JP2009033684A - Device and method for evaluating voice - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To reduce errors between a calculated voice evaluation value and a voice evaluation value, by a subjective evaluation when providing a voice-evaluating device having a buffer section and a thinning processing section. <P>SOLUTION: The voice evaluation method includes a step S301 for calculating a first evaluation value that is a voice evaluation value, based on a voice packet prior to thinning processing; a step S302 for calculating a second evaluation value that is a voice evaluation value, based on a voice packet after thinning processing; a step S304 for calculating the difference in a sequence number between the voice packets after thinning processing; steps S307, S308 for calculating the ratio of the difference with respect to a prescribed interval; and a step S309 for outputting the first evaluation value as a third value, when the ratio is not more than the threshold, and for bringing the third evaluation value closer to the second value as the ratio increases when the ratio exceeds the threshold. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a voice evaluation apparatus and a voice evaluation method for evaluating the quality of a voice call.

  2. Description of the Related Art Conventionally, VoIP (Voice over Internet Protocol), which is a technique for transmitting packets including encoded voice data (hereinafter, “voice packets”) via an IP (Internet Protocol) network such as the Internet, is known. .

  In addition, a communication terminal capable of performing a voice call using VoIP is provided with a buffer unit (jitter buffer) that absorbs jitter, that is, fluctuations in the reception interval of voice packets.

  The buffer unit accumulates the voice packets received by the packet receiving unit, and outputs the accumulated voice packets at predetermined intervals. The audio packet output from the buffer unit is decoded (reproduced) as audio from the speaker after being decoded.

  Further, in a voice call using VoIP, a voice evaluation device (for example, an RTP (Real Time Protocol) tester) that evaluates the quality of a voice call is used (see Patent Document 1).

Such a voice evaluation device measures voice packet transmission delay, jitter, packet loss, and the like based on the voice packet received by the packet receiver, and shows voice evaluation indicating the quality of the voice call according to the measurement result. Calculate the value. As the speech evaluation value, an R value obtained by a calculation formula called E-model is often used.
Japanese Patent Laying-Open No. 2004-222257 (FIG. 1)

  By the way, in a communication terminal capable of performing a voice call using VoIP, when the packet receiving unit receives a number of voice packets exceeding the capacity of the buffer unit, the voice packet stored in the buffer unit is thinned out at regular intervals. A processing unit may be provided.

  Furthermore, the interval between the audio packets after the thinning is reduced, so that continuous audio packets are accumulated in the buffer unit. By such thinning-out processing, free space is secured in the buffer unit.

  Since the voice packets stored in the buffer are thinned out and the intervals between the voice packets are reduced, when the voice packets after the thinning are reproduced, the person can hear the reproduced sound as the voice reproduced at the double speed. That is, even if the voice packets stored in the buffer unit are thinned out, if the number of thinned voice packets is small, that is, if the playback speed is equal to or lower than a predetermined value, the evaluation result by subjective evaluation (voice evaluation by a person) is It does not decline.

  Moreover, since it is preferable that a voice evaluation apparatus performs voice evaluation on the same conditions as a communication terminal, it is possible to provide a buffer part and a thinning-out process part in a voice evaluation apparatus. However, in the voice evaluation apparatus, the thinned voice packet is regarded as a packet loss. That is, when the voice packet stored in the buffer unit is thinned out, the voice evaluation value calculated by the voice evaluation device is greatly reduced.

  Therefore, when the buffer unit and the thinning processing unit are provided in the voice evaluation device, there is a problem that an error between the voice evaluation value calculated by the voice evaluation device and the voice evaluation value by subjective evaluation becomes large.

  Therefore, the present invention has been made to solve the above-described problem, and in the case where the buffer unit and the thinning processing unit are provided in the voice evaluation device, the calculated voice evaluation value, the voice evaluation value by subjective evaluation, An object of the present invention is to provide a voice evaluation device and a voice evaluation method capable of reducing the error.

