in step 403, the current time t* indicated by the transmission timer is, for example, 30:00, in step 404, the first communication device CD1 determines that the current block period is T4 and that the corresponding tick is 0.
    [00058] Then, the first communication device CD1 waits for possible Pki packets to be transmitted and marks them as either 1 or 0, as determined in step 404.
    [00059] In particular, if the current block period is an odd period T1, T3, ..., Tn-1, when a Pki packet needs to be transmitted (step 405), preferably the first communication device CD1 marks it by setting its bi bit to 1 (step 406), increases by one the value of the first counter C1 (step 407) and transmits the Pki packet along the L1 link (step 408). Otherwise, if the current block period is an even period T2, T4, ..., Tn, when a Pki packet needs to be transmitted (step 405'), preferably the first communication device CD1 marks it by its definition bit bi to 0 (step 406'), increases by one the value of the second counter C0 (step 407') and transmits the Pki packet along the L1 link (step 408').
    [00060] Therefore, during odd block periods T1, T3, ..., Tn-1, the Pki packets transmitted by the first communication device CD1 form odd blocks B1, B3, ., Bn-1 (shown in figure 5 ) and are marked by their bi bit set to 1 (in Figure 5, the reference “b(PF)” indicates the bi bit values for the Pki packets of the PF packet stream). Furthermore, in the manner also shown in Fig. 5, during odd block periods T1, T3, ., Tn-1, the value of the first counter C1 increases, while the value of the second counter C0 remains constant.
    [00061] On the other hand, during the even block periods T2, T4, ., Tn, the Pki packets transmitted by the first communication device CD1 form even blocks B2, B4, ..., Bn (shown in figure 5) and are set by their bit bit set to 0. Furthermore, as also shown in Figure 5, during the even block periods T2, T4, ..., Tn, the value of the second counter C0 increases, while the value of the first counter C1 stays constant.
    [00062] Figure 5 shows the blocks transmitted during a single marking period Tm. However, the operation of the first communication device CD1 is periodic, with period Tm. As a result, the PF packet stream is divided into a sequence of blocks of duration Tb, in which blocks that include Pki packets marked by bit bi equal to 1 alternate in time with blocks that include Pki packets marked by bit bi equal to 0. Since according to the exposed equation [3], the marking period Tm is an even multiple of the block period Tb, the alternation of blocks comprising packets marked 1 and blocks comprising packets marked 0 is the same in all Tm marking periods. In other words, in all marking periods Tm, the odd blocks B1, B3, ..., Bn-1 include Pki packets marked by bit bi equal to 1, while the even blocks B2, B4, ..., Bn include Pki packets marked by bit bi equal to 0.
    [00063] Blocks formed according to the above-described flowchart of figure 4 can include different numbers of Pki packets. In fact, the number of Pki packets that make up a given block depends on the number of packets actually transmitted by the first communication device CD1 during each block period. For example, during low traffic hours, blocks comprise a smaller number of packets than during high traffic hours. For the same reason, the first Pki packet of each block is not necessarily transmitted at the beginning of the corresponding block period.
    [00064] Again referring to Fig. 4, if the current block period determined in step 404 is an odd period T1, T3, ..., Tn-1, the first communication device CD1 determines that the value of counter C1 can is currently increasing, while the value of counter C0 is currently constant. Therefore, preferably, the first communication device CD1 sends the currently constant value of the second counter C0 to the management server MS (step 409).
    [00065] Otherwise, if the current block period determined in step 404 is an even period T2, T4, ..., Tn, the first communication device CD1 determines that the value of counter C0 may be currently increasing, while the value of counter C1 is currently constant. Therefore, preferably, the first communication device CD1 sends the currently constant value of the first counter C1 to the management server MS (step 409’).
    [00066] Preferably, in steps 409-409', the time in which the relevant counter value C1 or C0 is sent to the MS management server is delayed by a security timeout SWT with respect to the start time of the period of current block, for reasons that will be described in detail below.
    [00067] Now, by referring to figures 6a, 6b and 7, blocks B1, B2, B3, B4, ..., Bn-1, Bn are transmitted to the second communication device CD2, which receives additional corresponding blocks B' 1, B'2, B'3, B'4, ..., B'n-1, B'n. Each additional block B'1, B'2, B'3, B'4, ..., B'n-1, B'n differs from the corresponding block B1, B2, B3, B4, ..., Bn- 1, Bn where one or more Pki packets originally composed in the block were lost during transmission on L1 link.
    [00068] When the second communication device CD2 determines that a data loss measurement with respect to the PF packet stream needs to be started, it preferably initializes a third counter C'1 and a fourth counter C'0 to the value 0 ( step 601). Step 601 can be triggered, for example, by a management message transmitted from the management server MS and command from the second communication device CD2 to initiate a data loss measurement for the PF packet stream.
    [00069] Then, the second communication device CD2 monitors possible Pki packets received from the first communication device CD1 on L1 link (step 602). Every time a Pki packet is received, preferably the second communication device CD2 checks its marking, that is, it checks the value of its bi bit (step 603).
    [00070] If bit bi is equal to 1, preferably the second communication device CD2 increments by one the third counter C’1 (step 604). Otherwise, if bit bi is equal to 0, preferably the second communication device CD2 increases the fourth counter C’0 by one (step 604’).
