WO2007006194A1

WO2007006194A1 - A method and a device of detecting the quality of service in a communication network

Info

Publication number: WO2007006194A1
Application number: PCT/CN2006/000968
Authority: WO
Inventors: Cheng Chen
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2005-07-14
Filing date: 2006-05-15
Publication date: 2007-01-18
Also published as: CN1897547A; CN100459528C

Abstract

A method of detecting the quality of service in a communication network includes: A. the first point of the two address points in the communication network transmits the predetermined number of test packets to the second point, after the second point has received the test packets, it returns a corresponding test response packets; B. the first point determines the QoS parameter between of the two address points according to the return time and/or return quality of test response packets received form the second point. A device of detecting the quality of service in a communication network includes: a transceiving unit for transceiving the test packets, and a QoS parameter calculating unit for determining the QoS parameter between of the two address points according to the return time and/or return quality of test response packets of the opposite point. The invention not only automatically detects the onoff fault of the network, but also detects the QoS of the network, and in time switches the master network and the standby network to ensure the service on the rails; and automatically detects and orientates the fault point and the device and line which reduces the QoS to inform user to repair it.

Description

Method and device for detecting quality of service in communication network

The invention relates to a Qos (Quality Of Service) detection technology in the field of data communication, in particular to a Qos detection between a router, a Layer 3 switch or a Next Generation Network (NGN) device capable of outputting an IP interface in a communication network. Method and detection device. Background of the invention

In the field of data communication, Qos is used to measure the quality of IP networks, and is generally measured by indicators such as packet loss rate, jitter, delay, error, and out of order. Qos is generally determined by the devices and links on the transmission path. Congestion of routing devices may cause packet loss, delay, and jitter (delay and jitter if there is no congestion); simultaneous transmission of the same session on multiple paths It may cause out-of-order; the transmission link may cause bit errors and packet loss due to interference, signal attenuation, etc.

In the traditional data communication field, devices are connected by mesh. IP packets arrive from one device to another device. There are multiple reachable paths. The devices use IP routing to determine which IP packets are forwarded to that link. A router, unless the link or router in the routing direction fails, will cause the route to re-converge and select other routing directions to forward the packet. Otherwise, even if the QoS in the routing direction is poor (if there are a large number of packet loss and error) , and will not reroute to select other IP routes. In the field of telecommunications, especially in the field of next-generation networks (NGN), voice services require high quality of service for IP networks (generally, packet loss rate = 1%, network jitter = 20ms, delay = 100ms is poor) Network environment), if there is a serious packet loss, delay or jitter in the IP network, the quality of the call will be seriously affected. The current data communication network uses Diff-Serv, multi-protocol label switching - virtual private network (MPLS-VPN), virtual local area network Technology such as (VLAN) guarantees the QoS of the network, but it can only do QoS guarantee on the basis of existing resources, and cannot effectively solve the problems of link quality degradation and congestion. Especially for the case of active/standby networking, 'Ethernet networking as shown in Figure 1:

In general, when the primary Ethernet interface is faulty, if the standby Ethernet interface is normal, it is switched to the standby interface. However, if the backup interface is not interrupted and the line has a serious signal attenuation, if the repair is not detected in time, the service is switched to the backup link. Because the Qos is poor, the service is seriously affected. There are many situations in which the Qos needs to be detected. Therefore, a need is needed. A method for automatic detection and location of QoS, in order to detect network quality problems in time, and to locate which part of the link or which router has experienced a decline in service quality, and measures can be taken to correct it.

Existing data communication networks generally use both ping technology and route tracking (TraceRt) technology to detect and locate faults. The Pirig method can detect whether two IP addresses can communicate with each other. When the connection fails, the TraceRt method locates the fault on that router.

The Ping and TraceRt methods are available on general routers and Layer 3 Ethernet switches and are based on the Internet Control Message Protocol (ICMP) protocol defined by the standard RFC0792. The general usage method is that when a user finds that a certain network is unreachable, it first uses Ping to detect and then uses TraceRt to locate the fault. However, the prior art has the following disadvantages:

(1) The existing technology only gives the detection and positioning method for the continuity of the network, but the problem of poor service quality cannot be detected and located;

(2) Detection needs to rely on manual operation by the user, and cannot be automatically detected and located. Due to the sudden burst of IP network traffic, a large number of faults may occur and the network quality returns to normal. Summary of the invention

The object of the present invention is to provide a method and a detection device for detecting Qos in a communication network, which can automatically detect the connectivity and Qos quality of a communication path between any two addresses in a communication network.

