Performance measurement of live traffic transmitted over packet switched communication network

Abstract

A method for enabling performance measurements of multipoint packet flows carrying live traffic over a packet switched communication network is disclosed. Each packet includes a flag bit that, prior to its transmission, is set to a first value indicating that the packet will be subjected to a sample performance measurement, or to a second value indicating that the packet will not be subjected to a sample performance measurement. A measurement point placed on the path of the packet reads the value of its flag bit and, if it is set to a first value, it generates a sample performance parameter related to the packet; otherwise, it updates cumulative performance parameters associated with the multipoint packet flow as a whole. Sample measurements based on sample performance parameters and / or multi-point performance measurements based on cumulative performance parameters may then be performed.

Description

Technical Field

[0001] The present invention relates to the field of communication networks. In particular, the present invention relates to methods and systems for implementing performance measurements of packets conveying live traffic transmitted over a packet-switched communication network, and methods and systems for performing such performance measurements. Background Art

[0002] In a packet-switched communication network, a packet stream is transmitted from a source node to a destination node via possible intermediate nodes. Exemplary packet-switched networks are IP (Internet Protocol) networks, Ethernet networks and MPLS (Multi-Protocol Label Switching) networks.

[0003] Packets do not always reach their destination nodes, i.e. they may be lost during transmission through the network. Packet loss is caused by different reasons. For example, a node or a link may fail, or a node may drop packets due to congestion on its port. Furthermore, packets may be dropped by nodes because they contain bit errors.

[0004] In addition, each packet is sent by the source node at a send time and received by the destination node at a receive time. The time that elapses between the send time and the receive time is often referred to as the "one-way delay". The one-way delay of a packet depends primarily on the number of intermediate nodes that the packet may pass through from the source to the destination, the processing time of the packet at each node, and the propagation time along the link.

[0005] Furthermore, packets may have different one-way delays. The difference between the one-way delays of two packets of the same packet stream is called delay variation or "inter-arrival jitter" (or simply "jitter").

[0006] G. Fioccola et al., "Alternate-Marking Method for Passive and Hybrid Performance Monitoring", IETF (Internet Engineering Task Force) RFC 8321, January 2018, section 3.1, discloses a packet loss measurement technique that specifies virtually splitting the traffic flow to be measured into consecutive blocks, counting the number of packets in each block, and comparing the values ​​measured by different network devices along the path of the packet flow. To create the blocks, the packets of the packet flow are "colored" so that packets of the same block have the same color, while packets of different consecutive blocks have different colors. Traffic coloring can be achieved by setting a specific bit in the packet header and periodically changing the value of the bit. One-way delay measurement based on a dual-marking method is also envisioned (section 3.3.2), which specifies the use of a first mark to create an alternating flow, and the use of a second mark within the colored flow to select packets for measuring delay / jitter. The second mark creates a new set of marked packets that are fully recognized on the network, so that network devices can store timestamps for these packets.

[0007] G. Fioccola et al., "Multipoint Alternate-Marking Method for Passive and Hybrid Performance Monitoring", IETF (Internet Engineering Task Force) RFC 8889, August 2020, discloses an extension of the alternating marking technique of the above-mentioned RFC 8321, which is applicable to the most general case of multipoint packet flows. First, a monitoring network is derived on the multipoint path of the multipoint packet flow, so that all packets of the multipoint packet flow leaving the monitoring network have previously entered the monitoring network. The packets of the multipoint packet flow transmitted through the monitoring network are alternately marked. Multipoint packet loss measurements are obtained based on the counts provided by the measurement points of the monitoring network.

[0008] In this multipoint scenario, as disclosed in the IETF (Internet Engineering Task Force) draft "The BigData Approach for Multipoint Alternative Marking Method draft-c2f-ippm-big-data-alt-mark-01" by M Cociglio et al. on October 30, 2020, the packets of the multipoint packet stream to be measured at multipoint can be identified by applying address filtering techniques or by using additional bits. The default value of the additional bit is 0, and the value 1 is used to identify the packets of the traffic stream to be measured. Summary of the invention

[0009] Applicants have noted that in a multipoint scenario, it may be desirable to provide different types of measurements, including, for example, delay / jitter measurements of individual packets of a point-to-point packet stream included in a multipoint packet stream, as provided by RFC 8321, and multipoint measurements as provided by RFC 8889. In this case, if it is desired to avoid using address filtering techniques to identify packets of a multipoint packet stream, three different marker bits are required, namely:

[0010] - a first marking bit, used to implement alternating marking of all packets of a multipoint packet stream, as specified by both RFC 8321 and RFC 8889;

[0011] - a second marker bit for identifying packets in a point-to-point packet stream that are individually subject to delay / jitter measurement according to RFC 8321; and

[0012] - A third marker bit used to identify packets of a multipoint packet flow to be subjected to multipoint measurement according to RFC 8889.

[0013] FIG. 1 schematically illustrates the values ​​of the three marker bits in a packet of such a point-to-point packet flow according to the provisions of RFC 8321 and RFC 8889.

[0014] The first bit B1 is periodically switched between 1 and 0 with a period T (also referred to herein as a marking period) so as to implement an alternating marking technique whereby the traffic is divided into alternating blocks BL1, BL2, BL3, BL4, ... of duration T, as specified by both RFC8321 and RFC 8889. The second bit B2 is equal to 0 except for a single packet in each marking period that is to be subjected to delay / jitter measurement according to RFC 8321 (e.g., the first packet of each block BL1, BL2, BL3, BL4, ...). For those packets only, the second marking bit B2 is equal to 1. The third marking bit B3 is always equal to 1 to indicate that all packets belong to a multipoint packet flow and will therefore be considered for multipoint measurements according to RFC 8889. Packets that do not belong to a multipoint packet flow will be ignored (not shown in FIG. 1 ) and instead the third marking bit B3 is equal to a default value of 0.

[0015] While the above-described marking scheme will be able to support all desired measurements, the applicant has noted that there are currently only two bits in the packet's header available to implement marking of packets that support passive performance measurements (i.e., performance measurements performed on live traffic without the need to actively inject any special packets into the live traffic specifically for measurement purposes).

