TWI223508B - Method for objective playout quality measurement of a packet based network transmission - Google Patents

Abstract

A method for objective playout quality measurement of a packet based network transmission includes determining a normalized total delay of a packet for playout and assigning a corresponding delay mean opinion score (DMOS) to the packet, then further determining a normalized packet loss rate of the packet and assigning a corresponding loss mean opinion score (LMOS) to the packet. The method includes averaging the DMOS and the LMOS to determine a mean opinion score (MMOS) of the packet and outputting the MMOS of the packet to a display device. The DMOS and LMOS can be assigned referencing continuous modeling equations or can be assigned referencing discrete value lookup tables. The method is performed by a processor.

Description

1223508

TECHNICAL FIELD The present invention relates to a method for determining media playback of a packet transmitted by a media network in a playback mechanism, and in particular, a method for broadcasting a target quality in a mass. Prior art

The development of wireless communication technology, and the development of wireless communication technology have made possible real-time digital communication. Recently, new technologies for communication via the Internet have been developed maturely, such as the Internet Voice Protocol (Voice 0ver Internet Protocol, Vo IP) and other instant interactive communication systems. In network transmission packet communication, the biggest obstacle is the difference in network delay, which is the so-called delay disturbance (ji 11 er). Delay perturbation can be greatly reduced by delaying the playback of packets based on a playback delay. When the network delay is not fixed, reducing the number of delay disturbances of a transmission requires a reasonable measurement of the network delay and a correct judgment of the playback delay. However, the playback delay cannot be too long, because the purpose of the transmission is to achieve the effect of instant transmission, and an excessively long playback delay is contrary to this project. Therefore, for two-way communication such as Internet voice protocols, voice calls, and online games, it is important to minimize playback delays to avoid user inconvenience during use.

Page 6! 223508

Figure_ = Unintended packet information of a voice data 2 〇 is transmitted across a network 0. The data 20 includes audible sections 20a, 20c, and 20e, and 2 includes: a distinguishable audio message; and the data 20 also includes a stationary section 20 晌 ', and / or an unidentifiable audio message. A transmitter 1 2 can be 2 2 = A computer or other device. Packets P1 ~ P15 are sent at regular intervals, but because of network delay, the transmission of packet-P1 5 is delayed, and one of them reaches a receiver 14 Some of the packets must be delayed with a further amount of time to form a cohesive sound source. The receiver 14 can be a similar personal computer or device. / ^ 22 ^ Contains audible sections 22a, 22c and 22e and stationary section ΟΛ 'These sections correspond to section I in the material 20 to be transmitted ^ ^ P1 is transmitted by the transmitter at a given time 12 send out. The packet P1 2 2 Γ is delayed by the network 10 for any reason. The shadow in Figure 1 shows this delay and other further delay factors. The packet p 1 will be received by the receiver 1 4 The delay makes the packet p1 play continuously with the packet P2 that has been network 2 ^. If the packet P1 is not delayed by the receiver i by 7 steps, the packets P1 and P2 cannot be played continuously, and it will appear I heard interrupted. This hearing interruption f in the material 22 is heard by the user at the receiver 14, resulting in poor sound quality in the playback material 22. The packets P 2-P 5 are all delayed by the network 10 at the same time.

1223508

In addition, the packets P2-P5 are not necessarily delayed, and the receivers 14 and 14 receive further delays t. However, the packet P7 is broadcast before the packet timing, and the receiver 14 must delay the packet arrival. Until the packet P6 is received, the still segment L P7 of the material 22 is played. This delay is additively affected. Due to network delays and delays, \ so that the audible section 22c will not arrive at the same time as the packets P10 and P11. Seal ^ Quice, packet? 8 and P9 will be closed and jealous, while the female hit A seal P9 and P11 will be delayed because of this. However, there will be no further-information 22 delay. Packets, P13 and P14, and packets P6 and p7— 搂, receive / <You Zhisheng Packet -7 P1? N Pi Rail ^ τ ^ λ, 7 samples are subject to similar sorting confusion. The packets P12 and P15 reach the receiver 14 in the normal order. The above description with reference to FIG. 1 is a simplified process. The packet pi-Pb is assumed to depend on the integer product of its packet length to give the delay to the receiver. In fact, in a given transmission process, when there are network delays and delay disturbances, many packet numbers are delayed. Figure 1 does not show that the entire received data 2 is a delay of three segments. These three segments include a network delay and an additional playback delay added by the receiver 14. If the receiver 14 does not attach the additional delay, some packets may be out of order and other packets may not be played at all. Conventional techniques have shown many ways how to estimate the time to be delayed by the receiver. A basic and valid but controversial estimate of playback latency