The object of the invention is achieved in this way:

A method for detecting quality of service Qos in a communication network, comprising:

A. The first end of the two address endpoints in the communication network sends a predetermined number of detection packets to the second end, and the second end returns a corresponding detection response packet after receiving the detection packet;

B. The first end determines a Qos parameter between the two addresses according to the received return time and/or return quality of the second end detection response packet.

Preferably, the communication network has an active/standby interface and is connected to a third-party interface. When one of the primary and secondary interfaces performs service processing, step A and step B are performed between the primary and secondary interfaces and the third-party interface to detect the primary The QoS parameter between the interface and the third-party interface of the backup interface reaches the preset Qos parameter threshold between the interface of the current service interface and the third-party interface, and the interface between the other interface and the third interface When the Qos parameter does not reach the threshold of the threshold, the master/slave interface is switched.

Preferably, the method further comprises: setting a threshold value of the Qos parameter, and triggering an alarm when a Qos parameter between the two address endpoints reaches the threshold.

Preferably, the Qos parameter includes any one or a combination of the number of lost packets, the number of errored packets, the transmission delay, the jitter delay, the percentage of packet loss, the average transmission delay, and the average jitter delay.

Preferably, the Qos parameter includes the number of lost packets, and the method for determining the number of lost packets between the two address endpoints is: after the first end sends a detection packet, if the detection packet is not received within a predetermined time Corresponding detection response packet, then 'accumulate the number of lost packets.

Preferably, the Qos parameter further includes a percentage of packet loss, and the percentage of the packet loss is determined. The specific method is: dividing the accumulated number of lost packets by the total number of detected packets sent.

Preferably, the Qos parameter includes the number of error packets, and the method for determining the number of error packets between the two address endpoints is: after the first end sends a detection packet, if the detection is received within a predetermined time If the packet corresponding to the detection response packet is detected, it is further detected whether the content of the detection response packet is incorrect. When an error is detected, the number of error packets is accumulated.

Preferably, the Qos parameter includes a percentage error, and the specific method for determining the error percentage is: dividing the accumulated number of error packets by the total number of detected packets.

Preferably, the Qos parameter includes a transmission delay, and the method for determining a transmission delay between two address endpoints is: after the first end sends a detection packet, if the detection packet corresponding to the detection packet is received within a predetermined time The response packet further detects whether the content of the detection response packet is incorrect. When the content of the detection response packet is correct, the time difference from the transmission of the detection packet to the receipt of the corresponding detection response packet is determined as the transmission time of the detection packet. Delay.

Preferably, the Qos parameter includes an average transmission delay, and the specific method for determining the average transmission delay is: dividing the total of the determined transmission delays by the total number of correct response packets of the content.

Preferably, the Qos parameter includes a jitter delay, and the method for determining a jitter delay between two address endpoints is: after transmitting the detection packet twice, and determining the transmission delay of the two detection packets, The time difference between the transmission delays of the two detection packets is taken as a jitter delay.

Preferably, the Qos parameter between the two addresses includes an average jitter delay, and the method for determining an average jitter delay is: dividing the total of the determined jitter delay by the total number of detection response packets with the correct content. .

Preferably, whether the content of the detection response packet is incorrect is: detecting whether the content of the media access control layer, the IP protocol layer, and the Internet Control Message Protocol layer of the detection response packet is incorrect.

Preferably, the detection period is preset, and the step A and the step A are repeatedly performed in each detection period. Step B.

Preferably, the address is an interface address or an IP address.

Preferably, the detection packet is a Ping detection packet, an ARP protocol packet, an echo packet, or a link control protocol negotiation packet.

Preferably, the path between the two addresses includes at least one routing node, and the method further includes a process for detecting a Qos parameter of each segment, specifically:

Cl, the first end sequentially sends detection packets numbered 1, 2, the source address of each detection packet is a first endpoint, and the destination address is a second endpoint;

C2: Send a predetermined number of times for each number of detection packets; the routing node that receives the detection packet decrements the number by 1, and determines whether the number is 0, if yes, discards the detection packet, and returns a detection response to the first endpoint. a packet, including the address of the routing node; if the number is not 0, forwarding the detection packet to the next routing node, repeating step C2 until the detection packet reaches the second endpoint, from the second endpoint to the first end Point return detection response packet, including the address of the second endpoint;

C3. The first endpoint determines a Qos parameter between the first endpoint and each routing node according to the received return time and/or return quality of the detected response packet returned by each routing node.