  In order to solve the problems described above, the present invention has the following features. First, the first feature of the present invention is that a packet receiving unit (RTP test packet receiving unit 156a) for receiving a voice packet including a sequence number (SN), and storing the voice packet received by the packet receiving unit, A buffer unit (jitter buffer unit 159a) for outputting the stored voice packets at a predetermined interval (packet interval Tsub), and a voice evaluation value indicating the quality of the voice call based on the voice packets stored by the buffer unit An audio evaluation apparatus including an evaluation value calculation unit (statistics calculation unit 156b and evaluation value calculation unit 158) for calculating (R value), wherein the audio packets accumulated by the buffer unit are thinned out at regular intervals. The processing unit (buffer control unit 159b) and the voice packet after the thinning by the thinning processing unit A ratio calculation unit (reproduction speed calculation unit 159c) that calculates a difference of sequence numbers (sequence number difference Ssub) and calculates a ratio of the difference (reproduction speed Vspd) with respect to the predetermined interval, and the evaluation value calculation unit Is a first calculation unit (statistics calculation unit 156b and evaluation value calculation unit 158) that calculates a first evaluation value (evaluation value Md) that is the voice evaluation value based on the voice packet before thinning by the thinning processing unit. And a second calculation unit (statistical value calculation unit 156b and evaluation value calculation unit 158) that calculates a second evaluation value (evaluation value Ma) that is the voice evaluation value based on the voice packet after the thinning by the thinning processing unit. A third calculation unit (statistic calculation unit 156) that calculates a third evaluation value (evaluation value M) that is the voice evaluation value based on the first evaluation value, the second evaluation value, and the ratio. And the evaluation value calculation unit 158), the first evaluation value is output when the ratio is less than or equal to a threshold value, and the third evaluation value is output when the ratio exceeds the threshold value. .

  According to such a feature, the voice evaluation device outputs a first evaluation value that is a voice evaluation value based on a voice packet before thinning by the thinning processing unit when the ratio (that is, the reproduction speed) is equal to or less than a threshold value. To do. In addition, when the ratio exceeds the threshold, the voice evaluation device outputs a third evaluation value that is a voice evaluation value based on the first evaluation value, the second evaluation value, and the ratio. Therefore, an error between the voice evaluation value calculated by the voice evaluation device and the voice evaluation value by subjective evaluation can be reduced.

  A second feature of the present invention relates to the first feature of the present invention, wherein the third calculation unit subtracts (Md−Ma) the second evaluation value from the first evaluation value (evaluation value). A calculation unit 158), a division unit (evaluation value calculation unit 158) that divides the subtraction result by the subtraction unit by the ratio ((Md−Ma) / α (Vspd)), the second evaluation value, and the division It is summarized that an addition unit (evaluation value calculation unit 158) that adds the division result by the unit and uses the addition result (Ma + (Md−Ma) / α (Vspd)) as the third evaluation value is provided.

  A third feature of the present invention relates to the first feature of the present invention, and is related to transmission delays and packets in packet transmission paths (packet transmission paths R1, R2, R3) to and from a packet transmission device (IP telephone terminal 4). A transmission characteristic measurement unit (transmission characteristic measurement unit 15a) that measures transmission characteristic information including loss information in a first measurement period (a prescribed period of 60 seconds), and a packet transmission unit (RTP test packet transmission) that transmits the voice packet. Unit 155c) and a voice packet processing unit (emulator unit 157) that delays or discards the voice packet transmitted by the packet transmission unit based on the transmission characteristic information measured by the transmission characteristic information output unit. The packet receiving unit receives the voice packet output from the voice packet processing unit, and the transmission characteristic measuring unit Measured by the short period measurement unit (transmission characteristic information generation unit 152d) that measures the transmission characteristic information in a second measurement period (an evaluation target period of 1 second) shorter than the first measurement period, and the short period measurement unit A gist of the invention is that it includes a transmission characteristic information output unit (transmission characteristic information generation unit 152d) that repeatedly outputs the transmission characteristic information to the evaluation value calculation unit until the first measurement period expires.

  According to a fourth aspect of the present invention, there is provided a packet receiving unit that receives a voice packet including a sequence number, a buffer that stores the voice packet received by the packet receiving unit, and outputs the stored voice packet at a predetermined interval. Is a voice evaluation method using a part, wherein the step of thinning out the voice packets accumulated by the buffer part at a predetermined interval (step S302) and the difference of the sequence numbers between the voice packets after the thinning are calculated. A step (step S304), a step of calculating a ratio of the difference with respect to the predetermined interval (steps S307 and S308), and a step of calculating a first evaluation value which is a voice evaluation value based on the voice packet before thinning ( Step S301) and a voice evaluation value based on the voice packet after thinning A step of calculating an evaluation value (step S302), a step of calculating a third evaluation value that is a voice evaluation value based on the first evaluation value, the second evaluation value, and the ratio (step S309); The present invention includes the steps of outputting the first evaluation value when the ratio is equal to or less than the threshold value, and outputting the third evaluation value when the ratio exceeds the threshold value (steps S306 to S309).