    [00071] Therefore, as shown in figure 7, while the second communication device CD2 is receiving Pki packets from the odd blocks B'1, B'3, ..., B'n-1, the value of the third counter C '1 increases, while the value of the fourth counter C'0 remains constant. On the other hand, while the second communication device CD2 is receiving Pki packets from the even blocks B'2, B'4, ..., B'n, the value of the fourth counter C'0 increases, while the value of the third counter C'1 stays constant.
    [00072] Substantially parallel to the exposed steps of the flowchart of figure 6a, the second communication device CD2 also starts a cyclically receiving timer that counts from 0 until the marking period Tm (step 605).
    [00073] The detection timer in the second communication device CD2 and the transmission timer in the first communication device CD1 are preferably started at substantially the same time and are preferably synchronized with each other so that they indicate substantially the same current times . For this purpose, the transmission timer and the reception timer are preferably synchronized with the local clocks of the transmitting node N1 and the receiving node N2, respectively, which, in turn, are mutually synchronized. The synchronization of the local clocks in the CN communication network can be achieved by means of any known synchronization protocol, such as, for example, the Time Protocol for Networks (NTP).
    [00074] When the current time t** indicated by the reception timer is equal to kx Tb (k being equal to 0, 1, 2, ..., n-1) (step 606), preferably, the second communication device CD2 determines the current block period from the block periods T1, T2, T3, T4, ..., Tn-1, Tn (step 607). For this purpose, a local copy of the MTT marking timing table is also stored in the receiving node N2 and preferably the second communication device CD2 determines the current block period by comparing the current time t** indicated by the receiving timer with the contents of the MTT marking timing table. In particular, the current block period is determined as the block period whose start time is equal to the current read time t** indicated by the receive timer in step 606.
    [00075] Since the transmit timer and the receive timer are synchronized, the current block period determined by the second communication device CD2 in step 607 is substantially equal to the current block period that is simultaneously determined by the first communication device CD1 (see step 404 of figure 4).
    [00076] If the current block period determined in step 607 is an odd block period T1, T3, ..., Tn-1, the second communication device CD2 determines that the first communication device CD1 is marking packets at 1 and that, in this way, the value of counter C'1 is currently increasing, while the value of counter C'0 is currently constant. Therefore, preferably, the second communication device CD2 sends the currently constant value of the fourth counter C’0 to the management server MS (step 608).
    [00077] Otherwise, if the current block period is an even block period T2, T4, ..., Tn, the second communication device CD2 determines that the first communication device CD1 is marking packets at 0 and that, in this way, the value of counter C'0 is currently increasing, while the value of counter C'1 is currently constant. Therefore, preferably, the second communication device CD2 sends the currently constant value of the third counter C’1 to the management server (step 608’).
    [00078] Preferably, in steps 608-608', the time in which the relevant counter value C'1 or C'0 is sent to the MS management server is delayed by a security timeout SWT with respect to the time of start of the current block period.
    [00079] In fact, the window of time during which each block B'1, B'2, B'3, B'4, ..., B'n-1, B'n is actually received on the second device CD2 communication can be delayed with respect to the corresponding block period T1, T2, T3, T4, ..., Tn-1, Tn, as determined in step 607, as shown in Figure 7. This delay is due, mainly, to the delay of packet propagation on L1 link and processing times in CD1, CD2 communication devices. In view of this, the fourth counter C'0 may be increasing in a first part of the odd periods T1, T3, ., Tn-1, while the third counter C'1 may be increasing in a first part of the even periods T2, T4 , ., Tn. This delay can be further improved by possible timing tolerances between the transmit timer and the receive timer.
    [00080] The SWT safety timeout applied in steps 409-409' by the first communication device CD1 and in steps 608608' by the second communication device CD2 ensures that the values of counters C1, C0, C1' and C'0 sent to the MS management server only when they are actually constant. The safety timeout SWT is preferably composed between a minimum value greater than 0 and 50% of the block period Tb. The minimum value is preferably greater than the sum of the maximum expected delay between the first communication device CD1 and the second communication device CD2, the synchronization error between the transmission timer and the reception timer and the processing time required to change the marking on communication devices CD1, CD2. The minimum value can be equal to, for example, 1% of the block period Tb. For example, if the block period Tb is equal to 10 minutes, the safety timeout SWT can be equal to, for example, 3 minutes.
    [00081] Therefore, during the odd block periods T1, T3, ..., Tn-1, the management server MS receives the values of the counters C0 and C'0 while, during the even block periods T2, T4, ..., Tn, it receives the values of counters C1 and C'1.
    [00082] Then, in each block period, the MS management server can calculate the data loss as the difference between the received counter values (that is, the difference C1-C'1 during even periods and the difference C0- C'0 for odd periods).
    [00083] The data loss calculated in each block period actually refers to the block period preceding the current one, as it is calculated based on the values reached by the relevant counters at the end of the preceding block period. For example, the data loss calculated based on the counter values received during the third block period T3 indicates the data loss that affected the PF packet flow at the end of the second block period T2, as it is calculated based on values reached by counters C0 and C'0 at the end of the second period of block T2.
    [00084] Once the data loss values have been calculated for each block period Tb, they can be processed. So, for example, the number of Pki packets lost during a single block period Tb can be divided by the number of Pki packets transmitted in that block period (which can be derived by the values of the first counter C1 and the second counter C0), of that mode, providing a packet loss ratio (that is, the ratio of the number of lost packets to the number of transmitted packets).