Preferably, the method further includes: setting a threshold value of the Qos parameter, determining whether the Qos parameter between the first endpoint and each routing node reaches a threshold of the Qos parameter, and if so, stopping the pair of segments Detecting the Qos parameter of the path, and positioning the fault point and/or the Qos falling point between the corresponding routing nodes; otherwise, if the detection response packet of the second endpoint is received, determining the path detection between the two address endpoints After completion, the Qos parameter detection for each segment path is stopped.

As another aspect of the present invention, a device for detecting Qos in a communication network is disclosed, the device comprising: The detection packet transceiver unit is disposed on the detected endpoint in the communication network, and is configured to send a predetermined number of detection packets to the opposite end interval, return a detection response packet to the opposite end after receiving the detection packet, and receive the same. The detection response packet returned by the peer end;

The Qos parameter calculation unit is connected to the detection packet transceiver unit, and determines a Qos parameter between the two detected endpoints according to a return time and/or a return shield amount of the detection response packet of the opposite end received by the detection packet transceiver unit.

Preferably, the detecting device further includes an alarm unit, configured to set a threshold value of the Qos parameter, determine whether the Qos parameter determined by the Qos parameter calculation unit reaches the threshold, and if so, trigger an alarm.

Preferably, the communication network has a primary and secondary interfaces and is connected to the third interface; the detecting device is configured on the primary and secondary interfaces and the third interface, and is configured to detect Qos parameters of the primary and secondary interfaces and the third interface; And further comprising a switching unit, configured to be connected to the alarm unit, configured to trigger an alarm when the QoS parameter between the interface currently processing the service and the third interface in the active/standby interface, and the Qos parameter between the other interface and the third interface is not When the alarm is triggered, the active and standby interfaces are switched.

It can be seen from the above that the Qos detection method and the detection device provided by the present invention can automatically detect the faults and QoS values between any two addresses periodically, and can not only detect the on/off status of the network, but also detect the QoS of the network. If the network is faulty or the QoS value exceeds the threshold, you can also perform an alarm. If the active/standby interface is available, you can perform the active/standby switchover based on the fault and the QoS status to protect the service from being affected. Notify the user to maintain and automatically detect the point of failure. In addition, it can automatically detect the Qos from the source IP address to each routing node on the path, and record it as the basis for user fault location and Qos drop positioning. The present invention can be applied to a commonly used data communication network without special requirements for devices such as transmission links and routers.

The QoS of the QoS is high in the communication network. The present invention automatically detects the QoS of the IP bearer network, so as to timely protect the normal operation of the active/standby switchover and prompt the user to repair; Automatic detection is caused by the node or that link at the moment when the Qos quality degradation is reduced, so that the user can pinpoint the problem. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of a network structure including an active/standby interface in an Ethernet network;

2 is a flow chart of an embodiment of the method of the present invention;

3 is a schematic diagram of positioning a routing node fault according to the present invention;

4 is a schematic structural view of a detecting device according to the present invention;

FIG. 5 is a schematic diagram of a networking structure of the device according to the present invention applied to an Ethernet network including active and standby devices. Mode for Carrying Out the Invention The present invention will be described in detail below by way of specific examples and the accompanying drawings.

The core technical solution of the present invention is: A method for detecting a quality of service Qos in a communication network, comprising: A. The first end of the two address endpoints in the communication network sends a predetermined number of detection packets to the second end, and the second end After receiving the detection packet, a corresponding detection response packet is returned; B. The first end determines a Qos parameter between the two addresses according to the received return time and/or return quality of the second response detection packet.

The method of the present invention is suitable for a general IP network, and is particularly suitable for an NGN network, and has no special requirements for devices such as transmission links and routers.

Before the fault and QoS detection process begins, the user needs to configure the parameters that need to be used to automatically detect the fault and the QoS value. The parameters required to be configured are as follows: The Qos detection must be configured with the parameters including the source endpoint address between the detected two-point devices ( That is, the first endpoint address) and the destination endpoint address (ie, the second endpoint address, where the address may be an interface address or an IP address), a predetermined number of detection packets sent each time, and each detection packet The waiting response duration, the interval between each detection cycle, and the threshold of the Qos parameter.

