WO2016125415A1 - 体感品質最適化システム、体感品質最適化装置、レコメンド要求装置、体感品質最適化方法、レコメンド要求方法及びプログラム - Google Patents
体感品質最適化システム、体感品質最適化装置、レコメンド要求装置、体感品質最適化方法、レコメンド要求方法及びプログラム Download PDFInfo
- Publication number
- WO2016125415A1 WO2016125415A1 PCT/JP2015/086087 JP2015086087W WO2016125415A1 WO 2016125415 A1 WO2016125415 A1 WO 2016125415A1 JP 2015086087 W JP2015086087 W JP 2015086087W WO 2016125415 A1 WO2016125415 A1 WO 2016125415A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- qoe
- average
- function
- recommendation
- parameter
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/80—Responding to QoS
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
- H04N21/462—Content or additional data management, e.g. creating a master electronic program guide from data received from the Internet and a Head-end, controlling the complexity of a video stream by scaling the resolution or bit-rate based on the client capabilities
- H04N21/4621—Controlling the complexity of the content stream or additional data, e.g. lowering the resolution or bit-rate of the video stream for a mobile client with a small screen
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
- H04L65/60—Network streaming of media packets
- H04L65/61—Network streaming of media packets for supporting one-way streaming services, e.g. Internet radio
- H04L65/612—Network streaming of media packets for supporting one-way streaming services, e.g. Internet radio for unicast
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/266—Channel or content management, e.g. generation and management of keys and entitlement messages in a conditional access system, merging a VOD unicast channel into a multicast channel
- H04N21/2662—Controlling the complexity of the video stream, e.g. by scaling the resolution or bitrate of the video stream based on the client capabilities
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
- H04N21/44209—Monitoring of downstream path of the transmission network originating from a server, e.g. bandwidth variations of a wireless network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/472—End-user interface for requesting content, additional data or services; End-user interface for interacting with content, e.g. for content reservation or setting reminders, for requesting event notification, for manipulating displayed content
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/60—Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client
- H04N21/63—Control signaling related to video distribution between client, server and network components; Network processes for video distribution between server and clients or between remote clients, e.g. transmitting basic layer and enhancement layers over different transmission paths, setting up a peer-to-peer communication via Internet between remote STB's; Communication protocols; Addressing
- H04N21/637—Control signals issued by the client directed to the server or network components
- H04N21/6377—Control signals issued by the client directed to the server or network components directed to server
- H04N21/6379—Control signals issued by the client directed to the server or network components directed to server directed to encoder, e.g. for requesting a lower encoding rate
Definitions
- the present invention relates to an experience quality optimization technique for optimizing the quality (QoE: Quality of Experience) experienced by a user in a content distribution service.
- QoE Quality of Experience
- Nico Nico Douga registered trademark
- This method is a method of estimating the user's available bandwidth in some form and distributing the content at a bit rate corresponding to the available bandwidth.
- available bandwidth estimation methods There are several types of available bandwidth estimation methods. a. Passive type estimation method In the passive type estimation method, the available bandwidth is estimated from past communication history.
- estimation is performed using some tendency regarding the available bandwidth.
- the available bandwidth is estimated using time-series stationarity (see Non-Patent Document 3).
- Active type estimation method In the active type estimation method, short communication is additionally performed before distribution, and the available bandwidth is estimated with reference to the communication performance at that time. For example, a packet group that increases sequentially is sent before communication, and the available bandwidth is estimated from the amount of change in delay until arrival (see Non-Patent Document 4).
- c. Feedback system In the feedback system, the speed of the communication itself is measured during delivery, and the delivery rate is changed in real time according to the communication speed at that time. This method is mainly applied to video distribution in which the distribution rate can be changed on the way.
- content is divided into short time unit files called chunk type, network quality (NW quality) etc. is measured between server / client for each chunk, based on the measurement result, Next, the distribution rate of the chunk received by the terminal is determined, and the content is distributed.
- NW quality network quality
- Ohshiba et al. “Short-term Available Bandwidth Estimation Method to Ensure Real-Time Communication Quality”, Transactions of Information Processing Society of Japan, 53 (2), 698-711, 2012-2.
- Enomoto et al. "Dynamic coding rate control method suitable for high-speed mobile communication networks in media streaming", Transactions of Information Processing Society of Japan, 50 (10), p.2532-2542, 2009-10-1.
- Sazawa et al. "Examination of Dynamic Rate Control Method for TCP Video Streaming", IEICE Technical Report, IE, Image Engineering 102 (469), p.19-24, 2002-11-15.
- the conventional existing methods have the following problems.
- the method in which the user selects the distribution rate has the advantage that the video distribution rate can be set according to the user's preference, but it is higher than the actual communication environment.
- the method of selecting the distribution rate on the content distribution provider side is that the distribution rate can be set according to the amount of equipment on the distribution side, but the communication environment of each user Since the rate control is not performed according to the situation, content video distribution (rebuffering) at an excessive rate with respect to the communication environment, or content distribution at an excessive rate occurs. There is. (3) Method for selecting content distribution rate according to communication status a. Passive type estimation method Passive type estimation method requires a sufficient past end-to-end communication history, so far it has been difficult to prepare sufficient data. b. Active type estimation method The active type estimation method has a problem in that it takes time to measure the available bandwidth before communication, and thus it takes time to distribute the content.
- the waiting time until the start of communication greatly affects the quality of experience in web applications.
- c. Feedback method In this distribution method, the frequency of occurrence of rebuffering with playback stoppage is low, but the distribution rate (resolution) fluctuates during video viewing, so it is confirmed by subject evaluation that the viewer's QoE is not high. ing. In addition, since the distribution rate is set for the purpose of reducing rebuffering, the distribution rate may be lowered too much, causing an event that QoE is lower than when rebuffering occurs.
- the present invention has been made in view of such problems, accumulates information related to NW quality / QoE for each content distribution, and estimates a distribution method for optimizing QoE based on the information,
- the purpose is to provide technology that makes it possible to make recommendations.
- the experience quality optimization system is a experience quality optimization system for optimizing the quality (QoE: Quality of Experience) experienced by a user in a content distribution service.
- QoE Quality of Experience
- a recommendation requesting device that transmits a parameter candidate and receives a recommendation value that optimizes QoE, and a delivery parameter that estimates QoE based on the delivery parameter candidate input by the recommendation requesting device and optimizes the QoE Is calculated as a recommendation value, and a sensation quality optimizing device that outputs the recommendation value is provided.
- the experience quality optimization apparatus is an experience quality optimization apparatus for optimizing QoE in a content distribution service, and inputs distribution parameter candidates for content distribution and optimizes QoE. Based on a parameter input / output unit that outputs a recommendation value, and a distribution parameter candidate input by the parameter input / output unit, an estimation value that estimates a QoE and calculates a distribution parameter that optimizes the QoE as the recommendation value A recommendation value calculation unit.
- a recommendation request device is a recommendation request device for optimizing QoE in a content distribution service, and for requesting a distribution parameter for optimizing QoE as a recommendation value.
- a distribution parameter output unit that outputs a distribution parameter candidate to the sensation quality optimization device; and a recommendation value reception unit that receives the recommendation value from the sensation quality optimization device.
- FIG. 1 The figure which shows the system outline
- FIG. The figure explaining an example of the definition of the input-output data in the OTT interface of the said experience quality optimization apparatus 1.
- FIG. 1 The figure explaining an example of the definition of the input-output data in the OTT interface of the said experience quality optimization apparatus 1.
- FIG. The figure which shows an example of the table and parameter list which were memorize
- FIG. The figure which shows an example of the table and parameter list which were memorize
- FIG. The figure which shows an example of the table and parameter list which were memorize
- FIG. The figure which shows an example of the table and parameter list which were memorize
- FIG. The figure explaining the process performed by the throughput estimation function 11 of the said experience quality optimization apparatus 1.
- FIG. 12 The figure explaining the process performed by the throughput estimation function 11 of the said experience quality optimization apparatus 1.
- FIG. 12 The figure explaining the process performed by the reproduction
- FIG. 13 The figure explaining the process performed by the QoE estimation function 13 of the said experience quality optimization apparatus 1.
- FIG. 13 The figure explaining the process performed by the resolution and frame rate recommendation function 14 of the said experience quality optimization apparatus 1.
- FIG. The figure explaining the process performed by the initial coding parameter recommendation function 15 of the said experience quality optimization apparatus 1.
- FIG. 12 Processing included in the initial coding parameter recommendation function 15 (FIG. 12), which is executed by the function 15-1 that outputs the overall QoE and the coding parameter table T12 from the coding bit rate array and the average throughput distribution.
- FIG. 12 Processing included in the initial coding parameter recommendation function 15 (FIG. 12), which is executed by the function 15-1 that outputs the overall QoE and the coding parameter table T12 from the coding bit rate array and the average throughput distribution
- the figure explaining the process performed by the function 15-2 which is a function contained in the said initial coding parameter recommendation function 15 (FIG. 12), and outputs average throughput distribution table T11.
- the figure explaining the process performed by the function 15-3 which converts from table T12 to average throughput-QoE table T13.
- the function included in the function 16-2 (FIG. 21) for calculating the local average QoE and QoE of the real-time encoding parameter recommendation function 16 and executed by the function 16-4 for creating the average throughput-QoE table FIG.
- FIG. 1 The figure explaining the process performed by the throughput table information update function 21 of the said experience quality optimization apparatus 1.
- FIG. The figure explaining the process performed by the viewing log update function 22 of the said experience quality optimization apparatus 1.
- FIG. The figure explaining the process performed by the viewing log update function 22 of the said experience quality optimization apparatus 1.
- FIG. 1 is a diagram showing a system overview of a sensation quality optimization system according to an embodiment of the present invention.
- the experience quality optimization system includes a “recommendation requesting device 40” that requests parameters for optimal distribution in a video distribution system (hereinafter referred to as OTT (Over-The-Top)) of a content distribution company, It is composed of two devices, “Experience Quality Optimization Device 1” that estimates QoE from various delivery parameters and recommends the optimum parameters.
- OTT Over-The-Top
- the recommendation requesting device 40 outputs a delivery parameter candidate for content delivery to the experience quality optimizing device 1 in order to request a delivery parameter for optimizing QoE as a recommendation value, and the experience quality.
- a recommendation value receiving unit 42 for receiving the recommendation value from the optimization device 1.