[0016] In view of the above, the applicant has solved the problem of providing a method and a system for implementing performance measurements of a multipoint packet stream carrying live traffic through a packet-switched communication network, the method and the system being based on an alternating marking technique and being capable of implementing sample performance measurements of packets of a point-to-point packet stream of a multipoint packet stream and multipoint performance measurements of the multipoint packet stream as a whole by using only two bits, without the need to apply any address filtering technique to identify the packets to be considered for multipoint measurements.

[0017] In the present specification and claims, the expression "sample performance measurement" shall specify a performance measurement related to a specific packet (e.g., its one-way delay, such as, by way of example but not exclusively, the delay measurement specified by RFC8321) or a sequence of specific packets within a packet stream (e.g., their delay variation or jitter, such as, by way of example but not exclusively, the jitter measurement specified by RFC 8321). Whereas the expression "multipoint performance measurement" shall indicate a performance measurement related to a multipoint packet stream as a whole (e.g., its packet loss or average one-way delay, as, by way of example but not exclusively, specified by RFC 8889), wherein a "multipoint packet stream" indicates a packet stream whose packets follow at least partially non-overlapping paths. Furthermore, the expression "implementing performance measurement" shall indicate an operation of providing or generating performance parameters (such as timestamps or counters) that can be used by nodes of a network or an external entity (such as a management server that collects performance parameters from nodes of the network) to calculate performance measurements.

[0018] According to an embodiment of the present invention, the above-mentioned problem is solved by a method and system for alternately switching the value of the first marking bit B1 between two alternative marking values ​​(e.g., 1 and 0). As for the second marking bit B2, it is set to a first marking value (hereinafter also referred to as a "sampling value", e.g., 0) indicating that the packet will be subjected to sample performance measurement, or is set to a second marking value (hereinafter also referred to as a "non-sampling value", e.g., 1) indicating that the packet will not be subjected to sample performance measurement. One or more measurement points placed on the path of the marked packet read the second marking bit B2 of the received packet. If the second marking bit B2 is set to a sampling value, the measurement point determines that the packet will be subjected to sample performance measurement, and then generates a sample performance parameter (e.g., a timestamp) related to the packet. Otherwise, if the second marking bit B2 is set to a non-sampling value, the measurement point determines that the packet will be considered for multi-point measurement, and updates the cumulative performance parameter (e.g., a cumulative counter) related to the multi-point packet flow as a whole accordingly. Performance measurement is thus achieved, which includes sample measurement based on sample performance parameters and / or multi-point performance measurement based on cumulative performance parameters.

[0019] Therefore, based on different values ​​0 or 1 of the same bit (i.e., the second marking bit B2), the packets to be subjected to sample measurement alone and the packets to be considered for multi-point measurement are identified. Therefore, the packets to be considered for multi-point measurement can be identified without applying any address filtering techniques and without requiring any additional bits for this purpose. Since the second marking bit B2 merges two separate functions, only two bits are required to support all requested performance measurements based on alternating marking, i.e., sample measurement (e.g., delay or jitter measurement specified by RFC 8321) and multi-point measurement (e.g., multi-point measurement specified by RFC 8889).

[0020] It should be noted that, in general, only a small number of packets are required to perform sample measurements on a point-to-point packet stream, typically at most one per marking period T. Therefore, even assuming that sample measurements are required on all point-to-point packet streams of a multipoint packet stream, the vast majority of packets of the multipoint packet stream will still have their second marking bit B2 set equal to the non-sampled value. This advantageously results in the multipoint measurement being based on the vast majority of packets of the multipoint packet stream. Therefore, an accurate multipoint measurement can be provided.

[0021] According to a first aspect, the present invention provides a method for implementing performance measurement of a multipoint packet flow conveying live traffic through a packet switching communication network, each packet of the multipoint packet flow comprising a first marking bit and a second marking bit, the method comprising:

[0022] a) prior to transmitting each packet of the multipoint packet flow through the network, alternately switching the value of a first marker bit between two alternative marker values; and setting a second marker bit to a first value indicating that the packet is to be subjected to sample performance measurement or a second value indicating that the packet is not to be subjected to sample performance measurement; and

[0023] b) reading a second marker bit at at least one measurement point placed on the path of the packet, and:

[0024] - if the second flag bit is set to the first value, generating a sample performance parameter associated with the packet; and

[0025] - If the second marker bit is set to the second value, updating the accumulated performance parameter related to the multipoint packet flow as a whole.

[0026] Preferably, in step a), the two alternative marking values ​​are equal to a first value indicating that the packet is to be subjected to sample performance measurement and a second value indicating that the packet is not to be subjected to sample performance measurement.

[0027] More preferably, the first value indicating that the packet is to be subjected to the sample performance measurement is 0, and the second value indicating that the packet is not to be subjected to the sample performance measurement is 1.

[0028] According to one embodiment, in step a), a second marking bit in at least one packet of a point-to-point packet stream included in a multi-point packet stream is set equal to a first value indicating that the packet will be subjected to sample performance measurement, and other packets of the point-to-point packet stream have a second marking bit set equal to a second value indicating that the packet will not be subjected to sample performance measurement.

[0029] Preferably, at step a), the method comprises periodically switching the value of the first marking bit between two alternative marking values ​​at a marking period T before transmitting each packet of the multipoint packet flow through the network.

[0030] Preferably, at step a), according to the marking period T, in a packet of the point-to-point packet flow, a second marking bit is set equal to a first value indicating that the packet is to be subjected to sample performance measurement.

[0031] According to one variant, in step a), the second marker bit is set equal to the first value indicating that the packet will be subjected to sample performance measurement in a packet of the point-to-point packet flow according to the marking period T only when the first marker bit during the marking period T is set equal to a predetermined marking value among two alternative marking values.

[0032] Preferably, the predetermined one of the two alternative marking values ​​is equal to a second value indicating that the packet is not to be subjected to sample performance measurement.