The detection packet may be a Ping detection packet. ARP Association can also be used for Ethernet interfaces. The protocol packet detects the Qos parameter of the link between the two Layer 3 devices. The POS interface can also detect the between the two Layer 3 devices by using the PPP echo (Echo) packet and the PPP Link Control Protocol (LCP) negotiation packet. Qos of the link. If it is a Ping detection packet, the optional parameters configured above are optional parameters of the standard Ping command. In this embodiment, the Ping detection packet is taken as an example for description.

The Qos parameter includes any combination or one of a packet loss number, a number of error packets, a transmission delay, a jitter delay, a packet loss percentage, an average transmission delay, and an average jitter delay.

The threshold parameters of the Qos include the number of lost packets, the number of errored packets, the transmission delay, the threshold parameters of the jitter delay, and the threshold parameters of the percentage of packet loss, the threshold parameters of the delay average duration, and the threshold parameters of the jitter average duration. .

After the device is running normally, the detection of the network is started periodically, and the time of each detection is also determined. According to the configured parameters, the number of packets sent by each test determines the length of the test. The detection interval determines the time interval between the two tests. The waiting response time of each Ping detection packet is used to determine whether the response packet of the Ping detection packet can be received, that is, the Ping detection response packet.

2 is a flow chart of an embodiment of the method of the present invention. Referring to Figure 2, the process includes:

Step 201: The first end of the two address endpoints determines whether the time interval of each set detection period is up. If yes, step 202 is performed, otherwise, the process continues.

Step 202: The first end sends a Ping detection packet (the format of the Ping detection packet complies with the requirements of RFC0792), and records the time of transmission, and determines whether the Ping detection packet is received within a predetermined waiting period of each Ping detection packet. The Ping detection response packet (referred to as the response packet), if not, accumulates the number of packet loss, otherwise step 203 is performed.

Step 203: The first end receives the corresponding response packet within the waiting response time of each Ping detection packet, and detects whether the content of the response packet is correct, and specifically, the content of the MAC layer, the IP protocol layer, and the ICMP protocol layer may be performed. Detect, if not correct, accumulate the number of error packets, δ Otherwise, step 204 is performed;

Step 204: Determine, according to the time received by the response packet, a time difference from sending the detection packet to receiving the corresponding response packet as a transmission delay of the detection packet; further, detecting the packet every time, and determining the packet After the transmission delay of the two detection packets, the time difference of the transmission delays of the two detection packets is taken as a jitter delay. Accumulate the sum of the transmission delays and the sum of the accumulated jitter delays.

Step 205: Determine whether the number of sent Ping detection packets is greater than the number of detection detection packets per pre-configuration requirement. If not, return to step 202, otherwise, return to step 201.

After the detection is completed, the Qos parameters detected in the above steps can be further calculated to obtain a more optimized Qos parameter, and the calculation formulas of each parameter are as follows:

Packet loss percentage = number of lost packets I total number of detection packets;

Average delay = sum of delays I total number of correct response packets;

Average jitter duration = sum of jitter durations I Total number of correct response packets.

The first end may further determine whether the determined Qos parameter reaches the threshold according to a preset threshold of the Qos parameter, and if yes, trigger an alarm.

The above detection method can be applied to a communication network having an interface of the active and standby interfaces, for example, the communication network shown in FIG. 1. The access device has an active/standby interface, that is, a primary Ethernet interface and a backup Ethernet interface, and the active and standby interfaces are active and standby. The interfaces are respectively connected to the third-party interfaces of the Layer 3 devices. In this case, you need to configure an IP address for the active and standby interfaces, and use different MAC addresses. After the network access device runs normally, the Qos detection method described in Figure 2 is automatically run between the active and standby interfaces and the third-party interfaces. Detect the QoS parameter between itself and the connected three-party interface, and use the detection result as the basis for the active/standby switchover.

