Inference device

Abstract

This inference device comprises: an acquisition unit that acquires prescribed external information obtained in communication between a higher-order communication device and one or more lower-order communication devices connected to one or more user terminals; and an inference unit that infers the experiential quality of a network of the one or more user terminals on the basis of the acquired prescribed external information.　

Description

guessing device

[0001] The present invention relates to a prediction device.

[0002] As shown in Fig. 6, a conventional communication system includes an upper communication device and multiple lower communication devices. The upper communication device communicates with each of the lower communication devices. The upper communication device is a device installed in a communication building station, and the lower communication devices are terminals installed in users' homes. For example, if the communication system is a PON (Passive Optical Network) system, the upper communication device is an OLT (Optical Line Terminal), and the multiple lower communication devices are ONUs (Optical Network Units).

[0003] In recent years, the spread of social networking services (SNS) and the increase in remote work due to COVID-19 and other factors have placed a premium on the communication quality of communication services. Traditionally, communication carriers assessed network environments based on whether equipment was malfunctioning (whether communication was available or not). However, in recent years, understanding network conditions, such as network instability and slow communication speeds, and taking prompt measures to prevent users from abandoning their services has become important. Therefore, it is important for communication carriers to understand users' network quality of experience (QoE).

[0004] JP 2014-215988 A

[0005] However, conventional technologies cannot estimate QoE, and therefore must wait for a fault to be detected or for a user to report it. Furthermore, the need to maintain confidentiality of communications means that it is difficult to verify information inside frames, making it difficult to estimate QoE.

[0006] In view of the above circumstances, an object of the present invention is to provide a technique that can easily estimate a user's quality of experience.

[0007] One aspect of the present invention is an estimation device comprising one or more lower-level communication devices connected to one or more user terminals, an acquisition unit that acquires predetermined external information obtained in communication between the lower-level communication devices and a higher-level communication device, and an estimation unit that estimates the quality of experience of the network of the one or more user terminals based on the acquired predetermined external information.

[0008] The present invention makes it possible to easily estimate the user's quality of experience.

[0009] Fig. 1 is a diagram illustrating a configuration example of a communication system in an embodiment. Fig. 2 is a diagram illustrating an inference process (part 1) performed by an inference unit in an embodiment. Fig. 3 is a diagram illustrating an inference process (part 3) performed by an inference unit in an embodiment. Fig. 4 is a diagram illustrating an inference process (part 3) performed by an inference unit in an embodiment. Fig. 5 is a diagram illustrating a configuration example of a conventional communication system.

[0010] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0011] Fig. 1 is a diagram showing an example of the configuration of a communication system 100 according to an embodiment. The communication system 100 includes a lower communication device 10, an upper communication device 20, and a QoE estimation device 30. In Fig. 1, the communication system 100 includes m (m is an integer equal to or greater than 1) lower communication devices 10. One or more user terminals 40 are connected to each lower communication device 10. In the following description, as an example, it is assumed that a user terminal 40-1 is connected to the lower communication device 10-1, and a user terminal 40-n is connected to the lower communication device 10-m.

[0012] The communication system 100 may be a communication system that performs communication using, for example, TCP (Transmission Control Protocol), or may be a PON system. When the communication system 100 is a PON system, the lower-level communication device 10 is an ONU, and the upper-level communication device 20 is an OLT.

[0013] The lower-level communication device 10 is connected to a user terminal 40, and connects the user terminal 40 to the Internet.

[0014] The upper communication device 20 aggregates the lower communication devices 10 and notifies the QoE estimation device 30 of predetermined external information obtained in communication between each lower communication device 10 and the upper communication device 20. The predetermined external information is any one of the following: the number of retransmissions of an ACK (acknowledgement) packet, information for estimating the number of retransmissions of an ACK packet (for example, the number of transmission frames of the upper communication device 20, the number of reception frames of the lower communication device 10, the total number of bytes of a plurality of transmission frames, the total number of bytes of a plurality of reception frames, the transmission frame length, the reception frame length, etc., and time series information of frame arrival), information on the difference between the transmission request amount of upstream frames of the lower communication device 10 and the transmission permission amount of upstream frames of the lower communication device 10, or information on the packet discard rate (packet loss rate). The predetermined external information is information obtained without looking at the inside of the frame. Details of each piece of information included in the predetermined external information will be described later.