- the sensation quality optimizing device 1 has two functions of “estimation / recommendation function (estimated value / recommendation value calculation unit 10)” and “data storage function (setting value update unit 20)”. The following services are provided through the IP network.
- Each of the recommendation requesting device 40 and the sensation quality optimizing device 1 includes a CPU, which is a computer, a storage device, an input / output device, and a communication device, and the CPU is configured according to a program stored in the storage device. Various functions are executed by controlling the operation. Further, for each of the recommendation requesting device 40 and the sensation quality optimizing device 1, each unit (each function) may be a hardware circuit (such as an integrated circuit).
- the following operation is a specific example of the operation of the experience quality optimization system.
- a user When a user views a video, it sends a video request to the video distribution system. At that time, the user transmits user attribute information (e.g., usage location, network usage mode, terminal information) via an application or a browser.
- user attribute information e.g., usage location, network usage mode, terminal information
- the video distribution system that has received the video viewing request includes the user attribute information and the encoding conditions (distribution rate, resolution, frame rate, etc.) of each of the plurality of image quality prepared in the video distribution system for the corresponding video. These may be referred to as delivery parameter candidates) and sent to the sensation quality optimizing apparatus 1 to inquire about the optimal encoding conditions.
- the sensation quality optimization device 1 uses the technology described in detail below for the received request to calculate an encoding condition that maximizes QoE and returns it as a response to the video distribution system.
- the video distribution system distributes video to users under the received conditions. After viewing the moving image, the user inputs viewing log information (throughput information, network information, etc.) as a result to the experience quality optimization system.
- FIG. 2 is a diagram illustrating a functional configuration of the sensation quality optimization apparatus 1 in the sensation quality optimization system.
- the sensation quality optimizing device 1 includes an estimated value / recommendation value calculation unit 10, a set value update unit 20, and a parameter input / output unit 30.
- the estimated value / recommendation value calculation unit 10 includes data received from the recommendation requesting device 40 and the IPT via the IP network, a viewing log and a throughput table stored in the storage unit 24 of the setting value update unit 20, Based on set values such as constants, various estimated values and recommended values for optimizing QoE are calculated and output.
- the set value update unit 20 updates information of the viewing log and the throughput table based on data from the OTT. Further, based on the input data from the OTT and the system administrator of the sensation quality optimizing device 1, the setting value of the sensation quality optimizing device 1 is updated.
- the parameter input / output unit 30 receives a recommendation request from the recommendation requesting device 40 via an IP network or the like, receives a distribution parameter, and transmits a recommendation value calculated by the estimated value / recommendation value calculating unit 10.
- the estimated value / recommendation value calculation unit 10 includes a throughput estimation function 11, a reproduction stop state estimation function 12, a QoE estimation function 13, a resolution / frame rate recommendation function 14, an initial encoding parameter recommendation function 15, and a real-time encoding parameter recommendation function. It consists of sixteen functional blocks.
- the set value update unit 20 includes three functional blocks, a throughput table information update function 21, a viewing log update function 22, and a set value reference function 23, and a storage unit 24.
- the parameter input / output unit 30 receives and transmits IP packet data via the OTT interface with the OTT via the IP network.
- the set value reference function 23 of the set value update unit 20 communicates commands and data with the system administrator input / output device 25 of the sensation quality optimizing device 1 via an administrator interface (CUI). To output data.
- CUI administrator interface
- FIG. 3 is a diagram for explaining an outline of main functions in the sensation quality optimizing apparatus 1.
- FIG. 4 is a diagram showing a processing flow of the sensation quality optimizing apparatus 1.
- FIG. 5A is a diagram for explaining an example of the definition of input / output data in the OTT interface of the sensation quality optimization device 1.
- FIG. 5B is a diagram illustrating the definition of input / output data in the OTT interface of the sensation quality optimization system 1.
- FIG. 5C is a diagram illustrating the definition of input / output data in the OTT interface of the sensation quality optimization system 1.
- FIG. 5D is a diagram illustrating the definition of input / output data in the OTT interface of the sensation quality optimization system 1.
- FIG. 5E is a diagram for explaining the definition of input / output data in the OTT interface of the sensation quality optimization system 1.
- each input / output data “service ID”... “Average playback stop time” “reproduction stop count”. Define Format, Sample value, Lower limit, Upper limit, Data overview, and Related functions.
- the corresponding input / output data is input to and from which function between the OTT and the sensation quality optimizing device 1.
- Imn indicating the function mn of the input destination and "Imn” indicating the function of the input destination, and "Omn” indicating the combination of the function "O” indicating the output and the function mn of the output source for the output data.
- the input data “service ID” from the OTT is an ID assigned to each delivery service by the OTT, and the functions 11 to 16 of the estimated value / recommendation value calculation unit 10 and the viewing log update of the setting value update unit 20 are updated.
- the data is input to the function 22 (I11 to I16, I22).
- “average throughput average” indicates an average value [kbps] of average measured throughput calculated by the throughput estimation function 11 of the estimated value / recommendation value calculation unit 10 and output to the parameter input / output unit 30 ( O11).
- 6A and 6B are diagrams illustrating definitions of other data used in each function of the sensation quality optimization device 1.
- FIG. 7A is a diagram showing an example of a table and a list of parameters stored in the storage unit 24 that is referred to and updated in each function of the sensation quality optimization device 1.
- FIG. 7B is a diagram illustrating an example of a table and a list of parameters stored in the storage unit 24 that is referred to and updated in each function of the sensation quality optimization device 1.
- FIG. 7C is a diagram illustrating an example of a list of tables and parameters stored in the storage unit 24 that is referred to and updated in each function of the sensation quality optimizing device 1.
- FIG. 7D is a diagram illustrating an example of a table and a list of parameters stored in the storage unit 24 that is referred to and updated in each function of the sensation quality optimization apparatus 1.
- FIG. 7E is a diagram showing an example of a table and a list of parameters stored in the storage unit 24 that is referred to and updated in each function of the sensation quality optimizing device 1.
- “(day of the week, time) -calendar information conversion table T1” in FIG. 7A is referred to by the throughput estimation function 11 of the estimated value / recommendation value calculation unit 10, and “resolution table T8” in FIG.
- the “viewing log R” in FIG. 7E is updated by the viewing log update function 22 of the set value update unit 20, and is referred to by the resolution / frame rate recommendation function 14 of the estimated value / recommendation value calculation unit 10.
- FIG. 1 An overview of processing executed based on the various data, tables, and parameters shown in Fig. 1 will be described.
- the throughput estimation function 11 is a function for calculating information related to throughput from the input information of the OTT.
- the throughput estimation function 11 receives various input data (service ID) from the OTT via the IP network via the OTT interface (FIG. 2) for each content distribution request of the user of the content distribution service.
- service ID input data
- Design ID date and time
- the following are optional parameters use server, distribution server group, user ID, fixed / mobile identifier, server speed control, [IP address or carrier name], [cell ID or coordinate], wi-fi usage determination )
- Is input Is input, and information related to throughput [throughput average average, throughput average deviation, throughput deviation average, user actual average average, user actual average deviation, user actual deviation average] is calculated and output.
- the playback stop state estimation function 12 is a function of calculating information related to playback stop from the input information of the OTT and information related to the throughput.
- the reproduction stop state estimation function 12 receives various input data (from the OTT via the IP network via the OTT interface (FIG. 2) via the IP network for each content distribution request of the user of the content distribution service.
- Service ID, design ID, encoding bit rate, average throughput, throughput deviation, video duration, the following are input optional parameters (playback start threshold, playback restart threshold, playback stop threshold).
- coefficients coefficient tables 4 and 5 stored in the storage unit 24 for the internal system, the average playback stop count and the average playback stop time are calculated and output.
- the QoE estimation function 13 is a function for calculating QoE from information related to encoding.
- the QoE estimation function 13 performs various types of input data (service ID) via the OTT interface (FIG. 2) of the parameter input / output unit 30 via the IP network from the OTT for each content distribution request of the user of the content distribution service. , Design ID, encoding bit rate, resolution, frame rate, and the following optional parameters (reproduction stop count, reproduction stop time length array) are input, and QoE is calculated and output.
- the resolution / frame rate recommendation function 14 is a function for calculating the resolution and frame rate at which QoE is maximized from the encoding bit rate.
- the resolution / frame rate recommendation function 14 is executed every time a content distribution service is designed and redesigned (encoding, etc.), and is connected to the OTT interface of the parameter input / output unit 30 from the OTT via an IP network.
- Various input data service ID, design ID, encoding bit rate, and the following optional parameters (number of playback stop times, playback stop time length, resolution array, frame rate array)) are input via (FIG. 2), Calculate and output resolution, frame rate, and QoE.
- the initial encoding parameter recommendation function 15 is a function for calculating an encoding parameter that maximizes the average QoE (overall average QoE) of the entire user.
- the initial encoding parameter recommendation function 15 is executed every time the content distribution service is designed and redesigned (encoding, etc.), and is performed from the OTT via the IP network via the OTT interface ( Various input data (service ID, design ID, number of encodings, the following is an optional parameter (encoding bit rate array)) is input via FIG. 2), and the overall average QoE and encoding parameter table T12 are calculated. Output.
- the real-time encoding parameter recommendation function 16 is a function for calculating information related to encoding parameters from the input information of the OTT.
- the real-time encoding parameter recommendation function 16 performs various types of input data via the OTT interface (FIG. 2) of the parameter input / output unit 30 from the OTT via the IP network for each content distribution request of the user of the content distribution service.
- Service ID, design ID, date and time the following are optional parameters (use server, distribution server group, user ID, fixed / mobile identifier, server speed control, [IP address or carrier name], [cell ID or coordinate], wi- fi usage decision, OTT encoding parameter table, video duration, playback start threshold, playback restart threshold, playback stop threshold)), resolution, frame rate, encoding bit rate, QoE, local average Calculate QoE and output input / output as a log.
- the local average QoE is equivalent to the average QoE of all users in the same network environment as the user who is the recommendation target.
- the throughput table information update function 21 uses a throughput table ("(calendar information, base station ID) -speed conversion table T4", "(calendar information, NW name) -speed conversion table T5" or "(calendar information, use server). ) -Speed conversion table T6 ").