The detection result between the active/standby interface and its respective third-party interfaces (such as the number of lost packets, the number of errored packets, the transmission delay, the jitter delay, the percentage of packet loss, the average delay, and the average The jitter duration is recorded, and the Qos threshold parameters (including the packet loss threshold parameter, the error packet number threshold parameter, the transmission delay threshold parameter, the jitter delay threshold parameter, the threshold parameter of the packet loss percentage, and the delay parameter respectively) The threshold parameter of the average duration is compared with the threshold parameter of the average duration of the jitter. When the QoS parameter between the interface of the current processing interface and the third-party interface reaches the preset Qos parameter threshold, the other interface and the third interface are compared. If the QoS parameter between the interfaces does not reach the threshold, the active and standby interfaces are switched. For example, if the active/standby interface fails or the QoS parameters exceed the threshold, the percentage of detected packets is greater than the percentage of packet loss and error packets configured by the user. The average delay of detection and jitter is greater than the delay and jitter configured by the user. Average duration. In an extreme case, if the percentage of the packet loss is 100%, it is determined that an open circuit fault occurs between the corresponding primary interface or the backup interface and the third-party interface, and the primary and secondary interfaces are switched.

In addition to performing Qos detection on the primary and secondary interfaces, the present invention can also perform Qos detection on paths between non-active and standby interfaces and other interfaces. If the Qos parameter exceeds the threshold, Bay ij automatically alarms.

For example, the process of performing an active/standby switchover on the primary and backup interfaces, and performing the alarm on the non-active and standby interfaces may include the following steps:

Step 3: Record the test results in the system: so that the user can query at any time;

Step 12: When the main or backup ^interfaces: Interface standby failure or any Qos parameters of each index parameter generating parameters exceed the threshold, 'will carry out service switching or alarms.

Step 13: If the primary interface of the primary and secondary interfaces is faulty or the parameters of the QoS indicators exceed the threshold, if the backup interface is faulty and the QoS is met, the switchover is performed, and the service of the primary interface is switched to the standby interface. normal.

Step 14: The main interface of the other non-active or non-active and standby interfaces is faulty or the QoS parameters of the QoS exceed the threshold. The interface is alarmed and recorded.

From the above process, you can see the following two points:

(1) Using the Ping detection package to automatically detect the Qos between any two IP addresses periodically, statistics The Qos parameters such as packet loss rate, bit error rate, average delay, and average jitter are calculated, and the fault is further judged according to Qos.

(2) When Qos exceeds the threshold parameter or fault, the master is replaced by the protection service, and the alarm is notified to the user for maintenance.

In addition, if at least one routing node is included in the path between the two address endpoints, the present invention may also detect the Qos parameter of each segment path, and locate the fault point and the Qos falling point according to the detection result (ie, cause the Qos to decrease) Device or path).

FIG. 3 is a schematic diagram of positioning a routing node fault according to the present invention. Referring to FIG. 3, the present invention can automatically detect a faulty node on each routing node, that is, a source IP address to a QoS of each Layer 3 data communication device on the path, and use the detection result as a basis for user fault location, on the path. Each routing node performs automatic detection to locate the fault point and the device or path that causes the QoS parameters of the QoS to exceed the threshold. The steps specifically include:

Step 31: According to the implementation principle of the Tmcert command, the ping detection packet of the source address IP1, the destination address IP2, and the TTL number 1 is sent from the endpoint of the access device whose address is IP1. The router R1 receives the ping detection packet and puts the TTL. The value is decremented by 1. If the TTL value is equal to 0, the packet is discarded, and an Internet Control Management Protocol (ICMP) packet is sent to the source IP1 as the detection response packet. The reason is that the IP packet lifetime is invalid. If the TTL is greater than 0, Find the route, send the packet to the router R2; Send the Ping detection packet with the TTL number of 1, 2, in the IP header field in sequence, and the source destination addresses are the same, that is, the source address IP1 and the destination address IP2.

Step 32: As in the method shown in FIG. 2, for a ping detection packet with a TTL number of 1, the interval is sent a predetermined number of times, and the packet loss rate and error are counted according to the return time and return quality of the received detection response packet. Packet rate, delay, jitter and other parameters, and record each detection result; simultaneously record the IP address of each detection response packet, so that the Qos between the first end (ie, address IP1) and the router R1 can be determined. parameter.

Step 33: In turn, increase the TTL value by 1 to detect the Qos of the first end and the routers R2 and R3.