  According to the present invention, there is provided a voice evaluation device and a voice evaluation method capable of reducing an error between a calculated voice evaluation value and a voice evaluation value by subjective evaluation when the buffer unit and the thinning processing unit are provided in the voice evaluation device. Can be provided.

  Next, an embodiment of the present invention will be described with reference to the drawings. In the following, (1) overall schematic configuration of communication system, (2) configuration of wireless terminal, (3) operation of wireless terminal, (4) specific example, (5) operation and effect, (6) other embodiments It explains in order of.

  In the description of the drawings in the following embodiments, the same or similar parts are denoted by the same or similar reference numerals.

(1) Overall Schematic Configuration of Communication System FIG. 1 is an overall schematic configuration diagram of a communication system 10 according to the present embodiment. In the present embodiment, a communication system 10 applied to VoIP will be described. Moreover, this embodiment demonstrates the case where the audio | voice evaluation apparatus and communication terminal which were mentioned above are comprised integrally.

  The communication system 10 includes a plurality of wireless communication systems. Specifically, the communication system 10 is a wireless communication system compliant with cdma 2000 nx evolution-data only (EV-DO), which is a kind of third generation mobile phone system, and a wireless communication system defined by IEEE 802.16e. (Hereinafter referred to as “WIMAX” (registered trademark)) and a wireless LAN system (hereinafter referred to as “W-LAN”) defined by IEEE 802.11 or the like.

  The communication system 10 includes a wireless terminal 1, a wireless base station 31, a wireless base station 32, an access point 33, a communication network 3, and an IP telephone terminal 4. The communication network 3 is an IP network such as the Internet.

  The IP telephone terminal 4 and the wireless terminal 1 perform a voice call using VoIP via the communication network 3.

  Specifically, the IP telephone terminal 4 and the wireless terminal 1 encode voice data and transmit voice packets including the encoded voice data at regular intervals. Further, the IP telephone terminal 4 and the wireless terminal 1 decode voice data from the received voice packet and output voice.

  In this embodiment, RTP is used as a packetization and transfer protocol for voice data. Hereinafter, a packet generated according to RTP is referred to as an RTP packet. Note that, for example, SIP (Session Initiation Protocol) is used as the call control protocol.

  The radio base station 31 is included in a radio communication system compliant with EV-DO. The wireless base station 32 is included in a wireless communication system compliant with WiMAX. The access point 33 is included in a wireless communication system compliant with W-LAN.

  The wireless terminal 1 can execute wireless communication with the wireless base station 31, the wireless base station 32, and the access point 33. That is, the wireless terminal 1 can be connected to a wireless communication system compliant with EV-DO, a wireless communication system compliant with WiMAX, and a W-LAN defined by IEEE 802.11 or the like.

  The wireless terminal 1 performs a handover as the wireless terminal 1 moves. In the example of FIG. 1, the packet transmission path is switched in the order of R1, R2, and R3 when the wireless terminal 1 executes a handover.

  Further, the wireless terminal 1 receives the RTP packet transmitted by the IP telephone terminal 4 and calculates an evaluation value (R value) of voice quality based on the received RTP packet. That is, the wireless terminal 1 includes an RTP tester unit 15b (see FIG. 4) that calculates an evaluation value, and obtains an evaluation value using the RTP tester unit 15b.

  The voice evaluation in the RTP tester unit 15b requires a measurement time (specified period) of at least 60 seconds. However, when performing voice evaluation at the time of handover, it is necessary to perform voice evaluation in an evaluation target period of about 1 second. In the present embodiment, the wireless terminal 1 uses the RTP packet information measured in the evaluation target period of about 1 second, and uses the RTP tester unit 15b that requires a measurement time (specified period) of 60 seconds. Evaluate the voice quality.