    [00085] Advantageously, the final data loss values are independent of the number of Pki packets included in each block. This is due to the fact that they are calculated as a difference between packets transmitted in a block period Tb and packets received in a corresponding block period Tb.
    [00086] The above-described method has numerous advantages.
    [00087] First of all, it can be easily implemented by communication devices (ports or interfaces) that are typically available on nodes of a current communication network.
    [00088] Furthermore, for each block period, only the values of counters actually needed to calculate data loss are sent to the management server, while counters whose values are not currently constant (and which, in this way, are not should be used to calculate data loss) are not sent. Additionally, the values of the relevant counters are sent to the management server only once in each block period Tb. In other words, the block period Tb is also the measurement period.
    [00089] For example, if the block period Tb is 10 minutes according to the present invention, the values of two counters are sent to the management server once every 10 minutes. According to the method of WO 2010/072251, the values of four counters must be detected at least twice every 10 minutes. Considering that the detection period is, for example, 4 minutes, the values of the four counters should be detected about 2.5 times every 10 minutes. This means that the amount of information to be gathered by the MS management server to implement the method is advantageously reduced by a factor equal to about 5. Both the computational effort on the nodes and the use of bandwidth on the communication network links they are, in this way, advantageously reduced.
    [00090] Additionally, the algorithm implemented by the communication devices CD1, CD2 is advantageously very simple. In fact, to determine the counters to be sent to the management server that allow for a proper data loss calculation, the communication devices simply need to read the current time indicated by their respective timers and check the MTT tag timing table. It has been estimated that the algorithms shown in figures 4, 6a and 6b reduce the data loss computation time by about 1/10 compared to the algorithm of WO 2010/072251.
    [00091] On the one hand, this allows the application of the method to measure data loss in a larger number of data streams transmitted through the CN communication network, without increasing the block period Tb. In fact, in a single block period Tb, nodes can apply the exposed algorithms to calculate data loss on a very high number of data streams. On the other hand, given a number of data streams to be measured, this allows to reduce the block period Tb to provide a more frequent measurement of data loss from the data streams. This makes it possible to advantageously reduce the delay between a data loss affecting one of the measured data streams and the detection of such data loss.
    [00092] Additionally, advantageously, thanks to their synchronized timers and the MTT marking timing table, the communication devices CD1 and CD2 are always aware of the current marking applied to the packets and thus are able to determine which counters should be used to properly calculate data loss, even in block periods during which there are no packets to be transmitted.
    [00093] According to a second embodiment of the present invention, the above-described method can be used to perform, in addition to a data loss measurement, also a time measurement in the PF data stream.
    [00094] Referring to Figure 8, which shows the operation of the first communication device CD1 according to the second embodiment of the present invention, if, in step 404, the first communication device CD1 determines that the current block period is a odd period T1, T3 ..., Tn-1, after marking to 1 of each Pki packet to be transmitted (step 406) and incrementing the counter C1 by one (step 407), preferably, the first communication device CD1 determines whether the Pki packet is the first to be transmitted during the current odd block period (step 407a). If so, preferably the first communication device CD1 sets a first transmission time register S1 equal to the actual time indicated by the transmission timer (step 407b).
    [00095] Otherwise, if, in step 404, the first communication device CD1 determines that the current block period is an even period T2, T4 ..., Tn, after setting 0 in each Pki packet to be transmitted (step 406') and increasing the C0 counter by one (step 407'), preferably, the first communication device CD1 determines whether the Pki packet is the first to be transmitted during the current even block period (step 407a') . If so, preferably the first communication device CD1 sets a second transmission time register S0 equal to the actual time indicated by the transmission timer (step 407b’).
    [00096] Furthermore, if the current block period determined in step 404 is an odd period T1, T3, ..., Tn-1, in step 409, the first communication device CD1 sends the value to the management server MS. currently constant from the second counter C0 and, furthermore, the current value of the second transmit time register S0 (step 409). Otherwise, if the current block period determined in step 404 is an even period T2, T4, ..., Tn, in step 409', the first communication device CD1 sends to the management server MS the value currently contained in the first counter C1 and, in addition, the current value of the first transmission time register S1 (step 409').
    [00097] Referring to Figures 9a and 9b, which show the operation of the second communication device CD2 according to the second embodiment of the present invention, if, in step 603, the second communication device CD2 determines that the currently received packet is set to 1, after incrementing the counter C'1 by one (step 604), preferably the second communication device CD2 checks whether the previously received packet has the same tag as the current packet Pki (step 604a). If so, the second communication device CD2 determines that the Pki packet is not the first packet of a block and preferably takes no further action. If negative, the second communication device CD2 determines that packet Pki is the first packet of a block and preferably sets a first receive time register R1 equal to the current time indicated by the receive timer (step 604b).
    [00098] Otherwise, if, in step 603, the second communication device CD2 determines that the currently received packet is set to 0, after incrementing the counter C'0 by one (step 604'), preferably, the second device CD2 communication check that the previously received packet has the same tag as the current Pki packet (step 604a'). If so, the second communication device CD2 determines that the Pki packet is not the first packet of a block and preferably takes no further action. If negative, the second communication device CD2 determines that packet Pki is the first packet of a block and preferably sets a second receive time register R2 equal to the current time indicated by the receive timer (step 604b’).