II The parameter, that is, the detection of the TTL number 2, 3... is sequentially performed according to the method of step 32, and the corresponding Qos parameter is obtained, and further compared with the preset Q _0S parameter threshold, if the Q _0S parameter reaches the preset If the Qos parameter threshold is used, it is determined that a fault or a Qos drop occurs between the first end and the router. For example, if the packet loss rate is 100%, the disconnection fault occurs, and the detection is stopped, indicating that the fault point has been found; When the TTL is greater than n+1, the packet received on the router Rn is forwarded to the access device where the destination IP2 is located. In this case, the correct ping detection response packet is sent to the first end. It is determined that the routing node on the entire path has been detected, and the detection of the segments is stopped.

In addition, the detection response packet may be an ICMP packet whose route hop count exceeds the maximum value, in addition to the ping detection response packet.

4 is a schematic diagram of a device for detecting Qos in a communication network according to the present invention. Referring to Figure 4, the device includes:

The detection packet transceiver unit 41 is disposed on the detected endpoint in the communication network, for example, the first end of the address IP1 and the second end of the IP2 address are used to send a predetermined number of detection packets to the peer interval. After receiving the detection packet from the peer end, returning the detection response packet to the opposite end, and receiving the detection response packet returned by the opposite end;

The Qos parameter calculation unit 42 is connected to the detection packet transceiver unit, and determines the Qos parameter between the two detected endpoints according to the return time and/or the return quality of the detection response packet of the opposite end received by the detection packet transceiver unit 41.

The detecting device further includes an alarm unit 43 for setting a threshold value of the Qos parameter, determining whether the Qos parameter determined by the Qos parameter calculating unit 42 reaches the threshold value, and triggering an alarm if it is reached.

If the device is to be used in a communication network having a primary and secondary interfaces, it is further necessary to further include a switching unit. FIG. 5 is a schematic diagram of a networking structure of the device according to the present invention applied to an Ethernet including an active and standby device. Referring to FIG. 5, the Ethernet access device has a primary and secondary interfaces and is configured with The third-party interface (that is, the interface of the L3) is connected; the detecting device is configured on the primary and secondary interfaces and the third-party interface, and is configured to detect the QoS parameter of the primary and secondary interfaces and the third-party interface; and further includes the switching unit 51, and the The alarm unit is connected, and is configured to trigger an alarm when the QoS parameter is triggered between the interface of the current service interface and the third-party interface, and the QoS parameter is not triggered when the QoS parameter between the other interface and the third-party interface is not triggered. .

It should be noted that the above embodiments are merely illustrative of the present invention and are not intended to limit the technical solutions described in the present invention; therefore, although the present specification has been described in detail with reference to the various embodiments described above, It is to be understood that the invention may be modified or equivalently substituted without departing from the spirit and scope of the invention.

Claims

Claim

A method for detecting a quality of service Qos in a communication network, comprising:

.

The method of claim 1, wherein the communication network has an active/standby interface and is connected to a third-party interface; and when one of the primary and secondary interfaces performs service processing, the primary and secondary interfaces and the first Steps A and B are performed between the three interfaces to detect the QoS parameters between the active and standby interfaces and the third-party interface. The QoS parameters between the interface that processes the service and the third-party interface on the active and standby interfaces reach the preset Qos parameter threshold. If the value of the QoS parameter between the other interface and the third interface does not reach the threshold, the active and standby interfaces are switched.

The method according to claim 1, wherein the method further comprises: setting a threshold value of the Qos parameter, and triggering an alarm when a Qos parameter between the two address endpoints reaches the threshold value.

The method according to claim 1, wherein the Qos parameter includes the number of lost packets, the number of error packets, the transmission delay, the jitter delay, the percentage of packet loss, the average transmission delay, and the average jitter. Any combination of delays or one of them.

The method of claim 1, wherein the Qos parameter includes the number of lost packets, and the method for determining the number of lost packets between the two address endpoints is: after the first end sends a detection packet If the detection response packet corresponding to the detection packet is not received within a predetermined time, the number of lost packets is accumulated.

The method according to claim 5, wherein the Qos parameter further includes a percentage of packet loss, and the specific method for determining the percentage of the packet loss is: adding the accumulated packet loss The number is divided by the total number of detected packets sent.

The method according to claim 1, wherein the Qos parameter includes an error packet number, and the method for determining the number of error packets between the two address endpoints is: After the packet, if the detection response packet corresponding to the detection packet is received within a predetermined time, it is further detected whether the content of the detection response packet is incorrect, and when an error is detected, the number of error packets is accumulated.