  Further, when the wireless terminal 1 determines that the voice quality has deteriorated based on the calculated evaluation value (R value), the wireless terminal 1 switches the voice encoding method to a low-rate encoding method and maintains the voice call. On the other hand, when it is determined that the voice quality is good, the wireless terminal 1 switches the voice coding method to a high-rate coding method to improve the voice quality.

(2) Configuration of Wireless Terminal Next, the configuration of the wireless terminal 1 will be described.

(2.1) Schematic Configuration of Wireless Terminal FIG. 2 is a schematic configuration diagram of the wireless terminal 1. As shown in FIG. 2, the wireless terminal 1 includes antennas 11a to 11c, an EV-DO communication unit 12a, a WiMAX communication unit 12b, a W-LAN communication unit 12c, an operation unit 13, a voice codec unit 14, a control unit 15, and a microphone. 16a, a speaker 16b, a display unit 17, a storage unit 18, and a jitter buffer 19.

  The EV-DO communication unit 12a transmits / receives a radio signal according to EV-DO to / from the radio base station 31 via the antenna 11a. The WiMAX communication unit 12b transmits and receives wireless signals according to WiMAX to and from the wireless base station 32 via the antenna 11b. The W-LAN communication unit 12c transmits and receives a wireless signal according to the W-LAN to and from the access point 33 via the antenna 11c.

  In addition, the EV-DO communication unit 12 a, the WiMAX communication unit 12 b, and the W-LAN communication unit 12 c perform conversion between a radio signal and a baseband signal and transmit / receive the baseband signal to / from the control unit 15.

  The operation unit 13 is configured by a numeric keypad, function keys, and the like, and is an interface used for inputting user operation details.

  The audio codec unit 14 encodes the audio data acquired from the microphone 16 a and decodes the audio data acquired from the control unit 15. The microphone 16a collects sound and inputs sound data to the control unit 15 via the sound codec unit 14 based on the collected sound. The speaker 16 b outputs audio via the audio codec unit 14 based on the audio data acquired from the control unit 15.

  The display unit 17 displays the received image or operation details (input telephone number, address, etc.) via the control unit 15.

  The control unit 15 controls various functions provided in the wireless terminal 1. Further, the control unit 15 executes various protocols such as RTP, IP, and SIP. The storage unit 18 stores various information used for control and the like in the wireless terminal 1.

(2.2) Jitter Buffer The jitter buffer 19 is used for absorbing jitter. Further, the jitter buffer 19 rearranges the received RTP packets so that they are in the order of transmission of the RTP packets, based on the sequence numbers included in the RTP packets.

  Further, the jitter buffer 19 accumulates the received RTP packets and passes the accumulated RTP packets to the voice codec unit 14 at a predetermined interval. The audio codec unit 14 decodes the RTP packet from the jitter buffer 19. Thereby, the RTP packet is reproduced at an appropriate reproduction speed.

  When the wireless terminal 1 receives a large amount of RTP packets, the RTP packets will not fit in the jitter buffer 19. The behavior in the jitter buffer 19 in such a case includes the following types 1) to 3).

  1) Discard RTP packets that do not fit in the jitter buffer.

  2) Discard all the previously held RTP packets and store new RTP packets.

  3) The previously held RTP packet is thinned out and a new RTP packet is accumulated.

  In the above 1) and 2), RTP packets having consecutive sequence numbers are discarded. On the other hand, in 3) above, RTP packets with non-sequential sequence numbers are discarded by thinning. In the present embodiment, the jitter buffer 19 of the above type 3) is used.

  FIG. 3 is an explanatory diagram for explaining the thinning-out operation of RTP packets in the jitter buffer 19.

  In FIG. 3A, RTP packets with sequence numbers 1 to 16 are accumulated in the jitter buffer 19 and there is no free space. Assume that a new RTP packet is received in such a state.

  In FIG. 3B, RTP packets with sequence numbers 4, 8, 12, and 16 are thinned out. As a result, the data amount of the RTP packet stored in the jitter buffer 19 becomes 3/4.

  Here, one RTP packet is thinned out for every four RTP packets, but one RTP packet may be thinned out for every two or three RTP packets.

  In FIG. 3C, intervals between RTP packets after thinning are reduced. As a result, free capacity is secured in the jitter buffer 19 and new RTP packets can be stored.

  When an RTP packet as shown in FIG. 3C is played back, the playback is completed in the original 3/4 time. That is, the user recognizes the sound as being reproduced at double speed. However, just because the playback is performed at double speed does not mean that the user cannot hear the voice.