    [00099] Similar to the first mode, in parallel to the above-described steps, the second communication device CD2 also starts a cyclically receiving timer that counts from 0 to the marking period Tm (step 605) and, based on this, determines the current block period (steps 606 and 607).
    [000100] If the current block period determined in step 607 is an odd block period T1, T3, ..., Tn-1, the second communication device CD2 determines that the value of counter C'0 is currently constant and , in this way, sends the currently constant value of the fourth counter C'0 and the current value of the second reception timestamp R0 to the MS management server (step 608).
    [000101] Otherwise, if the current block period is a block period even T2, T4, ..., Tn, the second communication device CD2 determines that the value of counter C'1 is currently constant and thus , sends the currently constant value of the third counter C'1 and the current value of the first receive timestamp R1 to the management server (step 608').
    [000102] Therefore, according to the second mode, during odd block periods T1, T3, ..., Tn-1, the MS management server receives the values of the counters C0 and C'0 and the time registers S0 and R0 while, during even block periods T2, T4, ..., Tn, it receives the values of counters C1 and C'1 and of time registers S1 and R1.
    [000103] Then, in each block period, the MS management server can calculate the data loss as the difference between the received counter values (that is, the difference C1-C'1 during the even periods and the difference C0 -C'0 during odd periods).
    [000104] Furthermore, in each block period, the MS management server can calculate the delay as the difference between the values of the timestamps received (that is, the difference R1-S1 during the even periods and the difference R0 -S0 during odd periods). The calculated delay in a given block period is substantially the delay induced by propagation from the first communication device CD1 to the second communication device CD2 in the first transmitted packet during the block period preceding the current one. For example, if the block period Tb is equal to 10 minutes and the tick period Tm is equal to 60 minutes, when the transmit timer and receive timer (which are mutually synchronized) indicate a current time t* = t ** equal to 30 minutes, in steps 404 and 607, the current block period determined is T4, whose associated mark is 0. Therefore, communication devices CD1 and CD2 send to the MS management server the values of time stamps S1 and R1 respectively, which indicate the transmission time and reception time of the first packet transmitted during the preceding block period, i.e. T3.
    [000105] In addition to the delay calculation, the MS management server can also calculate an interarrived jitter as a difference between the delay calculated in the current block period and the delay calculated in the preceding block period.
    [000106] Preferably, in each block period, the MS management server calculates the interarrived delay and/or jitter only if the calculated data loss in the same block period equals 0. In fact, if the packet transmitted first during a lost block period before reaching the second communication device CD2, the value assigned to the reception time register R0 or R1 during this block period refers to a subsequent packet of this block, not the first. Therefore, a time measurement made on the basis of such a reception timestamp is inherently imprecise because it is calculated as a difference between timestamps that refers to different packets.
    [000107] However, if the first packet of a block period is lost during transmission from the first communication device CD1 to the second communication device CD2, this does not affect the accuracy of the interarrival delay and jitter measurements referred to the following blocks. In fact, the loss of the first packet (or any other packet) in a block does not prevent the second communication device CD2 from recognizing the first packet in the successive block and properly setting the reception time register for this block. Since then, the measurement of interarrival delay and instability will be immune from inaccuracy due to the loss of the first packet in the previous block.
    [000108] Time measurements are not necessarily based on the first packets of each block. According to variants of the present invention, the transmission times registers and the reception times registers can be defined for a different packet of the block. Additionally, to increase the interarrival delay and jitter measurement period without reducing the block period Tb, the transmit time register and receive time register can be set for more than one packet for each block (eg. the first, the 100°, the 200°, etc.). In each block period, the communication devices CD1 and CD2 send to the management server MS all the values of the transmission timestamp and the reception timestamp.
    [000109] The method according to the second modality of the present invention has basically the same advantages as the method according to the first modality, namely: reduced computational effort on the MS management server, low bandwidth usage in the connections of the communication network, possibility of providing a measurement even when traffic between CD1 and CD2 is absent for one or more block periods.
    [000110] As described above, both the first communication device CD1 and the second communication device CD2 are able to determine the current block period based on the current times indicated by the transmit timer and the receive timer, respectively. In particular, the current time is compared with the contents of the MTT tag timing table.
    [000111] According to advantageous variants of the present invention, both the first communication device CD1 and the second communication device CD2 are able to determine the current block period (and therefore the counters and possibly the time stamps to be sent to the management server MS) simply by processing the current times t* and t** indicated by the transmit timer and the receive timer, respectively, without using the MTT marking timing table.
    [000112] According to these other embodiments, the communication devices CD1 and CD2 know the duration Tb of the block periods and know the marking associated with the first block period T1 of each marking period Tm. Next, this tag is considered to be 1.
    [000113] Considering that the block period Tb is equal to an odd number of minutes (eg 5 minutes), the start time of the odd block periods T1, T3, ., Tn-1 expressed in minutes is 0 or an even number (10, 20, etc.), while the start time of the even block periods T2, T4, ...,Tn expressed in minutes is an odd number (5, 15, etc.).