[0015] The user terminal 40 is a terminal device that a user uses at home, etc. The user terminal 40 connects to the Internet via the lower-level communication device 10.

[0016] The QoE estimation device 30 estimates the quality of experience of each user terminal 40 based on predetermined external information obtained from the higher-level communication device 20 .

[0017] The QoE estimation device 30 includes a CPU (Central Processing Unit), memory, auxiliary storage device, and the like, all connected via a bus, and executes a program. By executing the program, the QoE estimation device 30 functions as a device including an acquisition unit 31, an estimation unit 32, and an output unit 33. Note that all or part of the functions of the QoE estimation device 30 may be realized using hardware such as an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), or an FPGA (Field Programmable Gate Array). The program may also be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs (Read Only Memory), and CD-ROMs, as well as storage devices such as hard disks built into computer systems. The program may also be transmitted and received via a telecommunications line.

[0018] The acquisition unit 31 acquires predetermined external information from the higher-level communication device 20 .

[0019] The estimation unit 32 estimates the quality of experience of the network of each user terminal 40 based on the predetermined external information acquired by the acquisition unit 31. The estimation unit 32 may determine in advance which information of the predetermined external information acquired by the acquisition unit 31 to use to estimate the quality of experience of the network of the user terminal 40, or may estimate the quality of experience of the network of the user terminal 40 based on each of multiple pieces of information.

[0020] The estimation unit 32 estimates the number of retransmissions of the ACK packet from the upper communication device 20 using information about the packet, for example, information for estimating the number of retransmissions of the ACK packet obtained from the upper communication device 20, and estimates the quality of experience of the network of the user terminal 40 based on the estimated number of retransmissions of the ACK packet. Furthermore, the estimation unit 32 estimates the quality of experience of the network of the user terminal 40 based on the number of retransmissions of the ACK packet obtained from the upper communication device 20.

[0021] The estimation unit 32 calculates the sum of the differences, for example, using information regarding the difference between the transmission request amount of upstream frames of the lower communication device 10 and the transmission permission amount of upstream frames of the lower communication device 10 for a predetermined period, which information is acquired from the upper communication device 20. The estimation unit 32 estimates the quality of experience of the network of the user terminal 40 based on the calculated sum of the differences.

[0022] For example, the estimation unit 32 acquires information indicating the discard rate of red frames, information indicating the discard rate of yellow frames, and information indicating the discard rate of green frames for each user terminal 40 connected from the upper communication device 20 to the lower communication device 10 based on information regarding the discard rate of packets acquired from the upper communication device 20, and calculates the ratio of the number of discarded red packets, the number of discarded yellow packets, and the number of discarded green packets. Here, the red frames, yellow frames, and green frames are frames indicated by 2R (Rate) 3C (Color) in QoS (Quality of Service) technology.

[0023] It is not necessary to use 2R3C, but may use 1R2C. In the case of 1R2C, the estimation unit 32 acquires information indicating the yellow frame discard rate and information indicating the green frame discard rate of each user terminal 40 connected from the upper communication device 20 to the lower communication device 10 based on information regarding the packet discard rate acquired from the upper communication device 20, and calculates the ratio of the number of discarded yellow packets to the number of discarded green packets. Then, the estimation unit 32 estimates the quality of experience of the network of the user terminal 40 based on the calculated ratio of the number of discarded packets.

[0024] The output unit 33 outputs the results estimated by the estimation unit 32 to an external device. The output unit 33 may output the results estimated by the estimation unit 32 to a device operated by a network administrator, for example, or may cause a display device to display the results estimated by the estimation unit 32. The output unit 33 may output only results in which the quality of experience is estimated to be lower than a threshold, rather than outputting all of the estimation results.

[0025] Next, specific processing by the estimation unit 32 will be described with reference to FIGS.