- the update cycle is based on the time of system operation, and is automatically started once every 1 to 5 minutes, for example, update target designation information (calendar) in the input information from the OTT.
- the update target table is updated based on the information, the use server, and the NW name.
- the viewing log update function 22 is a function for newly creating and updating the viewing log R.
- the viewing log update function 22 uses input data (user ID, fixed / mobile identifier, date / time, usage server, distribution server group, IP address, carrier name, cell ID, coordinates, average measured throughput) as input information of the OTT. Measured throughput deviation, service ID, design ID, video stop count, video file ID, browser used, OS used, wi-fi usage determination, file size, video bit rate, carrier identifier, NW name, option tags 1-10) If it exists, new addition or update to the viewing log R stored in the storage unit 24 is performed.
- the set value reference function 23 is a function for referring to and updating set values (FIGS. 7A to 7E) such as tables, parameters, and coefficients (constants) stored in the storage unit 24. Based on the input data from the OTT and the command data from the system manager of the sensation quality optimizing device 1, setting values used in the respective functions of the sensation quality optimizing device 1 are updated.
- the CPU controls the operation of each unit in accordance with instructions described in the program that controls each function stored in the storage device, and the software and hardware cooperate with each other.
- the throughput estimation function 11, the reproduction stop state estimation function 12, the QoE estimation function 13, the resolution / frame rate recommendation function 14, the initial encoding parameter recommendation function 15, A real-time encoding parameter recommendation function 16, a throughput table information update function 21, a viewing log update function 22, and a set value reference function 23 are realized.
- Each function of the information update function 21, the viewing log update function 22, and the set value reference function 23 can also be realized as a hardware circuit incorporating the processing.
- the technique used by the throughput estimation function 11 in the present embodiment is a technique for estimating the current throughput from a past viewing log (actual throughput value) of the user's network environment.
- the average / deviation of the throughput is calculated by using the usage network information and the time zone information as input and referring to the viewing log database of the usage base station / ISP. Detailed processing will be described below.
- FIG. 8A and 8B are diagrams for explaining processing executed by the throughput estimation function 11 of the sensation quality optimizing apparatus 1.
- FIG. 8A mainly shows a processing procedure
- FIG. 8B shows an example of a table used in the procedure.
- the day of the week label that matches the date may be selected by collating with calendar information or the like.
- a label indicating a holiday is selected, and in other cases, a label indicating a day of the week (Sunday to Saturday) is selected.
- Input data includes the following common data for mobile and fixed: service ID, design ID, date / time (year / month / day / time), optional parameters (use server, distribution server group, user ID, fixed / Mobile identifier, server speed control).
- Wi-fi usage determination data and data when wi-fi usage determination is false include carrier name, cell ID (or coordinates, or both), and wi-fi There is an IP address as data when the usage judgment is true. It should be noted that unnecessary information may be included in the input data. This is because it is input to the viewing log update function 22.
- step S112 based on the input data, it is determined whether the user terminal is a fixed terminal, a wi-fi using terminal, or a mobile terminal (mobile) not using wi-fi (step S112).
- the IP address of the user terminal is converted into an NW name according to the IP address-ISP conversion table T2 (step S113a).
- the carrier name corresponding to the user terminal is converted into a carrier identifier according to the carrier name-carrier identifier conversion table T3, and a cell ID is assigned (step S113b).
- the content of the viewing log R corresponding to the input data stored in the storage unit 24 is updated by the viewing log update function 22 of the set value update unit 20 (step S114).
- step S116 it is determined whether or not a utilization server is input as the input data.
- server speed control information When the usage server is input, it is determined whether or not server speed control information is input as the input data. When the server speed control information is input, the value of the server speed control information is determined. Is the average throughput of the server, and the average throughput deviation and the average deviation are set to “0”. If the server speed control information is not input, the average throughput, average throughput deviation, average throughput deviation of the usage server is determined according to (calendar information, usage server) -speed conversion table T6 (step S117). .
- the average throughput average of the NW acquired in step S115 is compared with the average throughput average of the utilization server acquired in step S117, and corresponds to the smaller average throughput average and the average throughput average.
- the throughput average deviation and the throughput deviation average are output (step S118).
- step S119 If the use server is not input, the throughput average average, throughput average deviation, and throughput deviation average of the NW acquired in step S115 are output (step S119).
- the elapsed time is within the past log search period (11_time) and the measured throughput average and measured If there is more than the number of past log calculations (11_limit) that have a throughput deviation and the same (user ID, distribution server group, NW name, calendar information), the latest number of viewing logs for that user (11_limit ) To calculate the throughput.
- the average value of the measured actual throughput average is calculated as the user actual average average
- the deviation of the actual measured throughput average is calculated as the user actual average deviation
- the average value of the actual measured throughput deviation is calculated as the user actual average deviation.
- the user terminal is a mobile terminal (mobile) that does not use wi-fi and there are coordinates as the input data
- the elapsed time is searched in the viewing log R stored in the storage unit 24.
- Logs that have a measured throughput average and measured throughput deviation within the period (11_time), and that match are the number of past logs calculated ( 11_limit) or more, the same throughput calculation as described above is performed for the latest calculation count (11_limit) of the viewing log of the user.
- the user terminal is a mobile terminal (mobile) that does not use wi-fi and there is no coordinates as the input data
- the elapsed time is searched in the viewing log R stored in the storage unit 24.
- Logs that have an average measured throughput and measured throughput deviation within the period (11_time) and have the same (user ID, distribution server group, cell ID information (base station ID), calendar information) are the number of past log calculations (11_limit) If there is the above, the same throughput calculation as described above is performed for the latest calculation count (11_limit) of the viewing log of the user.
- the value obtained from the above calculation is the user record. However, according to the calculation result flag (11_use_user_result) of the user result, depending on the calculation target user, this process (S120) is skipped, and the throughput average / average deviation / deviation average of the user result is not output.
- the throughput average average, throughput average deviation, and throughput deviation average estimated in step S118 or S119 and the user actual average average, user actual average deviation, and user actual deviation average calculated in S120 are OTT. Is output.
- the reproduction stop state estimation technique in the reproduction stop state estimation function 12 is a throughput average / deviation that is an output value of the throughput estimation technique, an encoding bit rate, a moving image length (duration), a reproduction start threshold value / reproduction. This is a technique for estimating the number of playback stops / playback stop time from the stop threshold / playback restart threshold.
- the terminal has a buffer for storing received data, and playback start, stop, and restart operations are determined according to the remaining capacity of the buffer.
- a model formula for estimating the playback stop state is constructed by performing a simulation or the like, and the playback stop state is estimated using the model formula.
- the number of playback stops is input using the throughput average / deviation, coding bit rate, video length (duration), playback start threshold / playback stop threshold / playback restart threshold as inputs. And the playback stop time is obtained.
- Each formula used in this embodiment is an example, and other formulas may be used.
- FIG. 9 is a diagram illustrating processing executed by the reproduction stop state estimation function 12 of the sensation quality optimization device 1.
- Video duration the following are optional parameters (playback start threshold, playback restart threshold, playback stop threshold, coefficients e1 to e8, coefficients f1 to f6, DR_norm (normalized video duration), BR_norm (normalized bit rate) ), Estimate_unit, IT_dr, RT_dr, ST_dr)) is input, the average playback stop count SC and the average playback stop time SL are estimated and output to the OTT according to the following formulas 5, 6, and 5-6. (Step S121).
- SC is the average playback stop count
- SL is the average playback stop time
- T is the average throughput
- T_std is the throughput deviation
- BR is the encoding bit rate
- DR is the video duration.
- IT is a playback start threshold value.
- RT is a reproduction restart threshold
- ST is a reproduction stop threshold.
- e1 to e8 are constants
- f1 to f6 are constants
- DR_norm is a normalized video duration
- BR_norm is a normalized bit rate
- EU is an estimation unit.
- the playback start threshold IT, playback restart threshold is set from the playback threshold table T7 based on the service ID and design ID. A value RT and a reproduction stop threshold value ST are acquired.
- the coefficients e1 to e8 and the coefficients f1 to f6 are not input, the coefficients e1 to e8 of the equation 5 are obtained from the coefficient table 4 based on the service ID and the design ID, and the equation 6 is obtained from the coefficient table 5. Get the coefficients f1 to f6.
- DR_norm, BR_norm, estimate_unit, IT_dr, RT_dr, and ST_dr are not input, the DR_norm, BR_norm, estimate_unit, IT_dr, RT_dr, and ST_dr are acquired from the configuration file F1.
- the average reproduction stop count SC is estimated from the input information using Equation 5
- the average reproduction stop time SL is estimated from the input information using Equation 6.
- the QoE estimation technique of the present embodiment is a technique for estimating QoE from the encoding bit rate, resolution, and frame rate.
- a relationship between QoE and various encoding bit rates / resolutions / frame rates is obtained by a subjective evaluation experiment, and the relationship is expressed as a model formula.
- QoE can be calculated by inputting an encoding bit rate, resolution, and frame rate into the model formula.
- Equation 2 or the like QoE can be calculated in consideration of the number of times playback is stopped and the length of playback stop time.
- Each formula used in this embodiment is an example, and other formulas may be used.
- FIG. 10 is a diagram for explaining processing executed by the QoE estimation function 13 of the sensation quality optimization device 1.
- various input data (service ID, design ID, encoding bit rate, resolution, frame rate, etc.) from the OTT content delivery server via the parameter input / output unit 30 are input optional parameters (number of playback stop times, playback stop time length array)), first, the coefficient array 1 is obtained from the coefficient table 1 based on the service ID and design ID, and the following formula 1 (sign) QoE_C (intermediate parameter) is calculated in accordance with the calculation quality estimation) (step S131).
- bitrate is the encoding bit rate
- RS is the resolution
- FR is the frame rate
- a 1 to a 5 and b 1 to b 4 are constants.
- the playback stop time length array is not input when the number of playback stops is 1 or more, it is processed as an error.
- the coefficient array 2 is acquired from the coefficient table 2 based on the service ID and the design ID. Based on the coefficient array 2 and the QoE_C calculated by the expression 1, and the input data, the following expression 2 QoE_CR (intermediate parameter) is calculated according to (QoE estimation) (step S132).
- Min (A, B) is a function that returns a smaller value of A and B
- n is the number of playback stops
- RL i is the playback stop time length that occurred i-th.
- c 1 and c 2 are constants.