    [000114] Therefore, in steps 404, the first communication device CD1 checks whether the current time t* indicated by the transmission timer (which, as stated, is equal to k * Tb, k being equal to 0, 1, 2, . .., n-1) is an odd or even number (expressed in minutes). If the current time t* expressed in minutes is an even number, the first communication device CD1 determines that an odd block period T1, T3, ..., Tn-1 is starting and then sends it to the management server MS the current value of the C0 counter and possibly the current value of the S0 transmission time register (step 409). Otherwise, if the current time t* expressed in minutes is an odd number, the first communication device CD1 determines that an even block period T2, T4, ..., Tn is starting and then sends it to the management server MS the current value of counter C1 and possibly the current value of transmission time register S1 (step 409').
    [000115] Similarly, in steps 607, the second communication device CD2 checks whether the current time t** indicated by the receive timer (which, as stated, is equal to k * Tb, k being equal to 0, 1, 2, ., n-1) is an odd or even number (expressed in minutes). If the current time t** expressed in minutes is an even number, the second communication device CD2 determines that an odd block period T1, T3, ., Tn-1 is starting and then sends to the management server MS o current value of counter C'0 and possibly the current value of reception time register R0 (step 608). Otherwise, if the current time t** expressed in minutes is an odd number, the second communication device CD2 determines that an even block period T2, T4, ., Tn is starting and then sends it to the management server MS the current value of counter C'1 and possibly the current value of the reception time register R1 (step 608').
    [000116] Considering now that the block period Tb is equal to an even number of minutes (eg 10 minutes), the start time of the odd block periods T1, T3, ..., Tn-1 expressed in minutes is 0 or an integer multiple of 2Tb (20, 40, etc.), while the start time of even block periods T2, T4, ..., Tn expressed in minutes is not an integer multiple of 2Tb (10, 30 , etc.).
    [000117] Therefore, in step 404, the first communication device CD1 checks whether the current time t* indicated by the transmission timer (which, as stated, is equal to k * Tb, k being equal to 0, 1, 2, . .., n-1) is an integer multiple of 2Tb (expressed in minutes). If so, the first communication device CD1 determines that an odd block period T1, T3, ..., Tn-1 is starting and then sends to the management server MS the current value of counter C0 and possibly the current value of the S0 transmission time register (step 409). If negative, the first communication device CD1 determines that a block period even T2, T4, ..., Tn is starting and then sends to the management server MS the current value of counter C1 and possibly the current value of the transmission time register S1 (step 409').
    [000118] Similarly, in step 607, the second communication device CD2 checks whether the current time t** indicated by the receive timer (which, as stated, is equal to k * Tb, k being equal to 0, 1, 2, ., n-1) is an integer multiple of 2Tb (expressed in minutes). If so, the second communication device CD2 determines that an odd block period T1, T3, ., Tn-1 is starting and then sends to the management server MS the current value of counter C'0 and possibly the current value of the reception time register R0 (step 608). If negative, the second communication device CD2 determines that a block period even T2, T4, ., Tn is starting and then sends to the management server MS the current value of counter C'1 and possibly the current value of the reception time register R1 (step 608').
    [000119] Therefore according to such modalities, the first and second communication devices CD1 and CD2 are able to determine the current block period (and therefore the counters and possibly the timestamps to be sent to the server MS management) without the need to cooperate with the MTT marking timing table.
    [000120] Although the method has been described in detail only for measuring data loss (and possibly also for taking time measurements) of a PF packet flow between two physically adjacent nodes (ie, transmitting node N1 and the receiving node N2) according to modalities not shown in the drawings, it can be used to measure a packet flow between two physically non-adjacent nodes. In particular, it can be used to measure data loss in relation to a complete packet stream transmitted from a given source node to a given destination node. In this case, preferably, the source node identifies the packets that belong to the complete packet flow (for example, by using its source address and its destination address) and only marks the identified packets. On the other hand, also, the destination node identifies the packets that belong to the complete packet flow and reads the tag of the identified packets only.
    [000121] The exposed method can be applied both in the case of point-to-point transmission, and in the case of point-to-multipoint transmission or by multicast. In the latter case, a separate data loss measurement can be performed for each point-to-multipoint transmission destination node.
    [000122] According to additional modalities, the above-described method is applied in a complete packet stream to measure the end-to-end data loss and also the amount of data lost in each intermediate hop. For this purpose, preferably each intermediate node comprises a communication device configured to apply the algorithm of figures 6a-6b or 9a-9b to packets received from an upstream node, and an additional communication device configured to apply the algorithm of figures 4 or 8 in the same packages before forwarding them to a downstream node. In this way, each intermediate node provides the MS management server with a first pair of counters (and possibly a first pair of timestamps) in relation to incoming packets marked 1 or 0 and a second pair of counters (and, possibly with a second pair of timestamps) against outgoing packets marked 1 or 0. Preferably, the MS management server processes the counters and timestamps received by the various nodes in each block period to derive edge measurements the edge, link measurements (ie, measurements in relation to transmission between two physically adjacent nodes) and inter-node measurement (ie, measurements in relation to the transition of packets through each node).
    [000123] According to additional embodiments not shown in the drawings, in addition to bit bi, at least one other bit of each Pki packet is reserved to mark Pki packets. For example, an additional bit in the header of each Pki packet can be used to mark the Pki packet as follows: when this additional bit is set to 0, it indicates that this Pki packet belongs to a packet stream currently not under measurement, while , when this additional bit is equal to 1, it indicates that the Pki packet belongs to a packet stream currently under measurement. This advantageously allows to distinguish packets from a packet flow to be measured from other packet flows that are not to be measured, without the need to read any other information in the packet header (such as, for example, the address of the originating node or the address of the destination node, as mentioned above). According to still further embodiments not shown in the drawings, numerous bits of the header of each Pki packet can be used to mark packets to indicate different packet flows to be measured. For example, 2 bits in the header will allow you to identify three different packet streams, while 3 bits will allow you to identify seven different packet streams.