[0033] According to some embodiments, step b) includes if the second marker bit is set to a first value indicating that the packet will be subject to sample performance measurement, and the first marker bit is equal to a second value indicating that the packet will not be subject to sample performance measurement, then also includes updating a cumulative performance parameter related to the multi-point packet flow as a whole.

[0034] According to a variant, the method further comprises periodically exchanging, according to a marking period T, a first value indicating that the packet is to be subjected to a sample performance measurement and a second value indicating that the packet is not to be subjected to a sample performance measurement.

[0035] Preferably, the sample performance parameter related to the packet comprises a timestamp indicating a reception time of the packet at at least one measurement point.

[0036] Preferably, the accumulated performance parameter comprises an accumulated counter counting packets of the multipoint packet flow received at the at least one measurement point and / or an accumulated or average timestamp indicating an accumulated or average reception time of packets of the multipoint packet flow at the at least one measurement point.

[0037] According to one embodiment, at step b), at least one measurement point determines whether the packet belongs to a point-to-point packet flow requiring sample performance measurement by applying an access command list technique.

[0038] According to one embodiment, in step a), the multipoint packet flow is injected into the network via at least two source nodes, and periodically switching the value of the first marking bit is performed by the at least two source nodes in a substantially synchronous manner.

[0039] According to a second aspect, the present invention provides a method for performing performance measurement in a packet-switched communication network, the method comprising the steps of the method for implementing performance measurement as described above, and:

[0040] c) performing performance measurements, including at least one of group-related sample measurements based on sample performance parameters and multi-point performance measurements based on accumulated performance parameters.

[0041] Optionally, step b) further comprises sending at least one of the sample performance parameter and the accumulated performance parameter to a management server cooperating with at least one measurement point, and step c) is performed by the management server.

[0042] According to a third aspect, the present invention provides a packet switching communication network supporting transmission of a multipoint packet stream carrying live traffic, each packet of the multipoint packet stream comprising a first marking bit and a second marking bit, the packet switching communication network comprising:

[0043] a) at least one node configured to, before transmitting each packet of the multipoint packet flow through the network, alternately switch the value of a first marker bit between two alternative marker values; and set a second marker bit to a first value indicating that the packet will be subject to sample performance measurement or a second value indicating that the packet will not be subject to sample performance measurement; and

[0044] b) at least one measurement point placed on the path of the packet, the at least one measurement point being configured to read the second marker bit, and:

[0045] - if the second flag bit is set to the first value, generating a sample performance parameter associated with the packet; and

[0046] - If the second marker bit is set to the second value, updating the accumulated performance parameter associated with the multipoint packet flow as a whole.

[0047] Preferably, the packet switching communication network further comprises:

[0048] c) A management server configured to perform performance measurement, including at least one of sample measurement based on sample performance parameters associated with the group and multi-point performance measurement based on accumulated performance parameters. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The present invention will become more apparent from the following detailed description given by way of example and not limitation, with reference to the accompanying drawings, in which:

[0050] - Figure 1 (described above) shows the known marking scheme specified by RFC 8321 and RFC 8889;

[0051] - Figure 2 An exemplary network supporting transmission of multi-point packet flows is schematically shown, in which a method for performing performance measurement according to an embodiment of the present invention is implemented;

[0052] - Figure 3 schematically illustrates an alternative network supporting transmission of other multipoint packet flows;

[0053] - Figure 4 Schematically illustrates the structure of packets exchanged in the communication network of FIG. 1 according to an embodiment of the present invention;

[0054] - Figure 5 A marking scheme based on two marking bits according to an embodiment of the present invention is shown;

[0055] - Figure 6 is a flow chart of the operation of a measurement point according to an embodiment of the present invention;

[0056] - Figure 7 is a flow chart of the operation of a measurement point according to another embodiment of the present invention;

[0057] - Figure 8 A marking scheme based on two marking bits according to another embodiment of the present invention is shown; and

[0058] - Fig. 9 and Fig.10 A marking scheme based on two marking bits according to another embodiment of the present invention is shown. DETAILED DESCRIPTION

[0059] Figure 2 An exemplary network 100 supporting transmission of multipoint packet flows PF is schematically shown, in which a method for implementing performance measurement according to an embodiment of the present invention is implemented. The network 100 may be an IP network or any other type of packet switching network (eg, MPLS or Ethernet).

[0060] The network 100 includes a plurality of nodes interconnected by links according to any known topology. As a non-limiting example, the network 100 includes eight nodes N1, N2, ... N8 interconnected according to a partial mesh topology. Specifically, the node N1 is connected to N2, N3 and N8, the node N2 is connected to N4 and N5, the node N3 is connected to N4 and N7, and the node N4 is connected to N6.

[0061] The network 100 supports the transmission of a multipoint packet flow PF. The multipoint packet flow PF includes K point-to-point packet flows.

[0062] These K point-to-point packet flows can originate from N = 1 source node and be addressed to M > 1 destination nodes. This is Figure 2 the case of the exemplary scenario described in, where the packet flow PF includes K = 5 point-to-point packet flows PF1, PF2,... PF5, with N = 1 source node N1 and M = 4 destination nodes N5, N6, N7, and N8. Specifically, PF1 and PF2 are addressed to N5, PF3 is addressed to N6, PF4 is addressed to N7, and PF5 is addressed to N8. The multi-point packet flow PF as a whole is injected into the network 100 via the source node N1 and then split. Specifically, as a non-limiting example, PF1, PF2, and the first part PF3' of the packet flow PF3 are transmitted from N1 to N2, while the other part PF3" of the packet flow PF3 is transmitted from N1 to N3 together with PF4, and PF5 is transmitted to N8. At node N2, PF1, PF2, and the first part PF3' of the packet flow PF3 are further split, i.e., PF1 and PF2 are transmitted to N5, while PF3' is transmitted to N4. At node N3, PF3" and PF4 are further split, i.e., PF3" is transmitted to N4, while PF4 is transmitted to N7. At node N4, the parts PF3' and PF3" are joined and forwarded to N6.

[0063] Alternatively, the multi-point packet flow PF can have N > 1 source nodes and M = 1 destination node. This is Figure 3 the case of the exemplary scenario described in a, where nodes N1, N2, N3 are three source nodes, and N7 is the only destination node.

