The content of the invention
The purpose of the present invention is that the shortcoming for overcoming above-mentioned prior art can realize monitoring OTT in real time there is provided one kind
Business is distributed to the operation conditions of the links such as user terminal sight from film source production, Broadcast Control, CDN content, quickly found and fixed
Position problem, General Promotion Consumer's Experience, with wider application OTT quality of services are realized based on user behavior analysis
The system and method for guarantee.
To achieve these goals, it is of the invention based on user behavior analysis realize OTT quality of services ensure system and
Method has following composition:
This realizes the system that OTT quality of services are ensured based on user behavior analysis, and it is mainly characterized by, described system bag
Include:
Film source quality-monitoring module, in product platform side to reaching the standard grade before piece source file detect;
Content delivery network node monitoring modular, in each content delivery network node by pseudo-terminal to platform
Electric program menu and program carry out testing;
Terminal monitoring module, to gather user using the rating parameter in OTT business procedures and according to described
Rating parameter calculates Consumer's Experience video quality score;
Fault locating analysis module, when described terminal monitoring module alarm, to be supervised according to described film source quality
The monitoring state for surveying module and content delivery network node monitoring modular is positioned to failure.
It is preferred that described system also includes:
Expertise library module, to according to described film source quality-monitoring module, terminal monitoring module and fault location
Analysis module warning message analysis fault type and user viewing experience information and when fault severity level exceed system
Operation maintenance personnel is notified during preset value.
More preferably, described expertise library module includes:
Expertise rule parsing device, to according to described film source quality-monitoring module, terminal monitoring module and failure
The warning message analysis fault type of positioning analysis module;
User's perception analysis device, to according to described film source quality-monitoring module, terminal monitoring module and fault location
The warning message of analysis module analyzes the viewing experience information of user;
Failure severity evaluator, to be analyzed failure severity and worked as fault severity level more than systemic presupposition
Operation maintenance personnel is notified during value.
The invention further relates to a kind of side for realizing the guarantee of OTT quality of services based on user behavior analysis by described system
Method, it is characterised in that described method comprises the following steps:
(1)Piece source file of the described film source quality-monitoring module before product platform side is to reaching the standard grade is detected;
(2)Described content delivery network node monitoring modular passes through pseudo-terminal pair in each content delivery network node
Platform clectronic program guide and program carry out testing;
(3)Rating parameter of the described terminal monitoring module collection user in using OTT business procedures and according to described
Rating parameter calculate Consumer's Experience video quality score;
(4)When described terminal monitoring module alarm, described fault locating analysis module is according to described film source matter
The monitoring state of amount monitoring modular and content delivery network node monitoring modular is positioned to failure.
It is preferred that the rating parameter described in described basis calculates Consumer's Experience video quality score, it is specially:
The ratio and program that duration and playing duration are buffered during being played according to program initial buffer duration, program delay again
Rush the frequency and calculate Consumer's Experience video quality score.
It is preferred that described fault locating analysis module is according to described film source quality-monitoring module and content distributing network
The monitoring state of node monitors module is positioned to failure, is comprised the following steps:
(41)Whether the described set top box of described fault locating analysis module detection normally runs, if it is, continuing
Step(42), otherwise determine abort situation in described set top box;
(42)Described fault locating analysis module judges that the alarm moment is front and rear in preset time range, described content
Whether the monitoring state of distribution network node monitors module is faulty, if it is, continuing step(43), otherwise continue step
(44);
(43)Described fault locating analysis module is according to the monitoring shape of described content delivery network node monitoring modular
State and Consumer's Experience video quality score carry out regional comprehensive analysis to described content delivery network node service ability;
(44)Described fault locating analysis module judges whether Broadcast Control platform normally runs, if it is not, then continuing step
(45);
(45)The interactive information that described fault locating analysis module is asked according to customer service analyzes the event of playing platform
Hinder link.