    [000124] The bit bi (and possibly additional bits to identify the packet flows to be measured) can be provided in the header of the packets to be transmitted by suitably modifying the protocols according to which the packets are formatted. For example, if Pki packets are formatted according to MPLS (Multi-Protocol Label Switching), the label field of the MPLS header can include the bit bit and also can be used to identify different packet streams to be measured. In this situation, the packet headers label field is used to alternately transmit Pki packets from the first communication device CD1 to the second communication device CD2 through a first tunnel T1 (for a block period Tb) and a second tunnel T2 (for a Tb block period). Therefore, a first packet stream PF1 is transmitted through the first tunnel T1 and a second packet stream PF2 is transmitted through the second tunnel T2. For measuring data loss on the link between CD1 and CD2, data loss is measured separately in the first tunnel T1 and in the second tunnel T2. In particular, the number of packets entering and leaving each tunnel is measured substantially with a same measurement period equal to the block period Tb according to the procedure described above. If a number of packet streams is to be transmitted from the first communication device CD1 to the second communication device CD2, each packet stream is assigned to a respective pair of tunnels. For each packet flow, the data loss measurement is performed as stated.
    [000125] According to alternative embodiments not shown in the drawings, the first communication device CD1 only marks the packets upon their transmission and appropriately increases the counters C1 and C0, and the second communication device CD2 reads only the marking of the received packets and appropriately increases counters C'1 and C'0, while the MS management server is responsible for deciding which counters should be gathered from communication devices CD1 and CD2 in each block period. For this purpose, the management server MS is preferably provided with a timer synchronized at least with respect to the transmission timer of the first communication device CD1. This allows the MS management server to determine the current block period and the marking that is currently applied by the first communication device CD1. Based on this determination, the MS management server can ask the communication devices CD1 and CD2 for counter values that are not currently increasing.
    权利要求:
    Claims (15)
    [0001]
    1. Method for performing a measurement on a data stream (PF) to be transmitted from a first communication device (CD1) to a second communication device (CD2) of a communication network (CN), the method comprising: a ) by transmitting the data stream (PF): - during the first block periods (T1, T3, ..., Tn-1) which alternate in time with the second block periods (T2, T4, ... , Tn), mark each data unit (Pki) of the data stream (PF) by setting a resource (bi) of data unit (Pki) in a first value and updating a first parameter (C1, S1) in relation to data units (Pki); b) upon receipt of the data stream (PF): - check the resource (bi) for each received data unit (Pki) and update a second parameter (C'1, R1) in relation to the data units (Pki) when the resource (bi) is equal to the first value; c) while steps a) and b) are carried out, determine whether a current block period is one of the second block periods (T2, T4, ..., Tn) and, if so, provide a current value of the first parameter ( C1, S1) and an actual value of the second parameter (C'1, R1); and d) perform a measurement in the data stream (PF) using the current value of the first parameter (C1, S1) and the current value of the second parameter (C'1, R1), characterized by the fact that step c) comprises processing a timer (t*, t**) to determine if a current block period is one of the second block periods (T2, T4, ..., Tn).
    [0002]
    2. Method according to claim 1, characterized in that: - in step a), the update comprises increasing a first transmission counter (C1) by transmitting each data unit (Pki) with the resource (bi) defined in the first value; - in step b), the update comprises increasing a first reception counter (C’1) upon receipt of each data unit (Pki) with the resource (bi) defined in the first value; - in step c), the provision comprises providing a current value of the first transmission counter (C1) and a current value of the first reception counter (C’1); and - in step d), performing a measurement comprises calculating a data loss as a difference between the current value of the first transmission counter (C1) and the current value of the first reception counter (C’1).
    [0003]
    3. Method according to claim 1 or 2, characterized in that: - in step a), the updating comprises updating a first transmission time record (S1) by transmitting a predetermined data unit (Pki) with the resource (bi) defined in the first value; - in step b), the updating comprises defining a first reception time register (R1) upon reception of the predetermined data unit (Pki); - in step c), the provision comprises providing a current value of the first transmission time register (S1) and a current value of the first reception time register (R1); and - in step d), performing a measurement comprises performing a time measurement using the current value of the first transmission time record (S1) and the current value of the first reception time record (R1).
    [0004]
    4. Method according to claim 3, characterized in that carrying out a time measurement comprises measuring at least one of an interarrived delay and instability of the packet flow (PF).
    [0005]
    5. Method according to any of the preceding claims, characterized by the fact that all the first block periods (T1, T3, ..., Tn-1) and the second block periods (T2, T4, ..., Tn) have the same duration (Tb).
    [0006]
    6. Method according to claim 5, characterized by the fact that, in step c), the timer (t*; t**) counts a marking period (Tm) equal to an even integer multiple of the duration (Tb).
    [0007]
    7. Method according to claim 6, characterized in that step c) comprises comparing a current time indicated by the timer (t*, t**) with a marking timing table (MTT) comprising information about time of start and/or end time of at least one of the first block periods (T1, T3, ., Tn-1) and of the second block periods (T2, T4, ., Tn).