  On the other hand, a new RTP packet cannot be accommodated even if it is thinned once, and if the thinning is repeated, a plurality of RTP packets having consecutive sequence numbers are discarded, and the user cannot hear as voice. In the present embodiment, the voice evaluation is performed in consideration of such behavior of the jitter buffer 19.

(2.3) Detailed Configuration of Wireless Terminal FIG. 4 is a detailed block diagram of the wireless terminal 1, specifically, a functional block diagram of each function executed by the control unit 15 and the storage unit 18. In the following, points related to the present invention will be mainly described.

  As shown in FIG. 4, the wireless terminal 1 includes a transmission characteristic measurement unit 15a, an RTP tester unit 15b, and an encoding control unit 15c.

  The transmission characteristic measurement unit 15a acquires RTP packets transmitted and received between each wireless communication unit (EV-DO communication unit 12a, WiMAX communication unit 12b, and W-LAN communication unit 12c) and the voice codec unit 14. In addition, the transmission characteristic measurement unit 15a acquires information related to wireless communication (hereinafter, “wireless information”) from the EV-DO communication unit 12a, the WiMAX communication unit 12b, and the W-LAN communication unit 12c.

  The transmission characteristic measurement unit 15a measures transmission characteristic information indicating the characteristics of the packet transmission path R1, R2, or R3 based on the acquired RTP packet and the acquired wireless information. Specifically, the transmission characteristic measurement unit 15a measures transmission delay and packet loss as transmission characteristic information.

  The RTP tester unit 15b calculates an evaluation value (R value) based on the transmission measurement information measured by the transmission characteristic measurement unit 15a. Specifically, the RTP tester unit 15b emulates the packet transmission path R1, R2, or R3 using the transmission measurement information, and enables measurement of transmission characteristic information for a specified period.

  The encoding control unit 15c controls the speech encoding method according to the evaluation value (R value) calculated by the RTP tester unit 15b.

  Hereinafter, a detailed configuration of the transmission characteristic measurement unit 15a and a detailed configuration of the RTP tester unit 15b will be described.

(2.3.1) Detailed Configuration of Transmission Characteristic Measurement Unit As shown in FIG. 4, the transmission characteristic measurement unit 15a includes a wireless information acquisition unit 151a, a packet reception unit 151b, a wireless information storage unit 18a, and a received packet information storage unit. 18b, stable period specifying unit 152a, transmission interval calculating unit 152b, relative delay calculating unit 152c, transmission characteristic information generating unit 152d, evaluation target period selecting unit 153a, wireless communication system specifying unit 153b, reference transmission delay storage unit 18c, time measuring unit 154a and a time extension control unit 154b.

  The wireless information acquisition unit 151a acquires wireless information from the EV-DO communication unit 12a, the WiMAX communication unit 12b, and the W-LAN communication unit 12c. The radio information is, for example, radio signal reception quality (RSSI or SNR) or a base station identifier. The acquired wireless information is stored in the wireless information storage unit 18a. The wireless information includes information indicating the wireless communication system to which the wireless terminal 1 is connected.

  The packet receiving unit 151b receives an RTP packet. The packet receiver 151b stores the reception time and sequence number of the received RTP packet in the received packet information storage unit 18b as received packet information.

  The stable period specifying unit 152a specifies a period during which wireless communication is stable (for example, the reception quality is equal to or higher than a threshold) from the reception quality information stored in the wireless information storage unit 18a.

  The transmission interval calculation unit 152b acquires the reception time of the RTP packet in the period specified by the stable period specification unit 152a from the reception packet information storage unit 18b. The transmission interval calculation unit 152b calculates a transmission interval at which the IP telephone terminal 4 transmits the RTP packet from the acquired reception time.

  The evaluation target period selection unit 153a selects an evaluation target period of about 1 second.

  The relative delay calculation unit 152c calculates a relative delay (transmission delay) in the evaluation target period selected by the evaluation target period selection unit 153a based on the received packet information. Details of the relative delay will be described later.

  Further, the relative delay calculation unit 152c sorts RTP packets in the evaluation target period according to the sequence number from the received packet information, thereby identifying a packet (packet loss) lost in the packet transmission path R1, R2, or R3.