- the coefficient array 3 is acquired from the coefficient table 3 based on the service ID and the design ID. Based on the coefficient array 3 and the QoE_CR calculated by the expression 2 and the input data, the following expression 3 QoE is calculated according to (terminal correction), and is output to the OTT as estimated QoE (step S133).
- r is 1 when there is a playback stop and is 0 when there is no playback stop.
- D 1 to d i are constants.
- Equation 3 QoE_CR
- FIG. 11 is a diagram illustrating processing executed by the resolution / frame rate recommendation function 14 of the sensation quality optimization device 1.
- various input data (service ID, design ID, encoding bit rate, the following are optional parameters (reproduction) from the OTT content distribution server via the parameter input / output unit 30 (Stop count, playback stop time length, resolution array, frame rate array) are input, the recommended resolution, frame rate, and QoE are calculated and output to the OTT (step S141).
- the QoE estimation process by the QoE estimation function 13 is repeatedly executed for all resolution and frame rate pairs included in the input resolution array and frame rate array (step S142). , The resolution and frame rate at which the QoE is maximized, and the QoE are output.
- a resolution array corresponding to the service ID and design ID is acquired from the resolution table T8.
- the frame rate array corresponding to the service ID and design ID is acquired from the frame rate table T9.
- FIG. 12 is a diagram for explaining processing executed by the initial coding parameter recommendation function 15 of the sensation quality optimizing apparatus 1.
- FIG. 13 is a function included in the initial coding parameter recommendation function 15 (FIG. 12), which is based on the function 15-1 for outputting the overall QoE and the coding parameter table T12 from the coding bit rate array and the average throughput distribution. It is a figure explaining the process performed.
- FIG. 14 is a diagram for explaining processing executed by the function 15-2 that is included in the initial encoding parameter recommendation function 15 (FIG. 12) and outputs the average throughput distribution table T11.
- FIG. 15 is a function included in the function 15-1 (FIG. 13) that outputs the overall QoE and the coding parameter table T12 from the coding bit rate array and the average throughput distribution of the initial coding parameter recommendation function 15.
- FIG. 16 is a diagram for describing processing executed by a function 15-3 for converting from an encoding parameter table T12 to an average throughput-QoE table T13.
- an encoded bit rate array corresponding to the input service ID and design ID is acquired from the encoded bit rate table T10.
- the average throughput distribution table T11 is acquired by the function 15-2 (FIG. 14) that outputs the average throughput distribution table T11 described later (step S152).
- step S151 all the codes included in the encoded bit rate array list L1 output in step S151 are input using the input service ID and design ID and the average throughput distribution table T11 acquired in step S152.
- the overall average QoE is calculated and the coding parameter table T12 is output for the coded bit rate array by the function 15-1 (FIG. 13) described later (steps S153 and S154). That is, the overall average QoE and the encoding parameter table T12 are obtained for each encoding bit rate array included in the encoding bit rate array list L1.
- step S154 the coding parameter table T12 and the overall average QoE when the overall average QoE output in step S153 is maximized and the overall average QoE are output to the OTT (step S154).
- the function 15-1 for outputting the overall QoE and the encoding parameter table T12 from the encoded bit array and the average throughput distribution in FIG. 13, first, it is included in the encoded bit rate array input from step S154 (FIG. 12). For all coding bit rates, QoE, resolution, and frame rate are calculated in accordance with the resolution / frame rate recommendation function 14 (see FIG. 11), and the QoE, resolution, and frame rate for each coding bit rate array are calculated. The associated encoding parameter table T12 is generated (steps S15-11 and S15-12).
- the function 15-3 (FIG. 15) described later converts the coding parameter table T12 into the average throughput-QoE table T13 (step S15-13), and the converted average throughput-QoE table T13 and the step Based on the average throughput distribution table T11 acquired in S152 (FIG. 12), the overall average QoE is calculated according to the following Equation 4 (step S15-14).
- N in the above expression indicates which class (n-th class) in the average throughput class in the tables T11 and T13.
- step S15-11 the coding parameter table T12 generated in step S15-11 and the overall average QoE calculated in step S15-14 are output to step S154 in FIG.
- the average measured throughput corresponding to the log of the date and time up to and including the current date and time is selected, and the average throughput array is extracted (step S15-21).
- an average throughput distribution table T11 is generated from the average throughput array extracted from the viewing log R, and the function 15 -1 (FIG. 13) is output to step S15-14 (step S15-22).
- step S15-3 In the function 15-3 for converting from the coding parameter table T12 to the average throughput-QoE table T13 in FIG. 15, all of the coding parameter table T12 generated in step S15-11 of the function 15-1 (FIG. 13). Is converted into an average throughput to generate a Tmp table T14 (step S15-31).
- the encoding bit rate that is the target of the encoding parameter table T12 is greater than or equal to the encoding bit rate class lower limit value set in the configuration file F4 and less than the encoding bit rate class upper limit value.
- An offset value corresponding to the configuration file F4 is added to the encoding bit rate to convert it into an average throughput. In other cases, the coding bit rate is directly converted into average throughput.
- the average throughput-QoE table T13 is initialized based on the minimum value MIN, maximum value MAX, and width RANGE of the throughput class set in the configuration file F3 (step S15-32).
- 500 means MIN to MIN + RANGE
- 1000 means MIN + RANGE to MIN + RANGE * 2.
- the resolution, frame rate, and QoE corresponding to each average throughput class in the initialized average throughput-QoE table T13 are set.
- the resolution and frame rate of the minimum average throughput of the Tmp table T14 are set as the average throughput-QoE table. Set to T13. Note that the QoE calculation method in this case will be described in detail later in function 15-4 (see FIG. 16).
- the resolution, frame rate, and QoE of the average throughput smaller than the average throughput class and the maximum average throughput are set in the average throughput-QoE table T13.
- the resolution and the frame rate of the existing average throughput are set in the average throughput-QoE table T13.
- the resolution and frame rate of the minimum average throughput are set in the average throughput-QoE table T13. Note that the QoE calculation method in this case will be described in detail later in function 15-5 (see FIGS. 17 and 18).
- FIGS. 16A and 16B show the case where the average throughput below the upper limit of the average throughput class does not exist in the Tmp table T14 in the conversion function 15-3 from the encoding parameter table T12 to the average throughput-QoE table T13 (T ⁇ It is a figure explaining the calculation function 15-4 of QoE of (T_min).
- FIG. 15 is a diagram illustrating 15-5.
- area S1 and RANGE are shown in FIG. 17B.
- area S2 and RANGE are shown in FIG. 17B.
- area S3 and RANGE are shown in FIG. 17B.
- FIG. 15 is a diagram illustrating 15-5.
- FIG. 19 is a diagram for explaining processing executed by the real-time coding parameter recommendation function 16 of the sensation quality optimizing apparatus 1.
- FIG. 20 is a diagram for explaining processing executed by the function 16-1 that is included in the real-time encoding parameter recommendation function 16 (FIG. 19) and that creates an average throughput probability density table.
- FIG. 21 is a diagram for explaining processing executed by the function 16-2 for calculating the local average QoE and QoE, which is a function included in the real-time encoding parameter recommendation function 16 (FIG. 19).
- FIG. 22 is a function included in the function 16-2 (FIG. 21) for calculating the local average QoE and QoE of the real-time encoding parameter recommendation function 16, and the function 16 for creating the average throughput-reproduction stop state table T17.
- FIG. 3 is a diagram illustrating a process executed by -3.
- FIG. 23 is a function included in the function 16-2 (FIG. 21) for calculating the local average QoE and QoE of the real-time encoding parameter recommendation function 16, and includes the function 16-4 for creating the average throughput-QoE table. It is a figure explaining the process performed.
- FIG. 24 is a function included in the function 16-4 (FIG. 23) for creating the average throughput-QoE table T18 of the real-time coding parameter recommendation function 16, and is executed by the QoE estimation function 16-5.
- FIG. 24 is a function included in the function 16-4 (FIG. 23) for creating the average throughput-QoE table T18 of the real-time coding parameter recommendation function 16, and is executed by the QoE estimation function 16-5.
- an average throughput probability table creation function 16-1 (FIG. 20) described later.
- An average throughput probability table T15 is created and output to the function 16-2 (FIG. 21) for calculating the local average QoE and QoE described later (step S161).
- the average throughput probability table T15 output from the average throughput probability table creation function 16-1 (FIG. 20) is input, and all the encoding bit rates, resolutions, and frames in the input OTT encoding parameter table are input.
- the local average QoE is calculated by the local average QoE and QoE calculation function 16-2 (step S162).
- an encoding parameter array list is acquired from the encoding parameter set table T16, and the set of the encoding bit rate, resolution, and frame rate is set to the local average QoE and QoE. Input to the calculation function 16-2 (step S163).
- the local average QoE and the local average QoE calculated by the QoE calculation function 16-2 are maximized, and the resolution, the frame rate, the encoding bit rate, and the local average QoE are output to the OTT.
- the frame rate and the encoding bit rate are input to the QoE estimation function 13 (FIG. 10) (step S163).
- the average throughput probability table T15 is initialized based on the minimum value MIN, maximum value MAX, and width RANGE of the throughput class set in the configuration file F3. (Step S16-11).
- the average throughput class in the average throughput probability table T15 is a value that increases from the minimum value MIN by the width RANGE.
- the user performance flag set in the configuration file F5 and the throughput average average, user performance average average, throughput average deviation, and user performance average deviation estimated by the throughput estimation function 11 are input.
- the probability of each row of the average throughput probability table T15 is calculated, and the average throughput probability table T15 is output to the local average QoE and QoE calculation function 16-2 (FIG. 21) (step 16). -12).
- the average throughput average is m.
- the average user performance average is m, but when the average user performance average is not input, the average throughput average is m.
- the throughput average deviation is set to s.
- the user performance flag is on “1”
- the user performance average deviation is set to s.
- the throughput average deviation is set to s.
- an average throughput-reproduction stop state table T17 is created by a function 16-3 (FIG. 22) for creating an average throughput-reproduction stop state table T17 described later (step S16-21).
- the input service ID, design ID, encoding bit rate, resolution, frame rate, video duration, and average throughput-reproduction stop state table T17 created by the function 16-3 (FIG. 22) are input.
- an average throughput-QoE conversion table T18 is created by a function 16-4 (FIG. 23) for creating an average throughput-QoE table, which will be described later (step S16-22).