[0064] Alternatively, the multi-point packet flow PF can have N > 1 source nodes and M > 1 destination nodes. This is Figure 3 the case of the exemplary scenario described in b, where nodes N1, N2, N3 are three source nodes, and N6 and N7 are two destination nodes. In this case, the multi-point packet flow PF can have fewer source nodes and several destination nodes, i.e., N < M. This is the case, for example, of multi-point packet flows carrying traffic of OTT Internet services in the downlink direction, i.e., from several OTT servers to multiple end-users, or of those multi-point packet flows carrying LTE (Long-Term Evolution) traffic in the downlink direction, i.e., from several packet gateways to several eNodeBs. Alternatively, the multi-point packet flow PF can have several source nodes and fewer destination nodes, i.e., N >> M. This is the case, for example, of multi-point packet flows carrying traffic of OTT Internet services in the uplink direction, i.e., from multiple end-users to several OTT servers, or of those multi-point packet flows carrying LTE traffic in the uplink direction, i.e., from several eNodeBs to several packet gateways.

[0065] Alternatively, the multipoint packet flow PF may have N=1 source nodes and M=1 destination nodes, but along different paths through the network 100. This is Figure 3 c, where N1 is the only source node and N5 is the only destination node.

[0066] Preferably, if Figure 4 As schematically described in , each packet Pk of the multipoint packet flow PF comprises a header H and a payload PL. The payload PL comprises user data. Preferably, the header H comprises information for routing the packet Pk. The header format depends on the protocol according to which the packet Pk is formatted.

[0067] According to a preferred embodiment of the present invention, the packets of the multipoint packet flow PF are marked before they are injected into the network 100 via their respective source nodes PF. The marking can be implemented at the source nodes PF or upstream of the source nodes PF through which the K point-to-point packet flows included in the multipoint packet flow PF are injected into the network 100.

[0068] More specifically, each packet Pk of the multipoint packet flow PF includes a first marking bit B1 and a second marking bit B2, such as Figure 4 The value of each marker bit B1, B2 can be set to any one of two alternative marker values, namely 1 and 0. The marker bits B1, B2 are preferably included in the packet header H.

[0069] Preferably, in the packets Pk of the K point-to-point packet flows of the multipoint packet flow PF, the value of the first marking bit B1 switches alternately between two alternative marking values ​​(e.g., 1 and 0). More preferably, the value of the first marking bit B1 switches periodically between the two alternative marking values ​​(e.g., 1 and 0) at a period T, which will be referred to as the "marking period" hereinafter. The marking period T can be set by the network operator according to the desired time measurement rate (as will be described in detail below, the marking period T is also the measurement period). For example, the marking period T can be equal to 5 minutes.

[0070] The switching of the value of the first marking bit B1 is substantially synchronized for all K point-to-point packet flows of the multipoint packet flow PF, i.e., the switching is performed substantially simultaneously (i.e., with a maximum mismatch of T / 2) for all K point-to-point packet flows of the multipoint packet flow PF. Thus, the packets Pk of the multipoint packet flow PF transmitted through the network 100 during a particular marking period have their first marking bit B1 set to substantially the same marking value 1 or 0.

[0071] For the second marking bit B2, its value is preferably set to be equal to a marking value indicating that the packet Pk will be subjected to sample measurement (also called "sampling value", such as 0) or a marking value indicating that the packet Pk will not be subjected to sample measurement (also called "non-sampling value", such as 1).

[0072] According to a first embodiment, in the packets Pk of each point-to-point packet flow PF1, PF2, ... PF5 for which sample measurement is required, the second marking bit B2 is preferably set to a non-sampled value for all packets Pk, except that the second bit B2 of one packet Pk in each marking period T is set equal to the sampled value. Preferably, only one packet Pk in each marking period T has its second bit B2 set to the sampled value. However, according to some variants, two or more packets Pk in each sampling period T may have their second bit B2 set to the sampled value.

[0073] Figure 5 The diagram schematically shows the values ​​of the first marker bit B1 and the second marker bit B2 in packets of a point-to-point packet flow that requires sample measurement and is included in a multipoint packet flow PF according to the first embodiment of the present invention.

[0074] The first marking bit B1 is periodically switched between 1 and 0 with a marking period T, thereby implementing an alternating marking technique, whereby the traffic is divided into alternating blocks BL1, BL2, BL3, BL4, ... of duration T, as specified by RFC 8321 and RFC 8889. The second bit B2 is equal to the non-sampled value 1 for all packets Pk, except for one packet in each marking period (as a non-limiting example, the first packet of each block BL1, BL2, BL3, BL4, ...). Only for those packets Pk, the second marking bit B2 is set equal to the sampled value 0.

[0075] For the point-to-point packet flow of the multipoint packet flow PF that does not require sample measurement, the second marker bit B2 of the packet Pk is changed to be fixed equal to a non-sampled value (eg, 1).

[0076] Preferably, the monitoring network is implemented on the path of the packets Pk of the multipoint packet flow PF marked as described above. The monitoring network includes at least one measurement point implemented on the path of the packets Pk of the multipoint packet flow PF marked as described above. More preferably, the monitoring network includes at least two measurement points implemented on the path of the packets Pk of the multipoint packet flow PF marked as described above. Each measurement point can be embedded in a corresponding node or implemented as a separate machine connected to a corresponding node.

[0077] In order to be able to perform multipoint measurements on a multipoint packet flow PF, the measurement points of the monitoring network are preferably arranged into clusters, which are defined as a group of measurement points that exhibit the same property that if no packet loss occurs, the set of packets Pk of the multipoint packet flow PF received at the (multiple) input measurement points of the cluster is the same as the set of packets Pk of the multipoint packet flow PF received at the (multiple) output measurement points of the cluster. In other words, if no packet loss occurs, each packet Pk of the multipoint packet flow PF received by any one of the cluster input measurement points is also received at one of the cluster output measurement points.