The method of claim 7, wherein the Qos parameter further includes a percentage error, and the specific method for determining the error percentage is: dividing the accumulated number of error packets by the sent The total number of packets detected.

The method according to claim 1, wherein the Qos parameter includes a transmission delay, and the method for determining a transmission delay between two address endpoints is: after each detection packet is sent by the first end, Receiving the detection response packet corresponding to the detection packet within a predetermined time, further detecting whether there is an error in the detection response packet, and determining whether to send the detection packet to the corresponding detection when the content of the detection response packet is correct The time difference of the response packet is used as the transmission delay of the detection packet.

The method according to claim 9, wherein the Qos parameter further includes an average transmission delay, and the specific method for determining the average transmission delay is: dividing the sum of the determined transmission delays by The total number of correct response packets is described.

The method according to claim 9, wherein the Qos parameter further comprises a jitter delay, and the method for determining a jitter delay between two address endpoints is: sending a detection packet twice, and determining After the transmission delays of the two detection packets, the time difference between the transmission delays of the two detection packets is taken as a jitter delay.

12. The method according to claim 11, wherein the Qos parameter between the two addresses further comprises an average jitter delay, and the method for determining the average jitter delay is: dividing the sum of the determined jitter delays The total number of response packets is correctly detected with the content.

The method according to any one of claims 7 to 12, wherein the detecting the content of the response packet is incorrect: detecting a media access control layer, an IP protocol layer, and an Internet control of the detection response packet The content of the message protocol layer is incorrect.

14. The method according to claim 1, wherein the detection period is preset, and the steps A and B are repeatedly performed in each detection period.

15. The method of claim 1, wherein the address is an interface address or an IP address.

The method according to claim 1, wherein the detection packet is a Ping detection packet, an ARP protocol packet, an echo packet, or a link control protocol negotiation packet.

The method according to any one of claims 1 to 16, wherein the path between the two address endpoints includes at least one routing node, and the method further includes a process of detecting a Qos parameter of each segment path, specifically For:

Cl, the first end sequentially sends the detection packets numbered 1, 2, and the source address of each detection 'packet is the first endpoint, and the destination address is the second endpoint;

C2, sending a detection packet interval of each number for a predetermined number of times; the routing node receiving the detection packet decrements the number by 1, and determines whether the number is 0, if yes, discarding the detection packet, and returning detection to the first endpoint a response packet, including an address of the routing node; if the number is not 0, forwarding the detection packet to the next routing node, repeating step C2 until the detection packet reaches the second endpoint, and the second endpoint is first The endpoint returns a detection response packet, including the address of the second endpoint;

The method according to claim 17, wherein the method further comprises: setting a threshold value of the Qos parameter, and sequentially determining between the first endpoint and each routing node Whether the Qos parameter reaches the threshold of the Qos parameter, and if so, stops the Qos parameter detection for each segment path, and locates the fault point and/or the Qos drop point between the corresponding routing nodes; otherwise, if received The detection response packet of the second endpoint determines that the path between the two address endpoints is detected, and stops the Qos parameter detection for each segment path.

A device for detecting a Qos in a communication network, the device comprising: a detection packet transceiver unit, disposed on a detected endpoint in the communication network, configured to send a predetermined number of detection packets to the opposite end interval, Receiving the detection packet sent by the peer end, returning the detection response packet to the opposite end, and receiving the detection response packet returned by the opposite end;

The Qos parameter calculation unit is connected to the detection packet transceiver unit, and determines a Qos parameter between the two detected endpoints according to a return time and/or a return quality of the detection response packet of the opposite end received by the detection packet transceiver unit.

The device for detecting Qos in the communication network according to claim 19, wherein the detecting device further comprises an alarm unit, configured to set a threshold value of the Qos parameter, and determine a Qos parameter determined by the Qos parameter calculation unit. Whether the threshold is reached and an alarm is triggered if it is reached.

The apparatus for detecting Qos in a communication network according to claim 20, wherein the communication network has a primary and backup interfaces and is connected to a third interface; the detecting device is disposed on the primary and secondary interfaces and The third interface is configured to detect the QoS parameter of the primary and secondary interfaces and the third interface, and further includes a switching unit, and is connected to the alarm unit, and is used to connect the QoS between the interface currently processing the service and the third interface in the primary and secondary interfaces. When the parameter triggers an alarm, and the QoS parameter between the other interface and the third interface does not trigger an alarm, the active and standby interfaces are switched.