It is preferred that described system also includes expertise library module, described step(3)With(4)Between, also including with
Lower step:
(31)Described expertise library module is according to described film source quality-monitoring module, terminal monitoring module and failure
Positioning analysis module warning message analysis fault type and user viewing experience information and when fault severity level exceedes
Operation maintenance personnel is notified during systemic presupposition value.
More preferably, described expert knowledge library is tight including expertise rule parsing device, user's perception analysis device and failure
Severe evaluator, described expertise library module is according to described film source quality-monitoring module, terminal monitoring module and failure
Positioning analysis module warning message analysis fault type and user viewing experience information and when fault severity level exceedes
Operation maintenance personnel is notified during systemic presupposition value, is comprised the following steps:
(311)Described expertise rule parsing device receives warning message and described expert is judged by cluster analysis
It whether there is the described corresponding fault type of warning message in knowledge base, if it is, continuing step(312), otherwise, continue
Step(313);
(312)Described expertise rule parsing device knowledge based storehouse rule of similarity is carried out described in fuzzy matching determination
The corresponding fault type of warning message, then proceedes to step(314);
(313)Described expertise rule parsing device creates new fault type according to described warning message, then
Continue step(314);
(314)Described user's perception analysis device tidies up the parameter evaluation failure pair according to described fault type and user
The influence that user causes;
(315)Described failure severity evaluator failure severity is analyzed and work as fault severity level exceed be
Operation maintenance personnel is notified during system preset value.
Employ in the invention and the system and method that OTT quality of services are ensured is realized based on user behavior analysis, have
Following beneficial effect:
OTT quality of services based on user behavior analysis ensure that invention helps OTT business operation maintenance personnel to monitor OTT industry in real time
Business is distributed to the operation conditions of the links such as user terminal sight from film source production, Broadcast Control, CDN content, quickly finds and positions
Problem, General Promotion Consumer's Experience ensures the long-run development of OTT core business, to realize that enterprise operation target provides strong branch
Support means, with wider application.
Its advantage is mainly reflected in the following aspects:
1st, active O&M, finds and handling failure prior to user, lifts Consumer's Experience;
2nd, Real-time Alarm, helps operation maintenance personnel quick response emphasis failure;
3rd, the service ability to CDN node is effectively assessed, and is easy to entire content to push the planning of network;
4th, terminal monitoring reflects the actual experience of OTT user;
5th, full process and full network, realizes integrated management;Accurate intelligence, realizes that operation is excellent.
Embodiment
In order to more clearly describe the technology contents of the present invention, carried out with reference to specific embodiment further
Description.
The present invention is expanded the business scenario under Lossless transport, according to OTT based on the testing standard for damaging transmission
The transmission feature of video traffic, based on user behavior, by the soft probe technique of terminal, on the master of real influence OTT Consumer's Experiences
Problem is wanted, such as buffering latencies, wait number of times, video code rate reduction carry out comprehensive analysis, evaluated.
The present invention sets up a set of OTT video service qualities guarantee system, including OTT quality of services(QoS, Quality
Of Service, service quality)And Consumer's Experience(QoE, Quality Of Experience, Quality of experience)E measurement technology, prison
Index system, OTT method for diagnosing faults opinion, OTT testing tools and end-to-end QoS monitoring system etc. are surveyed, soft probe is utilized
Technology is acquired to OTT terminal user's viewing-datas, analyzed, using testing technology to CDN(Content
DistributionNetwork, content distributing network)Node is monitored, so as to help user directly effectively to understand terminal
The problem of rating quality of user, quick positioning effects Consumer's Experience, so as to improve the business satisfaction of terminal user.
The present invention sets up a set of OTT video service qualities guarantee system, including OTT quality of services(QoS)And user's body
Test(QoE)E measurement technology, monitored data analysis, OTT method for diagnosing faults opinion, OTT testing tools and end-to-end QoS prison
Examining system etc., including following several key contents.
First, OTT quality of services measurement model is as shown in Figure 2.