Page 8 1223508 5. The method of invention description (4) is the mean delay and variance (MDV) method, which is written by R. Ramjee, j. Kurose, D. Towsley, and H. Schu1zrinne. "Adaptive Playout Mechanisms for Packetized Audio Applications in Wide-area Networks" is described, and is hereby proposed as a prior art of the present invention. The average delay difference method is further detailed in "Design and Experimental Evaluation of an Adaptive Delay / control Mechanism for Packetized Audio for use over the" by Marco Roccetti, Vittorio Ghini, Giovanni Pau, Paola Salomoni, and Maria Elena Bonfig1 i In the "Internet", it is also the prior art of the present invention. In short, the average delay difference method is to estimate the playback delay by a difference between the average network delay and a smoothing coefficient. This simple dynamic This method provides important key improvements over other non-dynamic methods. Other methods for estimating playback delay are described in the real-time transport protocol. Η ·

Schu 1 z r i nne ^ S. Casner, R. Frederick and V

&Quot; RTP: A Transport Protocol f0r by Jacobson

The standard specifications of the instant transmission communication protocol are described in detail in Rea Time Appl i cat ions · ', which is also the prior art of the present invention. The instant transmission protocol method used to estimate the delay is a ^ average delay difference number method with a fixed smoothing factor. Although it is simpler than the average delay difference method, but

Page 9 1223508

Time Transmission Protocol can only provide a rough estimate of network latency. Once the playback delay has been estimated and calculated, the receiver 14 can play packets transmitted over the network. The playback quality perceived by the user is related to the selected playback method practice. Therefore, in order to determine which method has the best playback quality, it is necessary to develop a quality measurement system. The following methods for determining the playback quality are also the prior art of the present invention: ITU-T P.861 (1996), MObjective quality measurement of telephone-band (300-3400Hz) speech codecs ", 02. 1998, ETSI (European Telecommunication Standard Institute ) EG 201 377-1 VI.1.1, '' Specification and measurement of speech transmission quality &quot;, 02 · 2001 and ITU-T G.8 62, fl Perceptual evaluation of speech qua 1 i ty (PESQ), an objective method for end-to-end speech quality assessment of narrow band telephone network and speech codecs &quot;, 02. 2001. These methods of sound compression and decompression are not applicable to two-way communication, and cannot effectively measure echo and stable interference. Furthermore, the previous method for measuring the playback quality cannot be applied to comparing different playback mechanisms. In addition, the prior technology also contains many patents on the target playback quality measurement, but whether it is measuring signals, sound or Video

1223508 V. Description of the invention (6) The qualitative method applications are only applicable to static data or one-way communication. SUMMARY OF THE INVENTION Therefore, the main object of the present invention is to provide a method for determining a target playback quality, so as to solve problems that cannot be solved by the prior art. ❿ In short, the present invention includes the use of a processor to determine the normalized total delay of a network transmission packet and specify a corresponding subjective score for the packet delay. The present invention further determines a packet loss rate and specifies a corresponding subjective score for packet loss. Finally, the present invention averages the delayed subjective score and the depleted subjective score to obtain an average subjective score and outputs it to a display device. Look at the Lord. Loss table Consumption and check score evaluation Appearance away from the main or late extension, Fang Limo special construction, please continue Shen Lian's examination, participation can be issued This rule is based on the fact that the main evaluation is equal. The evaluation and release of the system and the machine are broadcasted to the base of each machine. A step package is included in the count. The method can be duplicated. for. The amount of standard points is good. The range of the fan is requested. The quality of the product is broadcasted. The envelope can be divided and evaluated. The main equal flat is enclosed as the point. 03 ^ You are required to declare the scope of the other range.