[0078] For example, a cluster may include an input measurement point at each of the N source nodes and an output measurement point at each of the M destination nodes of the multipoint packet flow PF. Figure 2 As shown, a measuring point MP1 is implemented at a source node N1, and measuring points MP5, MP6, MP7, MP8 are implemented at destination nodes N5, N6, N7, N8.

[0079] If one or more measurement points ( Figure 2 (not shown in the figure), other clusters of measuring points can be defined, as described in EP3449602 in the name of the same applicant.

[0080] Preferably, the monitoring network is also equipped with a management server. For simplicity, Figure 2 The management server is not shown in FIG. The management server may be a stand-alone server connected to any node. Alternatively, the management server may be implemented on any node. The management server preferably cooperates with the measurement points to collect performance parameters therefrom, which will be described in detail below.

[0081] Each measurement point of the monitoring network is preferably configured to provide performance parameters related to a packet Pk of a multipoint packet flow PF, so as to enable performance measurements to be made, the performance measurements comprising at least one of a sample performance measurement of an individual packet Pk of at least one of the point-to-point packet flows PF1, PF2, ... PF5 and a multipoint performance measurement of the multipoint packet flow PF as a whole, as will be described hereinafter with reference to Figure 6 The flowchart is described in detail.

[0082] Each measurement point MPi (i=1, 2, . . . 8) receives all traffic (or its copy) received at the node where it is implemented or connected (step 600).

[0083] For each received packet Pk, the measurement point MPi first determines whether the packet Pk belongs to a point-to-point packet flow requiring sample measurement (step 601). To perform step 601, the measurement point MPi preferably applies ACL (Access Command List) technology.

[0084] If the packet Pk belongs to a point-to-point packet flow requiring sample measurement, the measuring point MPi determines whether the packet Pk is to be subjected to sample measurement, preferably based on the value of its second marker bit B2 (step 602).

[0085] If the second marker bit B2 of the identified packet Pk is equal to the sample value (eg 0), the measurement point MPi determines that the packet Pk is to be subjected to sample measurement (step 603).

[0086] Thus, the measuring point MPi provides sample performance parameters related to the packet Pk, based on which sample performance parameters a sample measurement can be provided (step 604). For example, in step 604, the measuring point MPi can provide a timestamp related to the packet Pk, also taking into account the value of the first marker bit B1 on which the alternating marking is implemented, as specified by chapter 3.3.2 of the above-mentioned RFC 8321 and also described in EP3085019 in the name of the same applicant.

[0087] Otherwise, if the second marker bit B2 of packet Pk is equal to a non-sampled value (eg 1), the measurement point MPi determines that the packet will not be subjected to sample measurement (step 605) and accordingly does not generate any sample performance parameters associated therewith.

[0088] Regardless of whether packet Pk belongs to a point-to-point packet flow requiring sample measurement (and then regardless of whether steps 602-605 have been performed on packet Pk), the measuring point MPi preferably determines again based on the value of its second marker bit B2 whether packet Pk is to be considered for multi-point measurement (step 606).

[0089] Specifically, if the second marker bit B2 is equal to the sampled value (eg, 0), the measurement point MPi determines that the packet Pk does not need to be considered for multipoint measurement (step 607). Then, the measurement point MPi takes no further action related to the packet Pk.

[0090] If, on the other hand, the second marking bit B2 is equal to a non-sampled value (eg 1), the measuring point MPi determines that the packet Pk has to be considered for the multipoint measurement (step 608).

[0091] The measuring point MPi accordingly updates at least one cumulative performance parameter related to the multipoint packet flow PF as a whole (step 609). For example, in step 609, the measuring point MPi may update a cumulative counter and / or a cumulative timestamp related to the multipoint packet flow PF as a whole, also taking into account the value of the first marking bit B1 on which the alternating marking is implemented, as specified by the above-mentioned RFC 8889 and also described in EP3449602 in the name of the same applicant.

[0092] Then, the measurement point MPi may send the value of at least one performance parameter to the management server (step 610), which will use the value to provide performance measurement. The at least one performance parameter sent in step 610 preferably includes the value of at least one cumulative performance parameter that may be updated in step 609, and the value of the sample performance parameter generated in step 604 (if any).

[0093] According to a preferred embodiment, step 610 is performed at the end of each marking period T, as indicated by a toggle of the value of the first marking bit B1. At the end of a certain marking period, the measuring point MPi preferably sends to the management server the value reached by at least one accumulated performance parameter at the end of the marking period and the value(s) of the sample performance parameter(s) generated for the sample(s) detected during the marking period.

[0094] The management server will then provide at least one performance measurement using the at least one performance parameter received from the measurement point(s) of the monitoring network.

[0095] Specifically, if sample performance parameters are received from one or more measurement points on the path of a point-to-point packet flow of a multipoint packet flow PF, they can be used to provide at least one sample measurement related to the point-to-point packet flow, such as one-way delay or jitter of its sample packets. The at least one sample measurement may include, for example, a delay measurement as specified in Chapter 3.3.2 of the above-mentioned RFC 8321 and also described in EP3085019 under the name of the same applicant; and / or a measurement of delay variation (jitter) as specified in Chapter 3.4 of the above-mentioned RFC8321 and also described in EP3085019 under the name of the same applicant.

[0096] In addition, at least one accumulated performance parameter received from the input measurement point and the output measurement point of the cluster can be used to provide at least one multi-point performance measurement, such as a multi-point packet loss measurement or a multi-point average delay measurement. For example, the multi-point performance measurement may include a multi-point packet loss measurement and / or a multi-point average delay measurement as specified by the above-mentioned RFC 8889 and also described in EP3449602 in the name of the same applicant.

[0097] Therefore, based on different values ​​of the same bit (i.e., the second marking bit B2), the packets Pk to be subjected to sample measurement alone and the packets Pk to be considered for multi-point measurement are identified. Therefore, the packets Pk to be considered for multi-point measurement are identified without applying any address filtering techniques and without requiring any additional bits for this purpose. Since the second marking bit B2 merges two separate functions, only two bits B1, B2 are required to support all requested performance measurements based on alternating marking, i.e., sample measurement (e.g., delay or jitter measurement specified by RFC 8321) and multi-point measurement (e.g., multi-point measurement specified by RFC 8889).