- the local average QoE is calculated based on the following Equation 8, and is output to the process of step S163 in FIG. 19 (step S16-23).
- LAQoE is a local average QoE.
- PDn is the probability of the average throughput class n in the average throughput probability table T15.
- QoEn is the QoE of the average throughput class n in the average throughput-QoE conversion table T18.
- MTC [] is an average throughput class sequence of the average throughput-probability table.
- the average throughput-reproduction stop is based on the minimum value MIN, maximum value MAX, and width RANGE of the throughput class set in the configuration file F3.
- the state table T17 is initialized (step S16-31).
- the average throughput-reproduction stop state table T17 For each row of the average throughput-reproduction stop state table T17, a value obtained by adding 1/2 of the RANGE to the corresponding average throughput class (average throughput class + RANGE / 2) is set as the average throughput, and the reproduction stop An average reproduction stop count SC and an average reproduction stop time SL are calculated by the state estimation function 12 (see step S121 in FIG. 9) (step S16-32).
- the average throughput-reproduction stop state table T17 thus created is output to the average throughput-QoE table creation function 16-4 (FIG. 23).
- the average throughput-QoE conversion table T18 is obtained based on the minimum value MIN, maximum value MAX, and width RANGE of the throughput class set in the configuration file F3. Initialization is performed (step S16-41).
- each corresponding QoE is calculated by the QoE estimation function 16-5 described later, and is substituted as the QoE of the corresponding average throughput class in the average throughput-QoE conversion table T18 (step S16-42).
- the average throughput-QoE conversion table T18 thus created is output to step S16-23 of the local average QoE and QoE calculation function 16-2 (FIG. 21).
- the input service ID, design ID, average playback stop count SC, average playback stop time SL, video duration duration, constants H1, H2, set by the configuration file F6 H3, duration_min, duration_max, the coefficient array 6 (g1, g2) and the coefficient array 7 (h1, h2) acquired from the coefficient tables 6 and 7 corresponding to the service ID and design ID are input, and the following formula 9 Based on the above, QoE is calculated (step S16-51).
- Min (A, B) in the above formula is a function that returns a smaller value of A and B.
- SC is the average number of playback stops
- SL is the average playback stop time.
- g 1 , g 2 , h 1 , h 2 are constants
- duration is a video duration.
- duration_min, duration_max, and H 1 to H 3 are constants defined in the configuration file.
- the intermediate parameter QoE_C2 used in Equation 9 is a coefficient table 8 corresponding to the input service ID, design ID, encoding bit rate Bitrate, resolution RS, frame rate FR, the service ID, and design ID.
- the coefficient array 8 (i1 to i7) acquired from the above is used as an input and calculated based on the following formula 10 (step S16-52).
- Bitrate is the encoding bit rate
- RS is the resolution
- FR is the frame rate.
- i 1 to i 7 are constants.
- Formulas 9 and 10 are also model formulas obtained through experiments as described above.
- QoE QoE_C2
- QoE_C2 is calculated by inputting the encoding bit rate, resolution, and frame rate.
- QoE is calculated by further adding the number of playback stops and the playback stop time.
- the resolution, the frame rate, the encoding bit rate, and the local average QoE that maximize the local average QoE obtained by the real-time encoding parameter recommendation function 16 are output to the OTT.
- FIG. 25 is a diagram illustrating processing executed by the throughput table information update function 21 of the sensation quality optimization device 1.
- the update target designation information (calendar information and usage) is automatically started once every 1 to 5 minutes.
- NW name a use server
- base station ID a use server
- the measured throughput average is included according to the calendar information and the usage server (calendar information, usage server) -log of the speed conversion table T6. Then, update is performed using the latest (21_log) logs acquired from the viewing log R stored in the storage unit 24 (step S212).
- the measured throughput average is included according to the calendar information and NW name (or base station ID).
- (Calendar information, NW name) -speed conversion table T5 (or (calendar information, base station ID) -speed conversion table T4) log acquired from the viewing log R stored in the storage unit 24 (21_log) Update is performed using the logs (step S213).
- step S257 the target row of the corresponding speed conversion table T4 / T5 / T6 is updated for the average throughput average, the average throughput deviation, and the average throughput deviation calculated in step S214 (step S215).
- the throughput table information update function 21 performs the (calendar information, base station ID) -speed conversion table T4, (calendar information, NW name) -speed conversion table T5, (calendar information) for each call by the automatic activation. , Use server)-for each of the speed conversion tables T6, the log including the information of the average measured throughput is a combination accumulated in the viewing log R more than the number (21_threshold) of the calculation conditions from the previous update. Run against.
- the location corresponding to the current calendar information and the calendar information one hour later is calculated based on the time when the system is operating.
- the (calendar information, base station ID) -speed conversion table T4 (calendar information, NW name) -speed conversion table T5, (calendar information, use server) -speed conversion table T6 is always updated to the latest state.
- calculation of each function in the estimated value / recommendation value calculation unit 10 can be executed.
- FIG. 26A and FIG. 26B are diagrams illustrating processing executed by the viewing log update function 22 of the sensation quality optimizing apparatus 1.
- FIG. 26A shows a processing procedure
- FIG. 26B shows a viewing log.
- the input data includes (date / time, user ID, service ID, design ID)
- the date / time refer to the (day of the week, time) -calendar information table T1.
- Calendar information corresponding to the date and time is added, and if an IP address is included, an NW name corresponding to the IP address is added by referring to the IP address-ISP conversion table T2, and a carrier name Is included, a carrier identifier corresponding to the carrier name is added with reference to the carrier name-carrier identifier conversion table T3 (step S222).
- step S223 it is determined whether or not there is a record in the viewing log R that matches the input (date, user ID, service ID, design ID) combination.
- step S224 If there is a record in the viewing log R that matches the combination of the input (date, user ID, service ID, design ID), the combination of the (date, user ID, service ID, design ID) The contents of the column input as an option of the combination are updated with respect to the records that match (step S224).
- the record of the viewing log R is always updated to the latest state, and the calculation of each function in the estimated value / recommendation value calculation unit 10 can be executed.
- the OTT recommendation requesting device 40 optimizes the QoE of the content distribution service to the experience quality optimization device 1 via the parameter input / output unit 30.
- a recommendation request and various distribution parameters are input.
- various setting values coefficient arrays are stored in the storage unit 24.
- the throughput average function, the throughput average deviation, and the like are estimated by the throughput estimation function 11.
- the real-time encoding parameter recommendation function 16 calculates an encoding parameter (including resolution and frame rate corresponding to each encoding bit rate) when the local average QoE is maximum, and uses the parameter as a recommendation value.
- the data is output to the content distribution system OTT via the input / output unit 30.
- the storage unit 24 corresponds to the input distribution parameters (including the encoding bit rate, resolution, and frame rate), the distribution parameters, and the input data from the system administrator.
- QoE is estimated by the QoE estimation function 13 based on various setting values (including the coefficient array) stored in. Then, the resolution and frame rate at which the estimated QoE is maximized are calculated by the resolution / frame rate recommendation function 14, and the coding parameters when the average QoE of the entire content distribution target users is maximized (for each coding bit rate). (Including the corresponding resolution and frame rate) is calculated by the initial encoding parameter recommendation function 15, and is output as a recommendation value to the content distribution system OTT via the parameter input / output unit 30.
- distribution parameters such as a distribution rate of the content distribution service can be set according to the encoding parameter that maximizes the local average QoE acquired as the recommendation value and the average QoE of the entire content distribution target users. Therefore, it is possible to select a content distribution method that maximizes the user's QoE, and an improvement in quality such as almost no rebuffering with a fixed distribution rate can be expected.
- various setting values (including various tables Tn and various coefficients (constants)) stored in the storage unit 24 are transmitted via the parameter input / output unit 30. Since the setting value update unit 20 updates the distribution parameter based on the newly input distribution parameter and the input data from the system administrator, the setting value is always updated to the latest state to estimate the QoE and the QoE. It is possible to calculate the resolution and frame rate that maximize, and the encoding parameter that maximizes the overall average QoE.
- the problems described in (1) to (3) of the problems to be solved by the invention can be solved by the technique described in the present embodiment. Specifically, it is possible to select a content distribution method that maximizes the user's QoE, and an improvement in quality such as almost no rebuffering with a fixed distribution rate can be expected.
- each processing method by the sensation quality optimization system described in each of the above embodiments that is, from the recommendation request processing and the sensation quality optimization device 1 with output of delivery parameters from the OTT by the recommendation request device 40 of FIG. Estimation value / recommendation value reception processing, processing of the throughput estimation function 11 shown in FIG. 8, processing of the reproduction stop state estimation function 12 shown in FIG. 9, processing of the QoE estimation function 13 shown in FIG. 10, resolution shown in FIG. Processing of the frame rate recommendation function 14, processing of the initial encoding parameter recommendation function 15 shown in FIGS. 12 to 18, processing of the real-time encoding parameter recommendation function 16 shown in FIGS. 19 to 24, throughput table information shown in FIG.
- Programs that can be executed by a computer include memory cards (ROM cards, RAM cards, etc.), magnetic disks (floppy (registered trademark) disks, hard disks, etc.), optical disks (CD-ROM, DVD, etc.), semiconductor memories, etc. It can be stored and distributed on a medium of the external recording device. Then, the computer (control device: CPU) of the electronic device reads the program stored in the medium of the external storage device into the storage device, and the operation is controlled by the read program.
- the recommended request function, the estimated value / recommendation value calculation function, and the set value update function can be realized, and the same processing as described above can be executed.
- the program data for realizing each of the above methods can be transmitted on the NW in the form of a program code, and the program data is taken into the electronic device from the computer device connected to the NW, and the storage device And the above-described recommendation request function, estimated value / recommendation value calculation function, and set value update function can be realized.