1

Page 1223508 5. Explanation of the invention (7) Comparison, evaluation and rating of different playback mechanisms. For implementation, please set 30k computer communication device instructions to the receiving network to place the packet to control the predetermined player 3 4 in the playback device. The brain device interferes with the reference picture receiving system, but is not placed. In this way, the sender 4 2 block mechanism in the path is omitted. In this way, the packet Figure 2 is a similar figure to the general structure diagram. In the present situation, it can be replaced with a message device, which can be broadcasted to the Internet. The wireless device 3 also has a transmission and settlement device 30. The receiver 30 includes a receiver buffer information of the terminal packet information. Play delay delays the packet. The statistical calculation of the playback delay of the audio and video output device is usually a speaker and a display of communication. The front and rear devices 30 can be played smoothly. The receiver receives the call mechanism, connects to the receiver to buffer the playback delay, and closes the communication device of the processor 14 or other but here to the network 32. Connectors 3, 4 and controllers then broadcast packet delays or other reductions in network communication equipment. Similar to simplifying 10 for receivers 32 for one broadcast 4 2 for buffering. Media Personal circuit delay The playback mechanism of the playback controller 42 can be the aforementioned average delay difference number method or an instant transmission protocol. In the preferred embodiment of the present invention, the 'playback mechanism' is software executed by the communication device 30, but it may also be hardware or software simulating hardware. For example, the playback controller 4 2 can calculate the playback delay of a received network transmission packet, which is represented by the arrow 70 in Figure 2. According to the average delay difference method, it is as follows:

Page 12 1223508 V. Description of the invention (8)

Di = K-Ri-1 &gt; &lt; si-Si-l) | MD ± = F x MD ^ + (1-F) x D ± v ± = | md ±-D ± | MV ± = F x W / ±, L + (1-F) x V ± PD ± = MV ± x SF Formula (1) where D is the network delay; R is a receiver time function; S is a transmitter time function; i points to a The index of the current packet; i-1 points to the index of a previous packet; MD is the average network delay; F is a smoothing factor; V is a network delay difference; MV is an average network delay difference; PD Is a playback delay; SF is a permutation coefficient. In this example, the playback controller 42 first calculates the network delay of the current packet by comparing the time function of the current packet and the previous packet, then calculates the average network delay difference based on the average network delay and the smoothing function, and finally uses An arrangement factor determines the playback delay of the current packet. The time function is provided by the receiver 32 to the playback controller 42. Smoothing factor

Page 13 1223508 V. Description of the invention (9) ------ and = column coefficients are set directly inside the playback controller 42 and can be adjusted. The playback controller 42 passes the playback delay to the playback buffer 34. ^ The average delay difference method described above, each packet transmitted over the network is continuous and real-time. In addition, the network transmission can be a video call, online gaming, or other instant interactive communications. As mentioned earlier, playback quality is closely related to the playback mechanism. For example, for the user of the communication device 30, the average delay difference method described by the equation (1) used is the key to the media output of the output device 36 as it is perceived. If the network communication is a voice communication, the listener will perceive a pause or pause when the network communication delay difference is too large (that is, there is a delay disturbance). If the network communication is an image or a game output, the viewer will perceive the screen to jump or the screen to be poor. These situations experienced by the listener or viewer detract from the overall playback quality. Referring to FIG. 2, a quality analyzer 44 is connected to the playback controller 42 to determine a target playback quality measurement value. The quality analyzer 44 includes a processor 46 for measuring the quality according to the parameter 52, a display device 48 for outputting the measured quality value to the user, and an interface 50 for the user to adjust the parameter 52. In a preferred embodiment of the present invention, the processor 46 is a computer processor and the display device 48 is a computer monitor. To facilitate the present invention, the quality analyzer 44 is connected to the playback controller 42 so that the playback controller 42 contains all the information required in the method of the preferred embodiment of the present invention. Or the playback controller 4 2 itself can directly perform playback quality