[0026] (Inference process (part 1) performed by the estimation unit 32) Fig. 2 is a diagram for explaining the estimation process (part 1) performed by the estimation unit 32 in the embodiment. As the estimation process (part 1) performed by the estimation unit 32, a method of estimating QoE based on the number of retransmissions of ACK packets in TCP communication will be explained. As shown in Fig. 2, it is assumed that the lower communication device 10 and the upper communication device 20 are connected via network A. Note that, in Fig. 2, for the sake of simplicity, the explanation will be given taking as an example a case where there is one lower communication device 10 and one user terminal 40.

[0027] In normal communication, data transmitted from a user terminal 40 reaches the upper communication device 20 via the lower communication device 10. In response to receiving the received data, the upper communication device 20 transmits an ACK packet to the lower communication device 10 connected to the user terminal 40 that is the sender of the data. If no failure or the like occurs in network A, the ACK packet is received by the lower communication device 10 and forwarded to the user terminal 40. On the other hand, if a failure or the like occurs in network A, it is assumed that the ACK packet will not reach the lower communication device 10. If the lower communication device 10 is unable to receive the ACK packet within a predetermined time, it requests the upper communication device 20 to retransmit the ACK packet.

[0028] The more the number of ACK packet retransmissions, the worse the communication environment of network A is expected to be. In other words, the more the number of ACK packet retransmissions, the worse the quality of experience of the user operating the user terminal 40 is expected to be. Here, assume that the number of ACK packet retransmissions from the upper communication device 20 in network A is N. In this case, the estimation unit 32 acquires information indicating the "transmission frame length," information indicating the "reception frame length," and information indicating the "frame arrival time series" from the upper communication device 20, and estimates the number of ACK retransmissions based on each acquired piece of information. The information indicating the frame arrival time series is time series information of frames arriving at the lower communication device 10 and the upper communication device 20, such as SYN (Synchronize) packet → SYN / ACK packet → ACK packet...

[0029] The frame length of an ACK packet is defined by the standard (ACK: 64 bytes). Therefore, when the estimation unit 32 acquires the above-mentioned information from the higher-level communication device 20, it only needs to determine the ACK packet based on its frame length. Then, the estimation unit 32 estimates the number of times an ACK packet has been transmitted within a predetermined period of time based on information indicating the time sequence of frame arrival and information on the frame length. The estimation unit 32 compares the estimated number of retransmissions of the ACK packet with an evaluation index, and if the number of retransmissions exceeds the evaluation index, it estimates that the quality of experience of the user terminal 40 is low. If the number of retransmissions is equal to or less than the evaluation index, the estimation unit 32 estimates that the quality of experience of the user terminal 40 is high. The estimation unit 32 performs this process for each user terminal 40.

[0030] The evaluation index may be a value calculated based on the network conditions reported by the user in the past, indicating how many ACK retransmissions the user must have before the user begins to feel uncomfortable, i.e., makes a report. In the above example, the estimation unit 32 is configured to estimate the number of ACK packet retransmissions, but the number of ACK packet retransmissions (e.g., information that the ACK packet has been retransmitted N times) may be collected by the upper communication device 20 and notified to the QoE estimation device 30. In this case, the estimation unit 32 may compare the number of ACK packet retransmissions notified by the upper communication device 20 with the evaluation index.

[0031] (Inference Process (Part 2) Performed by the Inference Unit 32) Fig. 3 is a diagram for explaining the estimation process (part 2) performed by the estimation unit 32 in the embodiment. As the estimation process (part 2) performed by the estimation unit 32, a method of estimating QoE based on the sum of the differences between REPORT (request amount of upstream frames from the lower communication device 10) and GATE (permitted amount of transmission of upstream frames from the lower communication device 10) for each user terminal 40 in bandwidth sharing (DBA: Dynamic Bandwidth Allocation) will be described. Note that, for simplicity of explanation, Fig. 3 will be described taking as an example a case where there is one lower communication device 10 and one user terminal 40.