- each of the embodiments includes inventions at various stages, and various inventions can be extracted by appropriately combining a plurality of disclosed constituent elements. For example, even if some constituent elements are deleted from all the constituent elements shown in each embodiment or some constituent elements are combined in different forms, the problems described in the column of the problem to be solved by the invention If the effects described in the column “Effects of the Invention” can be obtained, a configuration in which these constituent requirements are deleted or combined can be extracted as an invention.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Databases & Information Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Human Computer Interaction (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
- Information Transfer Between Computers (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
(1)利用者が配信レートを選択する方法
利用者側でQoEが影響する配信レートを設定するためのインターフェースを準備しておき、利用者が自身の通信環境などを考慮して配信レートを設定する方法が検討されている。配信レートとして、高/中/低などの3パターンの品質を準備する場合が多く利用されている。現在、YouTube(登録商標)などの映像配信サービスでは、画質を選択できるインターフェースが存在している(非特許文献1参照)。
(2)コンテンツ配信事業者側で配信レートを選択する方法
コンテンツ配信サービスの利用者数が多い時間帯は低レートの品質でのコンテツン配信を行ない、利用者数が少ない時間帯には中/高レートのより高品質なコンテンツ配信を行なう方法である。これにより、混雑時でもサービス利用者に最低限満足可能な品質でのサービス提供が可能となる。また、利用者数やプレミアムの支払いに応じて高品質なサービス提供も可能となる。現在、ニコニコ動画(登録商標)では無料会員に対して、混雑時に配信レートを低下させる様な施策を実施している(非特許文献2参照)。
(3)通信状況に応じてコンテンツの配信レートを選択する方法
この方法では、何らかの形で利用者の可用帯域を推定し、可用帯域に応じたビットレートでコンテンツを配信する方法である。可用帯域の推定方式から、幾つかに分類される。
a. パッシブ型推定方式
パッシブ型の推定方式においては、過去の通信履歴等から可用帯域を推定する。
b. アクティブ型推定方式
アクティブ型の推定方式においては、配信前に短い通信を追加で行ない、その際の通信性能を参考として、可用帯域を推定する。例えば、通信前にパケットトレインという、逐次的に大きくなるパケット群を送り、到着までの遅延の変化量から可用帯域を推定する(非特許文献4参照)。
c. フィードバック方式
フィードバック方式においては、配信を行っている途中に、その通信自体の速度を測定して、その際の通信速度に合わせて、リアルタイムで配信レートを変更するものである。この方式は、主に、途中で配信レートを変更できる、映像配信に適用される。この方式を取り入れている映像配信では、チャンク型と呼ばれる短い時間単位のファイルにコンテンツを分割し、チャンクごとにサーバ/クライアント間でネットワーク品質(NW品質)等を計測し、計測結果に基づいて、次に端末が受信するチャンクの配信レートを決定し、コンテンツを配信する。これにより、利用者の端末の再生バッファにコンテンツを蓄積し再生する事で、低レートでも品質の高いコンテンツ配信が可能となる(非特許文献5,6参照)。
(1)利用者が配信レートを選択する方法
利用者が配信レートを選択する方式では、利用者の好みにあった映像配信レートを設定出来るメリットはあるものの、実際の通信環境よりも高い配信レートを選択する事によって生じるリバッファリングによる体感品質低下や、出先などの自宅以外の環境で視聴する際に、利用者が通信環境に応じた適切な配信レートの設定が困難であるという問題がある。
(2)コンテンツ配信事業者側で配信レートを選択する方法
コンテンツ配信事業者側で配信レートを選択する方法は、配信側設備量に応じた配信レートの設定を行える半面、利用者個々の通信環境に応じたレート制御が行われないため、通信環境に対して過大なレートでコンテンツ映像配信を行ってしまう(リバッファリング)事象や、過少レートでコンテンツ配信を行ってしまう事象が発生するという問題がある。
(3)通信状況に応じてコンテンツ配信レートを選択する方法
a. パッシブ型推定方式
パッシブ型の推定方式は、エンドツーエンドの十分な過去の通信履歴が必要であるため、これまで、十分なデータを揃えるのが困難であった。
b. アクティブ型推定方式
アクティブ型の推定方式では、通信前に可用帯域を測定するための時間が追加でかかってしまうため、コンテンツの配信に時間がかかってしまうという問題がある。通信開始までの待ち時間はweb系のアプリケーションでは体感品質に大きく影響してしまう。
c. フィードバック方式
当該配信方式では、再生停止を伴うリバッファリングの発生頻度は低くなるが、映像視聴中に配信レート(解像度)が変動するため、視聴者のQoEは高くない事が被験者評価によって確認されている。また、リバッファリングを少なくする事を目的に配信レートの設定を行なうため、配信レートを低下させすぎ、QoEがリバッファリング発生時よりも低下する事象を引き起こす事がある。
本実施の形態におけるスループット推定機能11が使用する技術は、ユーザのネットワーク環境の過去の視聴ログ(スループットの実績値)から、現在のスループットを推定する技術である。
・日時:エラーコードを返す。
・ユーザID:視聴ログ検索機能をNoとして進める。
・固定/モバイル識別子:固定として扱う。
・固定なのにIPアドレスがない:エラーコードを返す。
・モバイルなのにwi-fi利用判定がない:エラーコードを返す。
・モバイルでwi-fi利用判定がtrueなのに、IPアドレスがない:エラーコードを返す。
・モバイルでwi-fi利用判定がfalseなのに、キャリア名がない:エラーコードを返す。
・モバイルでwi-fi利用判定がfalseなのに、座標もセルIDもない:エラーコードを返す。
視聴ログ更新機能22の必須入力パラメータ(日時,ユーザID,サービスID,設計ID)の全てがある場合は、他の情報が不足している場合も、当該視聴ログ更新機能22を実行する。
・モバイルの速度計算の際に値が取得できなかった場合:エラーコードを返す。
・ユーザ実績平均偏差が取得できなかった場合:ユーザ実績平均平均、ユーザ実績平均偏差、ユーザ実績偏差平均は付加しない。
・(曜日,時刻)-カレンダ情報変換テーブルT1:エラーコードを返す。
・IPアドレス-ISP変換テーブルT2及びキャリア名-キャリア識別子変換テーブルT3:エラーコードを返す。また、検索しようとして見つからなかった値を管理者に通知する。
・(カレンダ情報,NW名)-速度変換テーブルT5:エラーコードを返す。
・(カレンダ情報,利用サーバ)-速度変換テーブルT6:エラーコードを返す。
本実施の形態の再生停止状態推定機能12における再生停止状態推定技術は、スループット推定技術の出力値であるスループット平均/偏差、符号化ビットレート、動画長(デュレーション)、再生開始しきい値/再生停止しきい値/再生再開しきい値から、再生停止回数/再生停止時間を推定する技術である。なお、再生停止の仕組みに関し、例えば、プログレッシブダウンロード型の動画配信においては、端末は受信データを蓄積するバッファを保有し、バッファの残量に応じて、再生開始、停止、再開の動作が決定される。例えば、バッファ量が再生停止しきい値まで少なくなれば、再生が停止し、停止状態でバッファ量が再生再開しきい値になれば、再生が再開される。本技術では、シミュレーション等を行うことで、再生停止状態を推定するモデル式を構築し、モデル式を用いて再生停止状態推定を行う。
本実施の形態のQoE推定技術は、符号化ビットレート、解像度、及びフレームレートからQoEを推定する技術である。本技術では、例えば、主観評価実験等により、QoEと、様々な符号化ビットレート/解像度/フレームレートとの関係を求め、その関係をモデル式として表わす。以下の数式(モデル式)で具体例が示されているとおり、当該モデル式に、符号化ビットレート、解像度、フレームレートを入力することで、QoEを算出することができる。また、数式2等に示すように、再生停止回数と再生停止時間長も考慮してQoEを算出することができる。なお、本実施の形態で使用される各数式は例であり、他の数式を用いることとしてもよい。
図11は、前記体感品質最適化装置1の解像度・フレームレートレコメンド機能14により実行される処理を説明する図である。
図12は、前記体感品質最適化装置1の初期符号化パラメータレコメンド機能15により実行される処理を説明する図である。
QoE(T)=S1/RANGE=1 となる。この計算は、図16BにおけるCase1として図示されている。
QoE(T)=S2/RANGE
S2=(QoE_min-1)/B*(T_up-(T_min-B))^2/2
- (QoE_min-1)/B*(T-(T_min-B))^2/2 + 1*(T_up-T) となる。
QoE(T)=S3/RANGE
S3=(QoE_min-1)/B*(T_up-(T_min-B))^2/2 + 1*(T_up-T) となる。
QoE=S1/RANGE
S1=(1*(T[0]-B-A[0])+((QoE[0]-1)/2+1)*B+QoE[0]*(A[1]-T[0])) となる(ステップS15-51)。
QoE=S2/RANGE
S2=(QoE[0]*(T[1]-A[1])+QoE[1]*(A[2]-T[1])) となる(ステップS15-52)。
QoE=S3/RANGE
S3=(QoE[1]*(T[2]-A[3])+QoE[2]*(T[3]-T[2])+QoE[3]*(A[3]+RANGE-T[3])) となる(ステップS15-53)。
平均スループット階級A[0]のQoEの場合(ステップS15-51)は、
QoE[0]*(A[1]-T[0])がN個になり、
平均スループット階級A[1]のQoEの場合(ステップS15-52)は、
QoE[1]*(A[2]-T[1])がN個になり、
平均スループット階級A[3]のQoEの場合(ステップS15-53)は、
QoE[2]*(T[3]-T[2])+QoE[3]*(A[3]+RANGE-T[3])がN個になる。
QoE=S1/RANGE
S1=(QoE[0]-1)*B/2 - (QoE[0]-1)/B*(A[0]-(T[0]-B))^2/2
+ 1*(T[0]-A[0]) + QoE[0]*(A[1]-T[0]) となる(ステップS15-54)。
QoE[0]*(A[1]-T[0])がN個になる。
図19は、前記体感品質最適化装置1のリアルタイム符号化パラメータレコメンド機能16により実行される処理を説明する図である。
図25は、前記体感品質最適化装置1のスループットテーブル情報更新機能21により実行される処理を説明する図である。
図26A、図26Bは、前記体感品質最適化装置1の視聴ログ更新機能22により実行される処理を説明する図である。図26Aは処理の手順を示し、図26Bは視聴ログを示す。
10 …推定/レコメンド機能(推定値・レコメンド値算出部)
11 …スループット推定機能
12 …再生停止状態推定機能
13 …QoE推定機能
14 …解像度・フレームレートレコメンド機能
15 …初期符号化パラメータレコメンド機能
16 …リアルタイム符号化パラメータレコメンド機能
20 …データ蓄積機能(設定値更新部)
21 …スループットテーブル情報更新機能
22 …視聴ログ更新機能
23 …設定値参照機能
24 …記憶部
25 …入出力装置
30 …パラメータ入出力部
40 …レコメンド要求装置
41 …配信パラメータ出力部
42 …レコメンド値受信部
OTT…コンテンツ配信システム
Claims (9)
- コンテンツ配信サービスにおいて、ユーザが体感する品質(QoE:Quality of Experience)を最適化するための体感品質最適化システムであって、
コンテンツ配信のQoEを最適化するために、当該コンテンツ配信の配信パラメータ候補を出力し、当該QoEを最適化する配信パラメータをレコメンド値として受信するレコメンド要求装置と、
前記レコメンド要求装置から受信した前記配信パラメータ候補からQoEを推定し、当該QoEを最適化する配信パラメータをレコメンド値として算出し、当該レコメンド値を出力する体感品質最適化装置と、
を備えた体感品質最適化システム。 - コンテンツ配信サービスにおいて、ユーザが体感する品質(QoE:Quality of Experience)を最適化するための体感品質最適化装置であって、
コンテンツ配信のための配信パラメータ候補を入力し、QoEを最適化するレコメンド値を出力するパラメータ入出力部と、
前記パラメータ入出力部により入力された配信パラメータ候補に基づいて、QoEを推定し、当該QoEを最適化する配信パラメータを前記レコメンド値として算出する推定値・レコメンド値算出部と、
を備えた体感品質最適化装置。 - 前記推定値・レコメンド値算出部は、
前記配信パラメータ候補の符号化に関わる情報からQoEを算出するQoE推定機能部と、
前記QoE推定機能部により算出されたQoEが最大になる解像度,フレームレートを算出する解像度・フレームレートレコメンド機能部と、
コンテンツ配信対象ユーザ全体の平均QoEが最大になる符号化パラメータを計算する初期符号化パラメータレコメンド機能部と、
を有する請求項2に記載の体感品質最適化装置。 - 前記推定値・レコメンド値算出部は、前記パラメータ入出力部により入力された配信パラメータ候補とシステム管理者からの入力データに応じた設定値に基づいて、QoEを推定し、当該QoEを最適化する配信パラメータを前記レコメンド値として算出し、
前記パラメータ入出力部により新たに入力された配信パラメータ候補、又は前記システム管理者からの新たな入力データに基づいて、前記設定値の更新を行なう設定値更新部をさらに有する、
請求項2又は請求項3に記載の体感品質最適化装置。 - コンテンツ配信サービスにおいて、QoEを最適化するためのレコメンド要求装置であって、
QoEを最適化する配信パラメータをレコメンド値として要求するために、コンテンツ配信のための配信パラメータ候補を体感品質最適化装置へ出力する配信パラメータ出力部と、
前記体感品質最適化装置から前記レコメンド値を受信するレコメンド値受信部と、