    [0008]
    8. Method according to claims 5 or 6, characterized in that the duration (Tb) is equal to an odd number of time measurement units and in which step c) comprises determining whether a current time indicated by the timer (t*, t**) and expressed in time measurement units is a multiple of duration (Tb) and is an even number.
    [0009]
    9. Method according to claims 5 or 6, characterized by the fact that the duration (Tb) is equal to an even number of time measurement units and in which step c) comprises determining whether a current time indicated by the timer ( t*, t**) and expressed in time measurement units is not an integer multiple of duration (Tb) multiplied by 2.
    [0010]
    10. Method according to any one of the preceding claims, characterized in that: - step a) additionally comprises, during the second block periods (T2, T4, ..., Tn), marking each data unit (Pki ) of the data flow (PF) by setting the resource (bi) of the data unit (Pki) to a second value and updating a third parameter (C0, S0) in relation to the data units (Pki); - step b) further comprises updating a fourth parameter (C’0, R0) in relation to the data units (Pki) when the resource (bi) is equal to the second value; - step c) comprises, if negative, providing a current value of the third parameter (C0, S0) and a current value of the fourth parameter (C’0, R0); and - step d) comprises performing the measurement using the current value of the third parameter (C0, S0) and the current value of the fourth parameter (C’0, R0).
    [0011]
    11. Method according to any one of the preceding claims, characterized in that step c) is performed by the first communication device (CD1).
    [0012]
    12. Method according to claim 11, characterized in that step c) is also performed by the second communication device (CD2).
    [0013]
    13. Method according to claim 12, characterized in that step d) is performed by a management server (MS) that cooperates with the communication network (CN).
    [0014]
    14. Method according to claim 13, characterized in that step c) comprises: - in the first communication device (CD1), processing a transmission timer (t*) to determine whether a current block period is a of the second block periods (T2, T4, ..., Tn) and, if so, send the current value of the first parameter (C1, S1) to the management server (MS); and - in the second communication device (CD2), processing a receive timer (t**) synchronized to the transmission counter (t*) to determine whether a current block period is one of the second block periods (T2, T4, ..., Tn) and, if so, send the current value of the second parameter (C0, S0) to the management server (MS).
    [0015]
    15. Communication network (CN) comprising a first communication device (CD1) and a second communication device (CD2), wherein: - the first communication device (CD1) is configured to transmit a data stream (PF ) to the second communication device (CD2) and, by transmitting the data stream (PF), during the first block periods (T1, T3, ..., Tn-1) that alternate in time with the second periods of block (T2, T4, .., Tn), mark each data unit (Pki) of the data stream (PF) by setting a resource (bi) of the data unit (Pki) in a first value and updating one first parameter (C1, S1) in relation to data units (Pki); - the second communication device (CD2) is configured to receive the data stream (PF) from the first communication device (CD1) and, upon receipt of the data stream (PF), check the resource (bi) for each received data unit (Pki) and update a second parameter (C'1, R1) in relation to the data units (Pki) when the resource (bi) is equal to the first value, where the first communication device (CD1) and the second communication device (CD2) are further configured to determine whether a current block period is one of the second block periods (T2, T4, ., Tn) and, if so, to provide a current value of the first parameter ( C1, S1) and an actual value of the second parameter (C'1, R1) to perform a measurement on the data stream (PF); characterized by the fact that the first communication device (CD1) and the second communication device (CD2) are configured to process a timer (t*, t**) to determine whether a current block period is one of the second block periods (T2, T4, ..., Tn).
    类似技术:
    公开号 | 公开日 | 专利标题
    BR112013010739B1|2021-08-17|METHOD FOR PERFORMING A MEASUREMENT ON A DATA FLOW, AND, COMMUNICATION NETWORK
    KR101597255B1|2016-03-07|Performing a time measurement in a communication network
    KR101475347B1|2014-12-30|Measurement of data loss in a communication network
    US9450846B1|2016-09-20|System and method for tracking packets in a network environment
    EP2884697B1|2018-04-11|Measuring method, device and system for network packet loss
    US8427963B2|2013-04-23|Method and system for analyzing and qualifying routes in packet networks
    WO2014048137A1|2014-04-03|Measuring method, device and system for network packet loss
    US9762464B2|2017-09-12|Measurement on data traffic in a communication network
    US11121938B2|2021-09-14|Performance measurement in a packet-switched communication network
    Ramadža et al.2015|Network performance monitoring within MPLS traffic engineering enabled networks
    US20210409296A1|2021-12-30|Enabling a performance measurement in a packet-switched communication network
    EP3398296B1|2021-06-30|Performance measurement in a packet-switched communication network