  The wireless communication system specifying unit 153b specifies the wireless communication system to which the wireless terminal 1 is connected during the evaluation target period.

  The transmission characteristic information generating unit 152d acquires a reference transmission delay (system delay) corresponding to the wireless communication system specified by the wireless communication system specifying unit 153b from the reference transmission delay storage unit 18c. The reference transmission delay is a fixed delay time required for construction / decomposition of an RTP packet, and is a delay time specific to each radio communication system.

  The transmission characteristic information generation unit 152d calculates a transmission delay using the relative delay calculated by the relative delay calculation unit 152c and the reference transmission delay acquired from the reference transmission delay storage unit 18c.

  The packet loss information (hereinafter referred to as packet loss information) and the calculated transmission delay information (hereinafter referred to as “transmission delay information”) are stored in the transmission characteristic information generation unit 152d.

  The time extension control unit 154b instructs the transmission characteristic information generation unit 152d to start and stop transmission characteristic information output in accordance with the measurement start instruction and the measurement end instruction transmitted from the RTP tester unit 15b. Further, the time extension control unit 154b instructs the transmission characteristic information generation unit 152d to output the transmission delay information and the packet loss information based on the time information of the time measuring unit 154a and the elapsed time from the start of measurement.

(2.3.2) Detailed Configuration of RTP Tester Unit As shown in FIG. 4, the RTP tester unit 15b includes an emulator unit 157, a control message transmission / reception unit 155a, a transmission control unit 155b, an RTP test packet transmission unit 155c, and a control packet. Transmission / reception unit 155d, RTP test packet reception unit 156a, statistic calculation unit 156b, control packet transmission / reception unit 156c, RTT measurement unit 156d, measurement control unit 156e, control message transmission / reception unit 156f, evaluation value calculation unit 158, jitter buffer unit 159a, A buffer control unit 159b and a playback speed calculation unit 159c are included.

  The emulator unit 157 performs a retention process and a discard process for the RTP test packet and the control packet (ICMP packet) based on the transmission delay information and the packet loss information. The RTP test packet transmitted from the RTP test packet transmitter 155c is received by the RTP test packet receiver 156a after the emulator unit 157 stays in accordance with the transmission delay information. Furthermore, the emulator unit 157 discards the RTP test packet according to the packet loss information.

  The measurement control unit 156e instructs the transmission control unit 155b to start and stop the measurement via the control message transmission / reception unit 156f and the control message transmission / reception unit 155a. The measurement control unit 156e instructs the RTT measurement unit 156d to measure the round trip time (RTT).

  The RTT measurement unit 156d transmits a control packet via the control packet transmission / reception unit 156c. The RTT measurement unit 156d receives the reply from the control packet transmission / reception unit 155d and measures the round trip time.

  The transmission control unit 155b instructs the RTP test packet transmission unit 155c to start transmission in accordance with the measurement start instruction from the measurement control unit 156e.

  The statistic calculation unit 156b performs statistical processing on the RTP test packet and calculates statistical information. The statistical information includes average transmission delay (average value of transmission delay), jitter (transmission delay fluctuation), packet loss rate, and the like.

  The evaluation value calculation unit 158 calculates an evaluation value (R value) in the evaluation target period based on the statistical information calculated by the statistic calculation unit 156b and the round trip time measured by the RTT measurement unit 156d.

  The jitter buffer unit 159a accumulates the RTP test packet received by the RTP test packet receiving unit 156a. The buffer control unit 159b controls the jitter buffer unit 159a. The jitter buffer unit 159a is used in the same manner as the jitter buffer 19 shown in FIG.

  The reproduction speed calculation unit 159c calculates the reading speed (reproduction speed) of the RTP test packet read from the jitter buffer unit 159a. The evaluation value calculation unit 158 calculates the audio evaluation value in consideration of the reproduction speed calculated by the reproduction speed calculation unit 159c.

(2.4) Transmission Delay Measurement Processing Next, transmission delay measurement processing executed by the transmission characteristic measurement unit 15a will be described. FIG. 5 is an explanatory diagram for explaining the transmission delay measurement process.

(2.4.1) Transmission interval calculation process First, the transmission interval calculation process executed by the transmission interval calculation unit 152b will be described.

  The timing is not always the same between the RTP packet transmitting side (IP telephone terminal 4) and the RTP packet receiving side (wireless terminal 1). For example, there is an error of about several [msec] between the RTP packet transmitting side and the RTP packet receiving side.