を有するレコメンド要求装置。 - コンテンツ配信サービスにおいて、体感品質最適化装置が実行するQoEを最適化するための体感品質最適化方法であって、
コンテンツ配信のための配信パラメータ候補を入力する配信パラメータ入力ステップと、
前記配信パラメータ入力ステップにて入力された配信パラメータ候補に基づいて、QoEを推定し、当該QoEを最適化する配信パラメータをレコメンド値として算出する推定値・レコメンド値算出ステップと、
前記推定値・レコメンド値算出ステップにて算出されたレコメンド値を出力するレコメンド値出力ステップと、
を有する体感品質最適化方法。 - コンテンツ配信サービスにおいて、QoEを最適化するためにレコメンド要求装置が実行するレコメンド要求方法であって、
QoEを最適化する配信パラメータをレコメンド値として要求するために、コンテンツ配信のための配信パラメータ候補を体感品質最適化装置へ出力する配信パラメータ出力ステップと、
前記体感品質最適化装置から前記レコメンド値を受信するレコメンド値受信ステップと、
を有するレコメンド要求方法。 - コンピュータを、前記請求項2乃至請求項4の何れか1項に記載の体感品質最適化装置における各部として機能させるためのプログラム。
- コンピュータを、前記請求項5に記載のレコメンド要求装置における各部として機能させるためのプログラム。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/548,108 US10681416B2 (en) | 2015-02-04 | 2015-12-24 | Quality-of-experience optimization system, quality-of-experience optimization apparatus, recommend request apparatus, quality-of-experience optimization method, recommend request method, and program |
CN201580075264.6A CN107211178B (zh) | 2015-02-04 | 2015-12-24 | 体感质量最佳化系统、装置、方法以及建议请求装置、方法和可读存储介质 |
JP2016573208A JP6572245B2 (ja) | 2015-02-04 | 2015-12-24 | 体感品質最適化システム、体感品質最適化装置、レコメンド要求装置、体感品質最適化方法、レコメンド要求方法及びプログラム |
EP15881232.1A EP3255894B1 (en) | 2015-02-04 | 2015-12-24 | Quality-of-experience optimization system, quality-of-experience optimization device, recommendation request device, quality-of-experience optimization method, recommendation request method and program |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2015-020500 | 2015-02-04 | ||
JP2015020500 | 2015-02-04 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016125415A1 true WO2016125415A1 (ja) | 2016-08-11 |
Family
ID=56563772
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2015/086087 WO2016125415A1 (ja) | 2015-02-04 | 2015-12-24 | 体感品質最適化システム、体感品質最適化装置、レコメンド要求装置、体感品質最適化方法、レコメンド要求方法及びプログラム |
Country Status (5)
Country | Link |
---|---|
US (1) | US10681416B2 (ja) |
EP (1) | EP3255894B1 (ja) |
JP (1) | JP6572245B2 (ja) |
CN (1) | CN107211178B (ja) |
WO (1) | WO2016125415A1 (ja) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019016961A (ja) * | 2017-07-07 | 2019-01-31 | 日本電信電話株式会社 | 動画品質制御装置、ビットレート選択方法、及びプログラム |
WO2021124387A1 (ja) * | 2019-12-16 | 2021-06-24 | 日本電信電話株式会社 | 符号化データ生成方法、符号化データ生成装置、及びプログラム |
WO2022185534A1 (ja) * | 2021-03-05 | 2022-09-09 | 日本電信電話株式会社 | 符号化パラメータ導出装置、符号化パラメータ導出方法及びプログラム |
WO2022190263A1 (ja) * | 2021-03-10 | 2022-09-15 | 日本電信電話株式会社 | 情報処理装置、分析方法およびプログラム |
WO2022190264A1 (ja) * | 2021-03-10 | 2022-09-15 | 日本電信電話株式会社 | 情報処理装置、分析方法およびプログラム |
WO2022201225A1 (ja) * | 2021-03-22 | 2022-09-29 | 日本電信電話株式会社 | 制御装置、制御方法及びプログラム |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10674409B2 (en) * | 2017-06-09 | 2020-06-02 | At&T Intellectual Property I, L.P. | System and method for fine grained service management using SDN-NFV networks |
JP6992813B2 (ja) * | 2017-08-30 | 2022-01-13 | 日本電気株式会社 | 動画再生ビットレート推定装置、動画再生ビットレート推定方法、動画再生ビットレート推定プログラム及び通信品質測定装置 |
JP7082282B2 (ja) * | 2018-06-06 | 2022-06-08 | 富士通株式会社 | パケット解析プログラム、パケット解析方法およびパケット解析装置 |
US10819760B2 (en) | 2018-10-01 | 2020-10-27 | At&T Intellectual Property I, L.P. | Method and apparatus for streaming video applications in cellular networks |
US11128869B1 (en) * | 2018-10-22 | 2021-09-21 | Bitmovin, Inc. | Video encoding based on customized bitrate table |
US20230198645A1 (en) * | 2021-12-21 | 2023-06-22 | Dish Network Technologies India Private Limited | Methods, devices, and systems for distributing content in community of over-the-air broadcast content receivers |
US11824766B2 (en) * | 2022-04-13 | 2023-11-21 | Cisco Technology, Inc. | Adaptive selection of network paths based on long-term predictions of user experience |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2012070373A (ja) * | 2010-09-13 | 2012-04-05 | Ntt Docomo Inc | ビデオストリームを転送するための方法及び装置 |
US20130091248A1 (en) * | 2011-10-05 | 2013-04-11 | Alcatel-Lucent Bell N.V. | Method and apparatus for improving adaptive streaming video quality by optimizing resource allocation |
WO2013123467A1 (en) * | 2012-02-17 | 2013-08-22 | Vid Scale, Inc. | Hierarchical traffic differentiation to handle congestion and/or manage user quality of experience |
US20130275557A1 (en) * | 2012-04-12 | 2013-10-17 | Seawell Networks Inc. | Methods and systems for real-time transmuxing of streaming media content |
WO2014022017A1 (en) * | 2012-08-03 | 2014-02-06 | Intel Corporation | Methods for quality-aware adaptive streaming over hypertext transfer protocol |
Family Cites Families (31)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5742892A (en) * | 1995-04-18 | 1998-04-21 | Sun Microsystems, Inc. | Decoder for a software-implemented end-to-end scalable video delivery system |
FI109072B (fi) * | 1999-06-16 | 2002-05-15 | Nokia Corp | Menetelmä ja järjestely kanavakoodaus- ja lomitusmenettelyn valitsemiseksi eräissä pakettidatayhteyksissä |
US7424268B2 (en) * | 2002-04-22 | 2008-09-09 | Cisco Technology, Inc. | System and method for management of a shared frequency band |
US7747255B2 (en) * | 2003-03-26 | 2010-06-29 | Sony Corporation | System and method for dynamic bandwidth estimation of network links |
JP2005354126A (ja) * | 2004-06-08 | 2005-12-22 | Hitachi Communication Technologies Ltd | 無線通信端末、無線基地局及び無線通信システム |
CN1998255B (zh) * | 2004-07-05 | 2012-11-14 | 艾利森电话股份有限公司 | 用于推送消息启动服务的设备和方法 |
US8149264B2 (en) * | 2004-07-27 | 2012-04-03 | Telecom Italia S.P.A. | Video-communication in mobile networks |
US7773672B2 (en) * | 2006-05-30 | 2010-08-10 | Freescale Semiconductor, Inc. | Scalable rate control system for a video encoder |
FR2903259A1 (fr) * | 2006-06-29 | 2008-01-04 | Thomson Licensing Sa | Procede de gestion de demandes d'acces a distance a des contenus multimedia |
US20080016185A1 (en) * | 2006-07-11 | 2008-01-17 | Magix Ag | System and method for dynamically creating online multimedia slideshows |
US20080278582A1 (en) * | 2007-05-07 | 2008-11-13 | Sentinel Ave Llc | Video Fusion Display Systems |
US8819106B1 (en) * | 2008-12-12 | 2014-08-26 | Amazon Technologies, Inc. | Managing distributed execution of programs |
FR2954036B1 (fr) * | 2009-12-11 | 2012-01-13 | Thales Sa | Procede et systeme pour la determination de parametres de codage sur des flux a resolution variable |
KR20110102135A (ko) * | 2010-03-10 | 2011-09-16 | 엘지전자 주식회사 | 광대역 무선 접속 시스템에서 aGP서비스의 스케쥴링 방법 및 장치 |
US9107197B2 (en) * | 2011-04-13 | 2015-08-11 | Telefonaktiebolaget L M Ericsson (Publ) | Terminal-based selection of radio parameters among a parameter subset offered by the network |
KR101928910B1 (ko) * | 2011-08-30 | 2018-12-14 | 쏘닉 아이피, 아이엔씨. | 복수의 최대 비트레이트 레벨들을 사용하여 인코딩된 비디오를 인코딩하고 스트리밍하기 위한 시스템들 및 방법들 |
US9819717B2 (en) * | 2011-12-28 | 2017-11-14 | Intel Corporation | Video adaptation for content-aware wireless streaming |