[0098] It should be noted that, in general, only a small number of packets are required to perform sample measurements on a point-to-point packet flow, typically at most one per marking period T. Therefore, even assuming that sample measurements need to be performed on all point-to-point packet flows PF1, PF2, ... PF5 in a multipoint packet flow PF, the vast majority of packets of the multipoint packet flow PF still have their second marking bit B2 set equal to the non-sampled value. This advantageously results in the multipoint measurement being based on the vast majority of packets of the multipoint packet flow PF. As a result, an accurate multipoint measurement can be provided.

[0099] According to an advantageous variant, in order to determine whether a packet Pk should be taken into account for multipoint measurement, the measuring point MPi takes into account the values ​​of both marking bits B1 and B2 .

[0100] Specifically, refer to Figure 7 Flow chart of steps 600-605 and Figure 6 The steps 600-605 of the flowchart of FIG. 1 are the same as those of FIG. 1. Therefore, the detailed description of these steps will not be repeated.

[0101] As in Figure 6 As in the flowchart, according to Figure 7 Alternatively, the measuring point MPi preferably determines whether it is to be considered for the multipoint measurement by first considering the value of the second marker bit B2 of the packet Pk (step 606).

[0102] If the second marker bit B2 is equal to the sampled value (e.g. 0), then according to this variant, the measuring point MPi reads the value of the first marker bit B1 (step 606a). If the first marker bit B1 is also equal to the sampled value (e.g. 0), the measuring point MPi determines that the packet Pk does not need to be considered for multipoint measurement (step 607).

[0103] If, on the other hand, the first marker bit B1 is equal to a non-sampled value (eg 1), the measuring point MPi recognizes that the packet Pk must be considered for the multipoint measurement (step 608). Figure 6If the second flag bit B2 is equal to the non-sampled value, the measuring point MPi also makes the same determination (step 608). In this case, the measuring point MPi does not need to read the value of the first flag bit B1. Then, as described above, the measuring point MPi updates at least one accumulated performance parameter related to the multipoint packet flow PF as a whole (step 609).

[0104] Then, the measurement point MPi sends the value of at least one performance parameter to the management server (step 610), which will use the value to provide performance measurement, as described above in conjunction with Figure 6 The flowchart is described.

[0105] Therefore, according to this variant, the measurement point MPi considers for the multipoint measurement not only the packets Pk whose second marker bit B2 is equal to the non-sampled value, but also the packets Pk whose second marker bit B2 is equal to the sampled value, provided that their first marker bit B1 is equal to the non-sampled value (for example, 1). Therefore, only the packets Pk having both the first marker bit B1 and the second marker bit B2 equal to the sampled value (for example, 0) are excluded from the multipoint measurement.

[0106] For example, refer to Figure 5 According to this advantageous variant, only the first packet of block BL2 and the first packet of block BL4, whose first bit B1 and second bit B2 are both equal to the sample value 0, are excluded from the multipoint measurement. This advantageously leads to performing the multipoint measurement on a larger part of the packets Pk of the multipoint packet flow PF and thus produces more accurate measurements.

[0107] According to another embodiment, in the packets Pk requiring sample measurement of each point-to-point packet flow PF1, PF2, ... PF5, the second marker bit B2 is preferably set to a non-sampled value (e.g., 1) for all packets Pk, but with the following exception: only for a marking period T in which the first marker bit B1 is equal to a non-sampled value (e.g., 1), the second marker bit B2 of one packet Pk is set equal to a sampled value (e.g., 0). On the contrary, for a marking period T in which the first marker bit B1 is equal to a sampled value (e.g., 0), the second bit B2 is set to a non-sampled value (e.g., 1) for all packets Pk. Preferably, in each marking period T in which the first marker bit B1 is equal to a non-sampled value (e.g., 1), only one packet Pk sets its second bit B2 to a sampled value. However, according to some variants, during each sampling period T in which the first marker bit B1 is equal to a non-sampled value (e.g., 1), two or more packets Pk may set their second bit B2 to a sampled value.

[0108] Figure 8The values ​​of the first marker bit B1 and the second marker bit B2 in the packets of the point-to-point packet flow included in the multipoint packet flow PF requiring sample measurement according to this embodiment of the present invention are schematically shown.

[0109] The first marking bit B1 is periodically switched between 1 and 0 according to a marking period T, thereby implementing an alternating marking technique, whereby the traffic is divided into alternating blocks BL1, BL2, BL3, BL4, ... of duration T, as specified by both RFC 8321 and RFC 8889. The second bit B2 is equal to the non-sampled value 1 for all packets Pk, except for one packet in each marking period in which the first marking bit B1 is equal to the non-sampled value 1. Only for those packets Pk, the second marking bit B2 is set equal to the sampled value 0. For the marking period in which the first bit B1 is equal to the sampled value 0, the second marking bit B2 of all packets Pk is set equal to the non-sampled value 1. Therefore, as Figure 8 As shown, the first and third blocks BL1, BL3 are composed of packets whose first marking bit B1 is equal to the non-sampled value 1, so one packet of each of them (the first one as a non-limiting example) has the second marking bit B2 set to the sampled value 0. In contrast, for the second and fourth blocks BL2 and BL4, the packets Pk included therein have their first marking bit B1 equal to the sampled value 0. Therefore, all packets Pk of these blocks have their second marking bit B2 fixedly set to be equal to the non-sampled value 1.

[0110] It can be understood that when the measurement point MPi is based on Figure 6 When the flow chart is operated and when the measuring point MPi is based on Figure 7 When operating according to the flowchart Figure 8 The marking scheme of the embodiment further increases the number of groups Pk considered for multi-point measurement.