  For this reason, when the communication state is reproduced (emulated) based on the received RTP packet without correcting the error, a completely different communication state is reproduced as a cumulative error due to long-term reproduction. Will be.

  In order to avoid such a problem, the transmission interval calculation unit 152b uses the calculated transmission interval (Tjtr) instead of the scheduled transmission interval (for example, 20 [ms]).

The transmission interval calculation unit 152b calculates the average reception interval in the period in which the propagation environment is the best (for example, the reception quality value is the highest) among the periods using the same wireless communication system (type) as the evaluation target period, using the formula ( The transmission interval is calculated according to 1).
Tjtr = Mean (Trec (sn)-Trec (sn-1)) (1)

(2.4.2) Transmission Delay Calculation Processing Next, transmission delay calculation processing executed by the relative delay calculation unit 152c and the transmission characteristic information generation unit 152d will be described.

  The relative delay calculation unit 152c calculates a difference value between the reception elapsed time and the transmission elapsed time with reference to a certain RTP packet with respect to the RTP packet received during communication in the same wireless communication system (type). Here, the transmission elapsed time is calculated based on the transmission interval (Tjtr).

  The transmission characteristic information generation unit 152d identifies the RTP packet with the lowest calculated difference value as the reference packet (sn0). The delay (system delay) of the reference packet (sn0) is the shortest delay in the wireless communication system (type).

  Therefore, the delay of the specified RTP packet (reference packet; sn0) is considered to be only a delay (reference transmission delay (Tdlybas (type)) specific to the target wireless communication system (type).

  For example, when the evaluation target period is the period [T1, T2] in FIG. 6, the transmission characteristic information generation unit 152d determines that wireless communication is performed using WiMAX, and the reference transmission delay (Tdlybas (type ): type = WiMAX).

  Then, the transmission characteristic information generation unit 152d calculates a transmission delay Tdly (sn) for the packet (sn) in the evaluation value calculation target section.

Tdly (sn) = (Trec (sn)-Trec (sn0))-Tjtr × (sn-sn0) + Tdlybas (type) (2)
The transmission delay of each RTP packet is calculated by equation (2).

(2.5) Packet Loss Measurement Processing Next, packet loss measurement processing executed by the relative delay calculation unit 152c and the transmission characteristic information generation unit 152d will be described.

  The relative delay calculation unit 152c detects a missing SN (sequence number) in the received RTP packet. The missing SN (sequence number) means that the packet transmission path has been lost.

That is, the relative delay calculation unit 152c sorts the packets in the evaluation value calculation target section in order of the sequence number, and identifies the lost sequence number. The relative delay calculation unit 152c sets the loss flag (Lost (sn)) as follows in order to reproduce the packet loss (packet loss) in the emulator unit 157.
Lost (sn) = TRUE: Packet loss
FALSE: There is a packet

(2.6) Emulation Processing Next, the emulation processing executed by the RTP tester unit 15b will be described with reference to FIG.

  As described above, the evaluation target period (for example, 1 second) is shorter than the specified period (for example, 60 seconds). For this reason, the RTP tester unit 15b emulates (reproduces) the packet transmission path based on the transmission delay (Tdly (sn)) and loss flag (Lost (sn)) of the RTP packet in the evaluation target period. By emulating (reproducing) the packet transmission path, it is possible to perform measurement in a specified period (for example, 60 seconds).

  FIG. 7 is a graph with SN (sequence number) on the horizontal axis and transmission delay on the vertical axis. FIG. 7A shows an actual measurement result in the evaluation target period. FIG. 7B shows the time of evaluation processing for the period [N0, N1] of the sequence number corresponding to the specified period.

  The time extension control unit 154b specifies a sequence number corresponding to the elapsed time from the start of evaluation, and outputs transmission delay information and packet loss information based on the sequence number from the transmission characteristic information generation unit 152d to the emulator unit 157.

  That is, the transmission characteristic information generation unit 152d transmits the evaluation target period (corresponding sequence number) [N0, N1] (not including N1) with respect to the specified period (corresponding sequence number) [S0, Sm]. Delay information and packet loss information are repeatedly output.