US8762563B2 (en) * | 2012-04-16 | 2014-06-24 | Adobe Systems Incorporated | Method and apparatus for improving the adaptive bit rate behavior of a streaming media player |
US8792347B2 (en) * | 2012-06-01 | 2014-07-29 | Opera Software Ireland Limited | Real-time network monitoring and subscriber identification with an on-demand appliance |
US9571827B2 (en) * | 2012-06-08 | 2017-02-14 | Apple Inc. | Techniques for adaptive video streaming |
US9357272B2 (en) | 2012-08-03 | 2016-05-31 | Intel Corporation | Device orientation capability exchange signaling and server adaptation of multimedia content in response to device orientation |
US9172646B2 (en) * | 2013-03-15 | 2015-10-27 | International Business Machines Corporation | Dynamic reconfiguration of network devices for outage prediction |
US10187433B2 (en) * | 2013-03-15 | 2019-01-22 | Swyme Ip Bv | Methods and systems for dynamic adjustment of session parameters for effective video collaboration among heterogenous devices |
EP2806633A1 (en) * | 2013-05-23 | 2014-11-26 | Alcatel Lucent | Method and apparatus for improved network optimization for providing video from a plurality of sources to a plurality of clients |
US9413806B2 (en) * | 2013-05-24 | 2016-08-09 | Broadcom Corporation | Adaptive bit rate proxy |
CN105637925B (zh) * | 2013-10-16 | 2019-05-21 | 英派尔科技开发有限公司 | 具有细化策略管理的动态频率和功率资源分配 |
US10321361B2 (en) * | 2014-10-10 | 2019-06-11 | Huawei Technologies Co., Ltd. | Method and device for requesting a quality of experience in a communications network |
US10412182B2 (en) * | 2014-12-18 | 2019-09-10 | Nippon Telegraph And Telephone Corporation | Communication speed estimation apparatus, communication speed estimation method, and program |
WO2016172473A1 (en) * | 2015-04-24 | 2016-10-27 | Vid Scale, Inc. | Detecting man-in-the-middle attacks in adaptive streaming |
EP3145201A1 (en) * | 2015-09-17 | 2017-03-22 | Harmonic Inc. | Video processing with dynamic resolution changes |
GB2556925B (en) * | 2016-11-25 | 2020-07-29 | Canon Kk | Method and system for determining encoding parameters of video sources in large scale video surveillance systems |
-
2015
- 2015-12-24 US US15/548,108 patent/US10681416B2/en active Active
- 2015-12-24 EP EP15881232.1A patent/EP3255894B1/en active Active
- 2015-12-24 JP JP2016573208A patent/JP6572245B2/ja active Active
- 2015-12-24 WO PCT/JP2015/086087 patent/WO2016125415A1/ja active Application Filing
- 2015-12-24 CN CN201580075264.6A patent/CN107211178B/zh active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2012070373A (ja) * | 2010-09-13 | 2012-04-05 | Ntt Docomo Inc | ビデオストリームを転送するための方法及び装置 |
US20130091248A1 (en) * | 2011-10-05 | 2013-04-11 | Alcatel-Lucent Bell N.V. | Method and apparatus for improving adaptive streaming video quality by optimizing resource allocation |
WO2013123467A1 (en) * | 2012-02-17 | 2013-08-22 | Vid Scale, Inc. | Hierarchical traffic differentiation to handle congestion and/or manage user quality of experience |
US20130275557A1 (en) * | 2012-04-12 | 2013-10-17 | Seawell Networks Inc. | Methods and systems for real-time transmuxing of streaming media content |
WO2014022017A1 (en) * | 2012-08-03 | 2014-02-06 | Intel Corporation | Methods for quality-aware adaptive streaming over hypertext transfer protocol |
Non-Patent Citations (1)
Title |
---|
See also references of EP3255894A4 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2019016961A (ja) * | 2017-07-07 | 2019-01-31 | 日本電信電話株式会社 | 動画品質制御装置、ビットレート選択方法、及びプログラム |
WO2021124387A1 (ja) * | 2019-12-16 | 2021-06-24 | 日本電信電話株式会社 | 符号化データ生成方法、符号化データ生成装置、及びプログラム |
JPWO2021124387A1 (ja) * | 2019-12-16 | 2021-06-24 | ||
JP7280535B2 (ja) | 2019-12-16 | 2023-05-24 | 日本電信電話株式会社 | 符号化データ生成方法、符号化データ生成装置、及びプログラム |
WO2022185534A1 (ja) * | 2021-03-05 | 2022-09-09 | 日本電信電話株式会社 | 符号化パラメータ導出装置、符号化パラメータ導出方法及びプログラム |
WO2022190263A1 (ja) * | 2021-03-10 | 2022-09-15 | 日本電信電話株式会社 | 情報処理装置、分析方法およびプログラム |
WO2022190264A1 (ja) * | 2021-03-10 | 2022-09-15 | 日本電信電話株式会社 | 情報処理装置、分析方法およびプログラム |
WO2022201225A1 (ja) * | 2021-03-22 | 2022-09-29 | 日本電信電話株式会社 | 制御装置、制御方法及びプログラム |
WO2022201569A1 (ja) * | 2021-03-22 | 2022-09-29 | 日本電信電話株式会社 | 制御装置、制御方法及びプログラム |
JP7448091B2 (ja) | 2021-03-22 | 2024-03-12 | 日本電信電話株式会社 | 制御装置、制御方法及びプログラム |
Also Published As
Publication number | Publication date |
---|---|
JP6572245B2 (ja) | 2019-09-04 |
JPWO2016125415A1 (ja) | 2017-12-21 |
CN107211178B (zh) | 2020-10-09 |
EP3255894B1 (en) | 2021-01-20 |
US20180027293A1 (en) | 2018-01-25 |
US10681416B2 (en) | 2020-06-09 |
CN107211178A (zh) | 2017-09-26 |
EP3255894A4 (en) | 2018-09-26 |
EP3255894A1 (en) | 2017-12-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP6572245B2 (ja) | 体感品質最適化システム、体感品質最適化装置、レコメンド要求装置、体感品質最適化方法、レコメンド要求方法及びプログラム | |
EP3542537B1 (en) | Leveraging aggregated network statistics for enhancing quality and user experience for live video streaming from mobile devices | |
CN110268717A (zh) | 使用播放统计信息对多表示编码的比特率优化 | |
US20140149348A1 (en) | Application program management method and apparatus using context information | |
WO2017094847A1 (ja) | 体感品質最適化装置、ネットワーク品質推定装置、モバイル端末、体感品質最適化方法、ネットワーク品質推定方法、及びプログラム | |
JP2011170640A (ja) | キャッシュサーバ制御装置、コンテンツ配信システム、コンテンツ配信方法、及びプログラム | |
Liubogoshchev et al. | Adaptive cloud-based extended reality: Modeling and optimization | |
CN112752117B (zh) | 一种视频缓存方法、装置、设备及存储介质 | |
KR20220104200A (ko) | 사용자 경험을 향상시키기 위해 에지 컴퓨팅 시스템을 사용한 비디오 콘텐트 스트리밍 비트레이트 선택 | |
JP2011172021A (ja) | キャッシュサーバ制御装置、コンテンツ配信システム、コンテンツ配信方法、及びプログラム | |
US20170201630A1 (en) | System and method for time shifting cellular data transfers | |
Huang et al. | Personalized QoE enhancement for adaptive video streaming: A digital twin-assisted scheme | |
JP6823170B2 (ja) | コンテンツストリームにおける中断期間の予測 | |
JP5390369B2 (ja) | 映像品質推定装置および方法、符号化ビットレート推定装置および方法、並びにプログラム | |
CN111278039B (zh) | 用户感知压抑识别方法、装置、设备及介质 | |
US10491903B2 (en) | Delivery rate selection device, delivery rate selection method, and program | |
EP3788768B1 (en) | Methods and systems for streaming media data over a content delivery network | |
JP6466870B2 (ja) | クライアント装置および方法 | |
WO2020095767A1 (ja) | 配信設計支援方法、配信設計支援装置及びプログラム | |
JP7052768B2 (ja) | 配信設計支援方法、配信設計支援装置及びプログラム | |
Phong et al. | Joint preloading and bitrate adaptation for short video streaming | |
US11985363B2 (en) | Automatically determining parameter values | |
JP2018097393A (ja) | エンゲージメント指標値制御装置、エンゲージメント指標値制御方法、及びプログラム | |
US20240163511A1 (en) | Method and apparatus for receiving adaptive media streaming and adaptive media streaming system | |
WO2017145781A1 (ja) | ペーシング制御装置、ペーシング制御方法、及びプログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15881232 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2016573208 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15548108 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
REEP | Request for entry into the european phase |
Ref document number: 2015881232 Country of ref document: EP |