Analyzed by Internet packet capturing, and combine it is user watched during the index such as buffering situation, comprehensively counted according to model
Calculate MOS(Mean Opinion Score, mean opinion score)Score, embodies the viewing experience of terminal user.MOS scores are as follows
Table:
2nd, OTT quality of services measurement basic index
One)Basic index
1st, QoS Testing index:
-- program buffers duration first;
-- the buffering ratio in program buffering number of times, the program buffering frequency, program playing process;
--EPG(Electronic Program Guide, electric program menu)Request number of times, the frequency of failure;
-- EPG request responses average delay, maximum delay;
-- program ID;
-- program downloads time started and duration;
-- program total duration;
-- program download progress;
-- the CDN node of request;
-- program speed of download(Average, maximum, minimum);
-- request program fragment total degree and the frequency of failure;
-- program code check, code check switching;
2nd, network index:
-- the time of TCP foundation connections, TCP re-transmissions, TCP are out of order;
3rd, user behavior index:
-- user's open and close machine time;
-- user's online hours;
-- user watches programme information(Program code, request time, watch duration, coding parameter, code check etc.)
4th, terminal device operating index:
-- CPU usage;
-- memory usage;
-- disk space utilization rate;
Two)Consumer's Experience video quality score
Consumer's Experience video quality score(MOS, Mean Opinion Score, mean opinion score)It is reflection human eye master
See the MOS scorings for evaluating measurement.
For narrowly, MOS is calculated, main and following three relating to parameters in OTT video monitorings:
1)Program initial buffer duration
2)Program buffers duration and the ratio of playing duration during playing
3)Program buffers the frequency again
Broad sense is got on very well, and in order to correctly simulate the situation of program buffering, we are spliced according to packet is caught, and are put into
In local buffering, operation to buffering area when simulation normal video is played, by the model of Producer-consumer problem draw it is above-mentioned in terms of
Calculate MOS three parameters.
Carry out packet capturing, flow-mergence, data splicing, it is simulated cushioned during need following a few class parameters:
1)Network index class:Including average bandwidth, maximum bandwidth, packet loss number of times, number of retransmissions etc., network delay
2)Resource Properties class:Resource name, code check including request, code check rise and fall number of times etc.
3)User behavior class:Time shift, F.F. including user, pausing operation.
Therefore MOS calculates main according to above parameter, is drawn by certain data model and algorithm.
3rd, CDN(Content Delivery Network, i.e. content distributing network)Monitoring
Periodicity active testing is carried out to CDN node by video analysis equipment.Video analysis equipment can pass through Broadcast Control
The interface that platform is provided obtains the Token of certification.Contents for Monitoring includes EPG services and the order program service that POP points are provided.
Testing strategy, testing task can configure.
4th, intelligent expert storehouse
The warning message of terminal/CDN/ Broadcast Control platforms is mainly collected life by expertise library module according to knowledge base
Go out user into failure and intellectual analysis truly to perceive, not only draw failure cause, phenomenon of the failure, and decide whether according to severity
Notify attendant etc..
It is preliminary to report in the warning message input expertise rule parsing device that OTT terminals/CDN/ Broadcast Control platforms are reported
Alert information is carried out after cluster analysis, fuzzy matching is carried out using knowledge base rule of similarity, so as to obtain affiliated fault type of alarming.
By the fault type and the incoming user's perception analysis device of warning message, the analyzer is by failure and alarm
Video parameter input video information analyser, thus obtain the alarm failure to the user watched experience sense caused by.
User's perception analysis device experiences failure and user intelligence in the evaluation mechanism of the incoming intelligent trouble severity of result
The influence that the failure is caused to user can be evaluated, and decides whether to notify operation maintenance personnel.
Operation maintenance personnel confirms after failure and menace level notice is received and handles the failure.
5th, Efficient fault location algorithm
Failure effect analysis (FEA) is built upon on the basis of warning information, resource information and customer information synthesis, and failure is entered
Row temporal correlation, resource dependencies, the analysis of client's correlation.