[0111] By Figure 6 The flowchart is applied to Figure 8 In fact, the group Pk whose second marking bit B2 is equal to the sampling value is identified as the group to be measured by sample, while the group Pk whose second marking bit B2 is equal to the non-sampling value is identified as the group to be considered for multi-point measurement. Figure 5 This results in halving the number of packets Pk subjected to sample measurement compared to the marking scheme of , but on the other hand, increases the number of packets considered for multipoint measurement. Figure 5 In the marking scheme of , for multi-point measurement, one packet Pk is ignored in each marking period T, and Figure 8In the marking scheme of , only one packet Pk in half of the marking period T is ignored, i.e., those marking periods during which the first marking bit B1 is equal to the non-sampled value 1 (see the first packet of block BL1 and the first packet of block BL3). For the other marking periods in which the first marking bit B1 is equal to the sampled value 0, all packets Pk are considered for multi-point measurement.

[0112] In addition, by Figure 7 The flowchart is applied to Figure 8 With the marking scheme of the multipoint measurement, all packets Pk are identified as packets to be considered for the multipoint measurement. Only packets Pk whose second marking bit B2 is equal to the sampled value 0 do have a first marking bit B1 equal to the non-sampled value 1. This means that no packet Pk has both marking bits B1 and B2 equal to the sampled value 0, and therefore no packet is ignored for the multipoint measurement. Therefore, the resulting multipoint measurement is even more accurate.

[0113] The above-described operation of the measurement points MPi advantageously allows providing different types of sample measurements and multipoint measurements on the multipoint packet flow PF and its K point-to-point packet flows.

[0114] For example, refer to Figure 2 In the exemplary network 100 of FIG. 1 , based on the input cumulative counter provided by the measuring point MP1 and the output cumulative counter provided by the measuring points MP5, MP6, MP7 and MP8, the multipoint packet loss affecting the multipoint packet flow PF can be calculated as the difference between the sum of the input cumulative counter and the sum of the output cumulative counters, as described in EP3449602 in the name of the same applicant. The commensurability of the input cumulative counter and the output cumulative counter is ensured by the fact that the counters only consider packets entering and leaving the monitored network, wherein each measuring point MPi is based on the value of the second marker bit B2 as described above (see Figure 6 ), optionally by also taking into account the value of the first marking bit B1 (see Figure 7 to distinguish them from packets originating from within the monitoring network.

[0115] Other multi-point measurements are also possible. For example, if other measurement points are implemented at intermediate locations of the network 100, one or more clusters may be identified in the monitoring network, as described in EP3449602 in the name of the same applicant. Accordingly, multi-point measurements associated with each individual cluster of the monitoring network may be provided.

[0116] As for sample measurement, based on the input timestamps provided by the measurement point MP1 for a sample sequence identified in the point-to-point packet flow PF1 and the output timestamps provided by the measurement point MP5 for the same samples, a one-way delay measurement and / or jitter measurement of the point-to-point packet flow PF1 can be provided, as disclosed in EP3085019 in the name of the same applicant. Similarly, based on the input timestamps provided by the measurement point MP1 for a sample sequence identified in the point-to-point packet flow PF3 and the output timestamps provided by the measurement point MP6 for the same samples, a one-way delay measurement and / or jitter measurement of the point-to-point packet flow PF3 can be provided, as disclosed in EP3085019 in the name of the same applicant. The same applies to the point-to-point packet flows PF2, PF4 and PF5.

[0117] Of course, other types of performance measurements can also be performed on the K point-to-point packet flows of the multipoint packet flow PF, such as packet loss measurement or average delay measurement. In order to support these types of measurements, the measurement point MPi only needs to be based on, for example, ACL technology (see Figure 6 and 7 In step 601 of the embodiment of the present invention, the packet Pk of the point-to-point packet stream to be measured is identified, and then the value of its first marker bit B1 is read in order to identify the alternating blocks. For this type of measurement, it is not necessary to read the second marker bit B2, which is dedicated to identifying the packets to be subjected to sample measurement and the packets to be considered for multipoint measurement.

[0118] According to a further embodiment, the sampled value and the non-sampled value of the second marking bit B2 may be periodically inverted at each marking period T.

[0119] For example, Fig. 9 As shown, according to a further embodiment, the sampled value and the non-sampled value of the second marking bit B2 can be periodically reversed in each marking period T to ensure that during each marking period T, the sampled value of the second marking bit B2 is equal to the current value of the first marking bit B1. Fig. 9 As shown, in fact, during the first and third marking periods (blocks BL1 and BL3), the value of the first marking bit B1 is 1, and accordingly, the sampled value of the second marking bit B2 is also equal to 1, and the non-sampled value of the second marking bit B2 is equal to 0. On the contrary, during the second and fourth marking periods (blocks BL2 and BL4), the value of the first marking bit B1 is 0, so the sampled value of the second marking bit B2 is also equal to 0, and the non-sampled value of the second marking bit B2 is equal to 1.

[0120] According to these further embodiments, for each marking period T, the second marking bit B2 in one (or more) packets Pk may be set equal to the sampled value, such as Fig. 9 Alternatively, Fig.10As shown, in one (or more) packets Pk of each marking period T in which the first marking bit B1 is equal to one of its values ​​(eg, 1), the second marking bit B2 may be set equal to the sampled value, such as Fig.10 shown.

[0121] Also according to these variants, each measurement point MPi preferably identifies the packets Pk to be subjected to sample measurement as packets whose second marker bit B2 is equal to the sample value, such as Figure 6 and 7 As shown in the flowchart above and combined with Figure 5 and 8 However, since the sampled values ​​change periodically in each marking period T, the measuring point MPi should be able to determine which is the current sampled value. To this end, for example, the measuring point MPi can consider the current value of the first marking bit B1 and determine that the current sampled value is equal to the current value of the first marking bit B1.

[0122] This also applies to the identification of packets Pk to be considered for multipoint measurement. Each measuring point MPi indeed preferably identifies the packets Pk to be considered for multipoint measurement as those packets whose second marker bit B2 is equal to a non-sampled value, such as Figure 6 As shown in the flowchart of , the value of the first flag bit B1 may also be considered, such as Figure 7 As shown in the flowchart above Figure 5 and Figure 8 However, since the non-sampled value changes periodically at each marking period T, the measuring point MPi should be able to determine which is the current non-sampled value. To this end, for example, the measuring point MPi can consider the current value of the first marking bit B1 and determine that the current non-sampled value is different from the current value of the first marking bit B1.