Page 14 1223508 V. Measurement of invention description (ίο). The processor 46 of the quality analyzer 44 determines the subjective delay score of the current packet of a received network communication, as shown below:

DF TD, NTD ± = -1——

MAX _ TD DMOS ± = MAX MOS x (1 · NTDif11 formula (2) where NTD is a normalized total delay; TD is a total delay; DF is a delay coefficient; MAX_TD is a maximum total delay;

DM0S — delayed subjective score; MAX_M0S is a maximum subjective score; DM is a delayed subjective score coefficient.

The normalized total delay of the current packet is calculated with reference to the current total packet delay and the maximum total delay of all transmitted packets. The total delay includes a compression and decompression delay (ie, the time required for the communication device 30 to compress and decompress the packet data), a network transmission delay, and a packet playback delay. Then, the standardized total delay and the maximum subjective score are used to determine the delay subjective score of the current packet. The maximum subjective score depends on compression and decompression, and under ideal conditions can be determined empirically by measuring the target playback quality of multiple playback mechanisms. The index of the delay coefficient and the delay subjective score are also obtained through experience, and users can further adjust it.

Page 15 1223508 V. Description of the invention (11) In the preferred embodiment of the present invention, the processor 46 determines the subjective delay score of the current packet according to the calculation process of equation (2), and further determines the current packet as shown below. Loss subjective score: NPLR ± LMOS, PLR,

LF

_ PLR MOS X (1-NPLRJ

LH formula (3) Among them, NPLR is a standardized packet loss rate PLR is a packet loss rate; LF is a loss factor; MAX_PLR is a maximum packet loss rate: LM0S is a loss subjective score; LM is a loss subjective score coefficient . The loss rate is calculated before the calculation of the loss rate. The loss rate is the loss rate. The loss rate is the most expensive. The most successful package is the package. Preview before. Bao Wufeng, the person who lost knowledge of the newsletter, refers to the fact that the major sports skills of the middle and upper arts are at a loss, and they are sealed when it takes time. In order to get all the points you need, you can evaluate and judge the subject. The standardization loss of the seal owner is based on several steps of standardization and assessment. The winner decides the main loss of the sub-assessment and the sub-assessment of the sub-assessment. The C package is delayed. The front view is that once the sub-assessment is performed, the sub-assessment is performed.

Page 16 1223508 V. Description of the invention (12) The formula (4) of the sub-evaluation subjective average 1 is the value of the delayed subjective score, the attributable subjective score, and the average subjective score. The equations are given by equations (2), (3) And (4) can be obtained, and can also be classified according to compression and decompression quality (excellent-very poor G. 711 compression and decompression system-excellent quality loss delay 0 · 0 5 60 · 00 good 0 · 10 120 • 00 ordinary 0 · 15 240 • 00 Poor 0 · 25 480 • 00 Extreme Poor 0 · 35 720 • 00 Table G · 711 Subjective 'rating subjective rating 4.50 3. 50 3. 00 2. 00 1.0 0 Delay subjective rating 4. 50 4 · 00 3. 50 2. 50 2. 00 The list data in Table 1 can be regarded as the aforementioned equation: parameters, which can be applied to integrated circuits or digital applications (d tone. The phase loss rate and total delay of the packet are used for loss Queries on the subjective scoring of the extension.