Fault location strategy is main to carry out accident analysis positioning from 3 levels:OTT end sides, CDN side, Broadcast Control is put down
Platform side.
The monitoring threshold parameter that fault location is set according to operating personnel, when OTT terminal monitoring report and alarms, first
Detection set-top-box system situation, such as CPU, internal memory etc., if there is no problem, then in inquiry alarm t surrounding time section, terminal company
The CDN node and alarm, the performance data of Broadcast Control platform connect, further determines whether it is that CDN node or Broadcast Control platform are caused
The problem of.
If CDN node problem, then in the association analysis period, the receipts of the related OTT terminal users of the node are accessed
Optionally, with to the regional comprehensive analysis of CDN node service ability progress.
If the problem of Broadcast Control platform, the interactive information that the customer service recorded by Broadcast Control platform is asked enters one
Step analysis is which link in Broadcast Control platform goes wrong.
Fault location result can intuitively embody on the service link of OTT terminal users, help operation maintenance personnel quickly to send out
Existing root sex chromosome mosaicism.
Video flowing QoE parameter model
It is foundation objective indicator for the quality evaluation of video flowing, and assigns different weights, with reference to the true of people's observation
The obtained QoE indexs of experience.Video flowing QoE parameter model covers all levels of OSI seven layer models, as shown in Figure 3.
Under OTT video service quality guarantee systems, the monitoring of OTT business is mainly considered end from source to terminal to
End monitoring, schematic diagram such as Fig. 5.
First, film source quality-monitoring
Build film source skill in product platform side and examine system, to reaching the standard grade before piece source file detect.
2nd, CDN node is monitored
Platform side guarantee video file of reaching the standard grade meets content standard requirement, and the service ability of content distributing network is used terminal
The viewing experience at family also plays key effect.Video analysis equipment is disposed in CDN node, by pseudo-terminal to platform EPG
Testing is carried out with program, the service availability and service promptness of CDN node is reflected.
3rd, terminal monitoring
Gathered by terminal monitoring module and report relevant parameter, reflect user during using OTT video traffics
Viewing experience.Terminal monitoring emphasis is directed to Consumer's Experience and quality of service.
4th, fault locating analysis
The Real-time Alarm reported by terminal monitoring module, can be by inquiring about the peaceful stage piece source of same period CDN testings
The result of detection, helps user to carry out fault location.
5th, statistical analysis
The data reported for film source, CDN platforms, terminal monitoring module, system carries out comprehensive statistical analysis.
As shown in figure 1, the present invention's realizes that the system that OTT quality of services are ensured includes based on user behavior analysis:
Film source quality-monitoring module, in product platform side to reaching the standard grade before piece source file detect;
Content delivery network node monitoring modular, in each content delivery network node by pseudo-terminal to platform
Electric program menu and program carry out testing;
Terminal monitoring module, to gather user using the rating parameter in OTT business procedures and according to described
Rating parameter calculates Consumer's Experience video quality score;
Fault locating analysis module, when described terminal monitoring module alarm, to be supervised according to described film source quality
The monitoring state for surveying module and content delivery network node monitoring modular is positioned to failure;
Expertise library module, to according to described film source quality-monitoring module, terminal monitoring module and fault location
Analysis module warning message analysis fault type and user viewing experience information and when fault severity level exceed system
Operation maintenance personnel is notified during preset value;
Described expertise library module includes:
Expertise rule parsing device, to according to described film source quality-monitoring module, terminal monitoring module and failure
The warning message analysis fault type of positioning analysis module;
User's perception analysis device, to according to described film source quality-monitoring module, terminal monitoring module and fault location
The warning message of analysis module analyzes the viewing experience information of user;
Failure severity evaluator, to be analyzed failure severity and worked as fault severity level more than systemic presupposition
Operation maintenance personnel is notified during value.