[0032] As shown in FIG. 3 , bandwidth allocation and data transmission in DBA are performed as follows: (1) The lower communication device 10 receives upstream frames transmitted from the user terminal 40. (2) The lower communication device 10 notifies the upper communication device 20 of an MPCP (Multi Point Control Protocol) REPORT. In response, the lower communication device 10 notifies the upper communication device 20 of the transmission request amount of upstream frames. (3) The upper communication device 20 calculates the transmission amount using the DBA algorithm. Here, for example, the transmission request amount included in the MPCP REPORT is A, and the transmission permission amount included in the MPCP GATE is B. (4) The upper communication device 20 transmits an MPCP GATE including the calculated transmission permission amount B to the lower communication device 10. (5) The lower communication device 10 transmits upstream frames according to the transmission permission amount B included in the received MPCP GATE (upstream frames according to the allocated allocation amount) to the upper communication device 20.

[0033] The QoE estimation device 30 acquires information indicating a transmission request amount A of upstream frames of the lower communication device 10, which is included in the MPCP REPORT, and information indicating an allowed transmission amount B of upstream frames of the lower communication device 10, which is included in the MPCP GATE, from the upper communication device 20. The estimation unit 32 calculates the difference between the transmission request amount A of upstream frames and the allowed transmission amount B of upstream frames (allowed transmission amount B - requested transmission amount A) based on the acquired information.

[0034] The estimation unit 32 calculates multiple difference values ​​by repeating the above process at regular intervals. The estimation unit 32 then calculates the sum of the calculated multiple difference values. Note that an MPCP GATE greater than the transmission request amount A included in the MPCP REPORT is not permitted, and the sum is 0 if all of the user terminal 40's requests are met. The estimation unit 32 estimates the quality of experience of the user terminal 40 based on the calculated sum of difference values. The sum of difference values ​​has a maximum value of 0 when all of the user terminal 40's requests are met, and becomes more negative as the user terminal 40's request amount exceeds the permitted transmission amount. Therefore, the estimation unit 32 estimates that the quality of experience of the user terminal 40 is lower as the absolute value increases. In other words, the estimation unit 32 estimates that the quality of experience of the user terminal 40 is higher as the absolute value decreases (closer to 0). The estimation unit 32 performs this process for each user terminal 40.

[0035] (Inference Process (Part 3) Performed by the Estimation Unit 32) FIGS. 4 and 5 are diagrams for explaining the estimation process (Part 3) performed by the estimation unit 32 in the embodiment. As the estimation process (Part 3) performed by the estimation unit 32, a method of estimating QoE based on queue processing will be described. In explaining this content, the lower-level communication device 10 applies 2R3C and WRED (Weighted Random Early Detection) in QoS technologies. The lower-level communication device 10 performs 2R3C marking (1R2C may also be used) for each connected user terminal 40. The lower-level communication device 10 performs 3C marking with red (none in the case of 1R2C), yellow, and green. As shown in FIG. 4, the lower-level communication device 10 classifies frames equal to or less than the guaranteed bandwidth (CIR: Committed Information Rate) as green (green traffic in FIG. 4), frames exceeding the guaranteed bandwidth as yellow (yellow traffic in FIG. 4), and frames exceeding the maximum bandwidth (PIR: Peak Information Rate) as red (red traffic in FIG. 4). The graph shown in FIG. 4 shows which color a particular user's traffic is classified into when the traffic rate of that user increases over time.

[0036] The lower communication device 10 notifies the upper communication device 20 of information on the number of discarded packets of each color for each user terminal 40. The QoE estimation device 30 acquires information on the number of discarded packets of each color for each user terminal 40 from the upper communication device 20. The estimation unit 32 estimates the quality of experience of the user terminal 40 based on the acquired ratio of the number of discarded packets of each color.