Claims

1. A method for implementing performance measurement of a multipoint packet flow (PF) conveying live traffic through a packet-switched communication network (100), each packet (Pk) of the multipoint packet flow (PF) comprising a first marker bit (B1) and a second marker bit (B2), the method comprising: a) before transmitting each packet (Pk) of the multipoint packet flow (PF) through the network (100), alternately switching the value of the first marking bit (B1) between two alternative marking values; and setting the second flag bit (B2) to a first value indicating that the packet (Pk) is to be subjected to a sample performance measurement or to a second value indicating that the packet (Pk) is not to be subjected to the sample performance measurement; as well as b) at at least one measurement point (MPi) placed on the path of said packet (Pk), reading said second marker bit (B2) and: - if said second marking bit (B2) is set to said first value, generating a sample performance parameter associated with said packet (Pk); and - if said second marking bit (B2) is set to said second value, updating a cumulative performance parameter related to said multipoint packet flow (PF) as a whole.

2. The method according to claim 1, wherein in step a), the two alternative marking values ​​are equal to the first value indicating that the group (Pk) will be subjected to sample performance measurement and the second value indicating that the group (Pk) will not be subjected to the sample performance measurement.

3. A method according to claim 1 or 2, wherein in step a), the second marking bit (B2) in at least one packet (Pk) of the point-to-point packet stream (PF1, PF2, ... PF5) included in the multi-point packet stream (PF) is set equal to the first value indicating that the packet (Pk) will be subjected to sample performance measurement, and the second marking bit (B2) of other packets (Pk) of the point-to-point packet stream (PF1, PF2, ... PF5) is set equal to the second value indicating that the packet (Pk) will not be subjected to the sample performance measurement.

4. The method according to any one of the preceding claims, wherein in step a), the method comprises: Before each packet of the multipoint packet flow is transmitted through the network, the value of the first marking bit is periodically switched between two alternative marking values ​​according to a marking period T.

5. A method according to claim 4, wherein in step a), according to a marking period T, the second marking bit (B2) in a packet (Pk) of the point-to-point packet flow (PF1, PF2, ... PF5) is set equal to the first value indicating that the packet (Pk) will undergo sample performance measurement.

6. The method according to claim 4, wherein in step a), only when the first marking bit (B1) during the marking period T is set to be equal to a predetermined marking value among the two alternative marking values, the second marking bit (B2) in a packet (Pk) of the point-to-point packet flow (PF1, PF2, ... PF5) according to the marking period T is set to be equal to the first value indicating that the packet (Pk) will undergo sample performance measurement.

7. The method of claim 6, wherein the predetermined one of the two candidate marking values ​​is equal to the second value indicating that the packet (Pk) is not to be subjected to the sample performance measurement.

8. A method according to any one of the preceding claims, wherein: Step b) comprises: if the second marking bit (B2) is set to the first value indicating that the packet (Pk) will be subjected to sample performance measurement, and the first marking bit (B1) is equal to the second value indicating that the packet (Pk) will not be subjected to the sample performance measurement, then the accumulated performance parameter related to the multi-point packet flow (PF) as a whole is also updated.

9. The method according to any of the preceding claims, wherein the method further comprises periodically exchanging, according to the marking period T, the first value indicating that the packet (Pk) will be subjected to sample performance measurement and the second value indicating that the packet (Pk) will not be subjected to the sample performance measurement.

10. Method according to any of the preceding claims, wherein the sample performance parameters relating to the packet (Pk) comprise a timestamp indicating a reception time of the packet (Pk) at the at least one measurement point (MPi).

11. A method according to any of the preceding claims, wherein the cumulative performance parameters include a cumulative counter counting the packets of the multi-point packet flow (PF) received at the at least one measurement point (MPi) and / or a cumulative or average timestamp indicating the cumulative or average reception time of the packets of the multi-point packet flow (PF) at the at least one measurement point (MPi).

12. A method according to any of the preceding claims, wherein in step b), the at least one measurement point (MPi) determines whether the packet (Pk) belongs to a point-to-point packet flow (PF1, PF2, ... PF5) requiring the sample performance measurement by applying an access command list technique.

13. A method according to any of the preceding claims, wherein in step a), the multipoint packet flow (PF) is injected into the network (100) via at least two source nodes, and the periodic switching of the value of the first marking bit (B1) is performed by the at least two source nodes in a substantially synchronized manner.

14. A method for performing performance measurement in a packet-switched communication network, the method comprising the steps of the method for implementing performance measurement according to any one of claims 1 to 13, and further comprising: c) performing performance measurements, including at least one of sample measurements related to said group (Pk) based on said sample performance parameters and multi-point performance measurements based on said accumulated performance parameters.

15. The method according to claim 14, wherein step b) further comprises sending at least one of the sample performance parameter and the accumulated performance parameter to a management server cooperating with the at least one measurement point (MPi), and step c) is performed by the management server.

16. A packet-switched communication network (100) supporting transmission of a multipoint packet flow (PF) carrying live traffic, each packet (Pk) of the multipoint packet flow (PF) comprising a first marking bit (B1) and a second marking bit (B2), the packet-switched communication network (100) comprising: a) at least one node (N1) configured to, before transmitting each packet (Pk) of the multipoint packet flow (PF) through the network (100), alternately switch the value of the first marking bit (B1) between two alternative marking values; and set the second marking bit (B2) to a first value indicating that the packet (Pk) will be subjected to a sample performance measurement or to a second value indicating that the packet (Pk) will not be subjected to the sample performance measurement; as well as b) at least one measurement point (MPi) placed on the path of said packet (Pk), said at least one measurement point (MPi) being configured to read said second marking bit (B2) and: - if said second marking bit (B2) is set to said first value, generating a sample performance parameter associated with said packet (Pk); and - if said second marking bit (B2) is set to said second value, updating a cumulative performance parameter related to said multipoint packet flow (PF) as a whole.

17. The packet-switched communications network of claim 16, further comprising: c) A management server configured to perform performance measurements, including at least one of sample measurements related to the group (Pk) based on the sample performance parameters and multi-point performance measurements based on the accumulated performance parameters.

Images