Page 17

1223508 V. Description of the invention (13) Whether using continuous or discontinuous modeling, the average subjective score is output by the display device 48 of the quality analyzer 44 and used as a target playback quality measurement value. Or, under the circumstances, the delayed subjective score and the depleted subjective score can be used as the target playback quality measurement respectively. The parameters 52 that the user can adjust using the interface 50 include the delay coefficient (DF), the delay subjective rating coefficient (DM), the loss coefficient (LF), the wearable subjective rating coefficient (LM), and the maximum subjective rating (MAX). -M0S). The user can then use the average subjective score to set up network playback or determine an optimal playback mechanism. Referring to the flowchart of FIG. 3, the present invention can determine an optimal playback mechanism from a plurality of playback mechanisms. For a scheduled playback mechanism, the method is applied to a group of packets in a network transmission. For a balanced comparison of a playback mechanism, the distance between the network status and the network transmission in each playback mechanism must remain fixed. The steps of the flowchart in Figure 3 are as follows: Step 100: Start; β Step 102: Start a quality measurement playback mechanism or system, start network transmission and playback at the receiving pass-through device 30, ^ Step 104 : Determine the current packet transmission delay and loss rate of the network transmission according to the video playback mechanism; Step 106: Determine the current packet delay according to equations (2) and (3, evening outside μ7 days left% taxi / Table 1) = = (E.g. ^ Step 108: Use Equation (4) to calculate the average subjective score with a delayed subjective score and flat knife, Γί value / 7 ΐ, or refer to the 主 δ ten table (such as Table 1) Eight· -

Page 18

1223508

Step 11 0: Determine whether it is a network transmission — go to step 11 2; if not, the next packet will be sent. If it is, then the packet is indexed and the calculation returns to the average value under the playback mechanism. The average step 114 for network round-trip packets is to determine whether it is the last-need to compare 102; then go to step 116; if there is still a need to compare // release mechanism; if it is a 'Di mechanism, return to step 11 6: The playback mechanism is subjective on average, and is displayed through the display device 48; the average value of the knife is output, and step 118: terminate. Where to use the quality analyzer 44 In practice, this is done in the network of Figure 3 above. Users can benefit from evaluating and rating each playback mechanism. The processor 46 performs steps 104 to 114. The process can be used in a physical network or a model to output the average subjective score data. Please refer to Figure 4. Fig. 4 is an example of a graphical output 6 displayed on the display device 8 of the quality analyzer 44. The measurement of the playback quality of most playback mechanisms is performed according to the method of the present invention in the range of delay disturbance (network delay difference number) from 60ms to 720ms, and a bar graph of PMla-PM4a to PMle-PM4e is generated, where pMia-pM4a Is the playback mechanism, and 3 to e indicate the range of the delay perturbation. The height of the bar indicates the average subjective rating from 0 · 0 0 to 5 · 0 0. The value of the delay disturbance range in Figure 4

Page 19 1223508 V. Explanation of the invention (15) and average subjective score are only examples. Users can compare the average subjective score of each playback mechanism within a range of delay disturbances to determine the most suitable playback mechanism for practical applications. Compared with the prior art, the method of the present invention provides a target quality measure for a packet-based network transmission. The playback quality (or average subjective score) is determined by a processor using standardized total packet delay, standardized packet loss rate, compression and decompression information, and adjustable parameters. Finally, the user can obtain the graphic output of the average subjective score displayed by the display device.

The above description is only a preferred embodiment of the present invention, and any equivalent changes and modifications made in accordance with the scope of the patent application of the present invention shall fall within the scope of the patent of the present invention.

Page 20

Claims (1)