[0037] An example of WRED settings is shown in Figure 5. In Figure 5, the vertical axis represents the packet drop probability, and the horizontal axis represents the queue length. A 0% drop threshold and a 100% drop threshold are set. As an example, let's assume that the following values ​​are set: Red (graph of line segment L1): 0% drop threshold is 10%, 100% drop threshold is 25% Yellow (graph of line segment L2): 0% drop threshold is 25%, 100% drop threshold is 50% Green (graph of line segment L3): 0% drop threshold is 50%, 100% drop threshold is 100%

[0038] The lower-level communication device 10 calculates the average queue length at regular intervals. If the average queue length is equal to or less than the 0% drop threshold, packets are guaranteed and not discarded. If the average queue length is between the 0% drop threshold and the 100% drop threshold, packets are randomly discarded based on the drop probability. If the average queue length is equal to or greater than the 100% drop threshold, 100% of packets are discarded. For example, if the average queue length is the length of the symbol ▲ shown in Figure 5, packets classified as red will be discarded 100%, packets classified as yellow will be discarded at a random rate, and packets classified as green will not be discarded.

[0039] The QoE estimation device 30 acquires information indicating the discard rate of red frames, information indicating the discard rate of yellow frames, and information indicating the discard rate of green frames for each user terminal 40 connected from the upper communication device 20 to the lower communication device 10. The estimation unit 32 uses the acquired information to calculate the ratio of the number of discarded red packets, the number of discarded yellow packets, and the number of discarded green packets. The estimation unit 32 estimates the quality of experience of the user terminal 40 based on the ratio of green packets to all discarded packets. For example, if the ratio of green packets to all discarded packets increases, the estimation unit 32 assumes that the average queue length is increasing and estimates that the quality of experience of the user terminal 40 is decreasing. The estimation unit 32 performs this process for each user terminal 40.

[0040] According to the communication system 100 configured as described above, the QoE estimation device 30 includes an acquisition unit 31 that acquires predetermined external information obtained in communication between one or more lower-level communication devices 10 connected to one or more user terminals 40 and the upper-level communication device 20, and an estimation unit 32 that estimates the quality of experience of the network for each of the one or more user terminals 40 based on the predetermined external information acquired by the acquisition unit 31.

[0041] The predetermined external information can be obtained without looking inside the frame. This makes it possible to easily estimate the user's quality of experience. Such real-time QoE estimation enables proactive measures for the user, improving the user's QoE. As a result, it becomes possible to continue using communication services.

[0042] (Modification) The QoE estimation device 30 may be provided in the higher-level communication device 20 or in a device higher than the higher-level communication device 20 .

[0043] Although an embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to this embodiment, and includes designs within the scope of the gist of the present invention.

[0044] The present invention can be applied to a communication system via a network.

[0045] 10, 10-1 to 10-m... lower-level communication device, 20... upper-level communication device, 30... QoE estimation device, 31... acquisition unit, 32... estimation unit, 33... output unit, 40, 40-1 to 40-n... user terminal

Claims

1. An estimation device comprising: an acquisition unit that acquires predetermined external information obtained in communication between one or more lower-level communication devices connected to one or more user terminals and a higher-level communication device; and an estimation unit that estimates the quality of experience of the network of the one or more user terminals based on the acquired predetermined external information.

2. The estimation device described in claim 1, wherein the specified external information is information regarding the number of retransmissions of a response signal from the upper communication device, or information regarding packets transmitted and received in communication between the one or more lower communication devices and the upper communication device for estimating the number of retransmissions of the response signal, and the estimation unit estimates the number of retransmissions of the response signal from the upper communication device using the information regarding the packets, and estimates the quality of experience of the network for the one or more user terminals based on the estimated number of retransmissions of the response signal or the acquired number of retransmissions of the response signal.

3. The estimation device described in claim 1, wherein the specified external information is information regarding the difference between the requested transmission volume of upstream frames of the one or more lower-level communication devices and the permitted transmission volume of upstream frames of the one or more lower-level communication devices, and the estimation unit calculates a sum using information regarding the difference for a specified period of time, and estimates the network quality of experience of the one or more user terminals based on the calculated sum.

4. The estimation device according to claim 1, wherein the predetermined external information is information relating to a packet discard rate, and the estimation unit estimates the network quality of experience of the one or more user terminals based on the information relating to the packet discard rate.

Images