    BR112014002014B1|2021-12-21|METHOD TO PERFORM A MEASUREMENT ON DATA TRAFFIC TO A NODE OF A COMMUNICATION NETWORK, AND, NODE TO A COMMUNICATION NETWORK
    Liu2014|A novel method for estimating the variable and constant components of one-way delays without using the synchronized clocks
    同族专利:
    公开号 | 公开日
    EP2636185A1|2013-09-11|
    CN103262471B|2015-12-16|
    KR101443237B1|2014-09-23|
    EP2636185B1|2014-07-02|
    WO2012059138A1|2012-05-10|
    JP2014502441A|2014-01-30|
    JP5661941B2|2015-01-28|
    US9264336B2|2016-02-16|
    US20130223274A1|2013-08-29|
    BR112013010739A2|2016-08-09|
    CN103262471A|2013-08-21|
    KR20130093647A|2013-08-22|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US6188674B1|1998-02-17|2001-02-13|Xiaoqiang Chen|Method and apparatus for packet loss measurement in packet networks|
    US6662223B1|1999-07-01|2003-12-09|Cisco Technology, Inc.|Protocol to coordinate network end points to measure network latency|
    US6480899B1|1999-09-08|2002-11-12|Nortel Networks Limited|Differentiated services IP quality of services round trip time aware intelligent traffic conditioner in an ingress node of virtual private networks|
    JP4767443B2|2001-07-04|2011-09-07|富士通株式会社|Network storage type video camera system|
    US6850769B2|2002-02-14|2005-02-01|Qualcomm Incorporated|Method and apparatus for adaptive measurement of round-trip time in ARQ protocols and using the same for controlling flow of data in a communication system|
    US7372865B2|2003-07-09|2008-05-13|Fujitsu Limited|Processing data packets using time stamped marker packets|
    JP4070029B2|2005-01-14|2008-04-02|アンリツ株式会社|Communication quality evaluation system, apparatus and method|
    CN1881908A|2005-06-13|2006-12-20|华为技术有限公司|Method for measuring MPLS network performance parameter|
    JP4698550B2|2006-03-21|2011-06-08|富士通株式会社|Communication processing device and aggregation system|
    JP4736957B2|2006-05-31|2011-07-27|日本電気株式会社|Quality measurement system, communication device, communication terminal, and streaming distribution quality measurement method used therefor|
    US8451734B2|2008-12-22|2013-05-28|Telecom Italia S.P.A.|Measurement of data loss in a communication network|
    KR101871411B1|2011-07-28|2018-06-26|텔레콤 이탈리아 소시에떼 퍼 아찌오니|A method for performing a measurement on data traffic at a node of a communication network, node and management server for such a communication network, and recording medium storing a program for implementing such a method|
    CN103959713B|2011-09-19|2017-12-12|意大利电信股份公司|Measurement to the data flow in communication network|EP2903212B1|2012-09-29|2018-08-08|Huawei Technologies Co., Ltd.|Network delay measuring method and device|
    US9286620B2|2012-11-05|2016-03-15|Broadcom Corporation|Annotated tracing for data networks|
    US9438497B2|2013-05-06|2016-09-06|Viavi Solutions Inc.|Method and system for measuring packet loss|
    KR102059986B1|2013-05-31|2020-02-11|텔레콤 이탈리아 소시에떼 퍼 아찌오니|Performance measurement of a link of a packet-switched communication network|
    CN104301000A|2013-07-18|2015-01-21|中兴通讯股份有限公司|Method for data processing by utilizing sample point stage accelerator and sample point stage accelerator|
    WO2015090364A1|2013-12-17|2015-06-25|Telecom Italia S.P.A.|Time measurement in a packet-switched communication network|
    US20150200843A1|2014-01-15|2015-07-16|Cisco Technology, Inc. A Corporation Of California|Packet Labels For Identifying Synchronization Groups of Packets|
    US9641416B2|2014-04-18|2017-05-02|Cisco Technology, Inc.|Operations analysis of packet groups identified based on timestamps|
    CN106301987B|2015-06-03|2020-02-14|华为技术有限公司|Message loss detection method, device and system|
    WO2017071779A1|2015-10-30|2017-05-04|Telecom Italia S.P.A.|Performance measurement in a packet-switched communication network|
    CN109196823A|2016-04-29|2019-01-11|意大利电信股份公司|To the performance measurement of multiple spot stream of packets|
    US10831760B2|2016-09-13|2020-11-10|Viavi Solutions Inc.|Data stream monitoring|
    US11121938B2|2016-09-14|2021-09-14|Telecom Italia S.P.A.|Performance measurement in a packet-switched communication network|
    US10277493B2|2017-05-12|2019-04-30|Ciena Corporation|Packet throughput and loss ratio measurements of a service function chain|
    CN108989237B|2017-06-01|2021-03-23|华为技术有限公司|Method and device for data transmission|
    CN110233770A|2018-03-06|2019-09-13|华为技术有限公司|Network flow label and measurement method, node|
    WO2020016216A1|2018-07-18|2020-01-23|Telecom Italia S.P.A.|Performance measurement in a packet-switched communication network|
    法律状态:
    2018-03-27| B15K| Others concerning applications: alteration of classification|Ipc: H04L 12/26 (2006.01) |
    2019-01-08| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|
    2020-02-11| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|
    2021-06-08| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|
    2021-08-17| B16A| Patent or certificate of addition of invention granted [chapter 16.1 patent gazette]|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 05/11/2010, OBSERVADAS AS CONDICOES LEGAIS. PATENTE CONCEDIDA CONFORME ADI 5.529/DF, QUE DETERMINA A ALTERACAO DO PRAZO DE CONCESSAO. |
    优先权:
    申请号 | 申请日 | 专利标题
    PCT/EP2010/066955|WO2012059138A1|2010-11-05|2010-11-05|"measurement on a data flow in a communication network"|
    [返回顶部]