The present invention's realizes that the method that OTT quality of services are ensured comprises the following steps based on user behavior analysis:
(1)Piece source file of the described film source quality-monitoring module before product platform side is to reaching the standard grade is detected;
(2)Described content delivery network node monitoring modular passes through pseudo-terminal pair in each content delivery network node
Platform clectronic program guide and program carry out testing;
(3)Rating parameter of the described terminal monitoring module collection user in using OTT business procedures and according to program
Initial buffer duration, program buffer duration during playing and the ratio and program of playing duration buffer frequency calculating user's body again
Test video quality score;
(31)Described expertise library module is according to described film source quality-monitoring module, terminal monitoring module and failure
Positioning analysis module warning message analysis fault type and user viewing experience information and when fault severity level exceedes
Operation maintenance personnel is notified during systemic presupposition value;
(4)When described terminal monitoring module alarm, described fault locating analysis module is according to described film source matter
The monitoring state of amount monitoring modular and content delivery network node monitoring modular is positioned to failure;Specifically include following five
Sub-step:
(41)Whether the described set top box of described fault locating analysis module detection normally runs, if it is, continuing
Step(42), otherwise determine abort situation in described set top box;
(42)Described fault locating analysis module judges that the alarm moment is front and rear in preset time range, described content
Whether the monitoring state of distribution network node monitors module is faulty, if it is, continuing step(43), otherwise continue step
(44);
(43)Described fault locating analysis module is according to the monitoring shape of described content delivery network node monitoring modular
State and Consumer's Experience video quality score carry out regional comprehensive analysis to described content delivery network node service ability;
(44)Described fault locating analysis module judges whether Broadcast Control platform normally runs, if it is not, then continuing step
(45);
(45)The interactive information that described fault locating analysis module is asked according to customer service analyzes the event of playing platform
Hinder link.
As shown in figure 4, described expertise library module is according to described film source quality-monitoring module, terminal monitoring module
With the warning message analysis fault type of fault locating analysis module and the viewing experience information of user and when the serious journey of failure
Degree notifies operation maintenance personnel when exceeding systemic presupposition value, comprises the following steps:
(311)Described expertise rule parsing device receives warning message and described expert is judged by cluster analysis
It whether there is the described corresponding fault type of warning message in knowledge base, if it is, continuing step(312), otherwise, continue
Step(313);
(312)Described expertise rule parsing device knowledge based storehouse rule of similarity is carried out described in fuzzy matching determination
The corresponding fault type of warning message, then proceedes to step(314);
(313)Described expertise rule parsing device creates new fault type according to described warning message, then
Continue step(314);
(314)Described user's perception analysis device tidies up the parameter evaluation failure pair according to described fault type and user
The influence that user causes;
(315)Described failure severity evaluator failure severity is analyzed and work as fault severity level exceed be
Operation maintenance personnel is notified during system preset value.
Employ in the invention and the system and method that OTT quality of services are ensured is realized based on user behavior analysis, have
Following beneficial effect:
OTT quality of services based on user behavior analysis ensure that invention helps OTT business operation maintenance personnel to monitor OTT industry in real time
Business is distributed to the operation conditions of the links such as user terminal sight from film source production, Broadcast Control, CDN content, quickly finds and positions
Problem, General Promotion Consumer's Experience ensures the long-run development of OTT core business, to realize that enterprise operation target provides strong branch
Support means, with wider application.
Its advantage is mainly reflected in the following aspects:
1st, active O&M, finds and handling failure prior to user, lifts Consumer's Experience;
2nd, Real-time Alarm, helps operation maintenance personnel quick response emphasis failure;
3rd, the service ability to CDN node is effectively assessed, and is easy to entire content to push the planning of network;
4th, terminal monitoring reflects the actual experience of OTT user;
5th, full process and full network, realizes integrated management;Accurate intelligence, realizes that operation is excellent.
In this description, the present invention is described with reference to its specific embodiment.But it is clear that can still make
Various modifications and alterations are without departing from the spirit and scope of the present invention.Therefore, specification and drawings are considered as illustrative
And it is nonrestrictive.