-&'; Yagen i-: · * _1] -L p ——— 1— 1 Case No. 92105703 Amendment 6 、 Scope of Patent Application! 1. A method for measuring a network in a network using a processor Objective of road transmission: A method of objective playout quality. The network transmission includes packet information for playing by using a playback mechanism. The method includes the following steps: Determine the total delay of a packet of a packet. The total delay depends on the transmission time of the packet at two nodes in the network; the maximum of one of the total packet delays of a plurality of packets is used to normalize the total packet delay; A delay mean opinion score (DMOS) is assigned to the packet according to the normalized total packet delay; the packet loss rate of the packet is determined, and the loss rate refers to the packet being transmitted between two nodes. Extra time of the predetermined time; using one of the packet loss rates of the plurality of packets to normalize the packet loss rate; assigning a loss subjective score I (loss mean opinion score, LMOS) to based on the standardized packet loss rate to The packet; Use the delayed subjective score and the attrition subjective score to determine a mean mean opinion score (MMOS) of a packet of the packet; and output the average subjective score of the packet to a display device. 2. The method according to item 1 of the patent application scope, wherein the standardized packet total delay, the standardized packet loss rate, the delay subjective score, and the loss subjective score are based on a continuous modeling equation (c ο n t i n u 〇 u s Page 23 Amendment b 丨 Case No. 92105703 rij VI. Decision and shooting of patent application scope modeling equations 3. If the method of patent application scope item 1 is used, 1 will be improved 仏 ± ^ Total delay, the standardized packet loss rate / The two subjective scores of the two non-dagger seals are based on the discrete values (^^ delayed subjective scores and lookup tables) and the designated timetable (diSCrete vahe 4). ^ ^ Refers to the delayed subjective score and the average subjective score in the attrition subject. The leader 砚 one of the arithmetic mean ❿ If the method of the fourth member of the patent application, the method also includes the average subjective one of a plurality of packets. The average value is used to determine the average time value of one of the playback mechanisms in the average subjective Ping 82 = day-to-day cycle. The flat knife is used in the Internet transfer wheel. The time i 'of the average subjective rating of the player's playback mechanism is based on the broadcasting system. J Ten constructive values for several players Page 24 repair options i 丨 I &amp; tr {: Case No. 92105703 3 a ___________________Kim Jong ________________________________________________________ I 6. Application for Patent Scope j I 8. If the method of the first item of patent application is applied, the network transmission Contains sound, video calls, online games, and other instant interactive communications transmitted in accordance with the Voice over Internet Protocol (Vo IP). 9. The method of claim 1 in which the network includes a computer network or a wireless transmission network of a wireless telephone. 1 0. A method for using a processor in a network to determine the playback quality of a network transmission target, wherein the network transmission includes packet information for playing using a playback mechanism, and the method includes the following steps: decision The total delay of a packet, the total delay includes the transmission time of the packet at one of two nodes in the network; calculate a standardized total delay, which is the ratio of the total packet delay to a maximum total delay Display, and use a delay coefficient as an index; calculate the delay subjective score of a packet, the packet delay subjective score is 1 minus the normalized total delay, and a delay subjective score factor is used as an index, and then multiplied by a maximum subjective score; And output the packet delay subjective score to a display device. 1 1. The method according to item 10 of the scope of patent application, wherein the delay coefficient and the delay subjective scoring coefficient are constants obtained experimentally. -1, 1223 m if- X Λ-ί 93. β * — 本 ———____________ 1 7 WWJ- ^ Λ: W * urw--T ΊΓ1 m- i-jmtm VI. Application for patent coverage 92105703 ^ J________________ Day ___________________________________ Amendment! 12. The method of item 10 in the scope of patent application, wherein the maximum subjective rating i is a maximum value set according to the playback mechanism. ί I 1 3. —A method for using a processor in a network to determine the playback quality of a network transmission target. The network transmission includes packet information for playing using a playback mechanism. The method includes the following steps: Determine the packet loss rate of a packet. The packet loss rate includes the extra time when the packet is transmitted between two nodes that exceeds a predetermined time. Calculate a standardized packet loss rate. The standardized packet loss rate is based on the packet loss rate and one. The ratio of the maximum packet loss rate is expressed, and a loss coefficient is used as an index; a packet loss subjective score is calculated. The packet loss subjective score is 1 minus the normalized loss rate, and a loss subjective score coefficient is used as an index, then multiplied A previous maximum subjective score; and outputting the packet loss subjective score to a display device. 14. The method according to item 13 of the scope of patent application, wherein the attrition coefficient and attrition subjective scoring coefficient are constants determined by experiments. 15. The method according to item 13 of the scope of patent application, wherein the maximum subjective score is a maximum value set according to the playback mechanism.