  That is, the transmission delay information and packet loss information of the sequence number Sa in the period [S1, S2) (not including S2) include the sequence number Na (= Sa−S1 + N0) corresponding to the period [N0, N1) (not including N1). ) Transmission delay information and packet loss information are used.

  Accordingly, it is possible to obtain an evaluation value in the evaluation target period by extending to the specified period length necessary for measurement based on the short evaluation target period.

(2.7) Evaluation Value Calculation Processing Next, evaluation value calculation processing executed by the RTP tester unit 15b will be described.

  The reproduction speed calculation unit 159c calculates the sequence number difference Ssub of the RTP test packet reproduced (read) from the jitter buffer unit 159a and the reproduction interval (hereinafter referred to as “packet interval”) Tsub of the RTP test packet.

Specifically, the packet interval Tsub is a time difference (T _{N + 1} −T _N ) between times (T _N , T _{N + 1} ) when the RTP test packets N and N + 1 are reproduced, and a packet transmission interval (for example, 20 [ms]). Is a value normalized for comparison with the sequence number difference Ssub.

  The playback speed calculation unit 159c calculates the playback speed Vspd as an average value in the evaluation target period according to the following equation (3).

式（３）及び図８に示すように、パケット間隔Ｔｓｕｂがシーケンス番号差Ｓｓｕｂ以上である場合、スロー再生とみなされ、再生速度Ｖｓｐｄ＝０になる。

As shown in Expression (3) and FIG. 8, when the packet interval Tsub is greater than or equal to the sequence number difference Ssub, it is regarded as slow reproduction, and the reproduction speed Vspd = 0.

  Further, the statistic calculation unit 156b and the evaluation value calculation unit 158 include an evaluation value Md when thinning is not performed in the jitter buffer unit 159a, and an evaluation value Ma when thinning is performed in the jitter buffer unit 159a. Is calculated.

  The evaluation value calculation unit 158 calculates a final evaluation value M according to the following equation (4).

ここで、図９に示すように、α（Ｖｓｐｄ）は、再生速度Ｖｓｐｄの増加に応じて、指数関数的に増加する。

Here, as shown in FIG. 9, α (Vspd) increases exponentially as the reproduction speed Vspd increases.

  As can be seen from the equation (4), when the reproduction speed Vspd is 0 (α (Vspd) = 1), the evaluation value M becomes the evaluation value Md. When the reproduction speed Vspd is 1, the evaluation value M is a value between the evaluation values Md and Ma. When the reproduction speed Vspd becomes a large value, the evaluation value M approaches the evaluation value Ma.

(3) Operation of Wireless Terminal Next, the operation of the wireless terminal 1 will be described.

(3.1) Outline of Evaluation Value Calculation Procedure FIG. 10 is a flowchart showing an outline of the evaluation value calculation procedure executed by the wireless terminal 1.

  In step S101, the stable period specifying unit 152a specifies a good (stable) period of the wireless state from the wireless information stored in the wireless information storage unit 18a.

  In step S102, the transmission interval calculation unit 152b acquires the received packet information of the RTP packet received in the period specified by the stable period specifying unit 152a.

  In step S103, the transmission interval calculation unit 152b calculates the transmission interval Tjtr according to Equation (1).

  In step S104, the evaluation target period selection unit 153a selects an evaluation target period.

  In step S105, the wireless communication system specifying unit 153b specifies the wireless communication system from the wireless information.

  In step S106, the transmission characteristic information generation unit 152d acquires a reference transmission delay (Tdlybas (type)) corresponding to the wireless communication system (type).

  In step S107, the transmission characteristic information generation unit 152d calculates a transmission delay based on an arbitrary packet in a period in which the wireless communication system (type) is used, and identifies a packet having the lowest value of the transmission delay (sn0). .

  In step S108, the transmission characteristic information generation unit 152d calculates a transmission delay according to Expression (2).

  In step S109, the transmission characteristic information generation unit 152d sorts the RTP packets in the evaluation value calculation target section in order of sequence number, and identifies the sequence number of the missing RTP packet.

  In steps S110 to S112, the RTP tester unit 15b executes an emulation process (virtual measurement operation) in a specified period. Details of the emulation processing will be described later.

  In step S113, the evaluation value calculation unit 158 calculates an evaluation value using the measurement result of the emulation process.

(3.2) Emulation Procedure FIG. 11 is a flowchart showing the emulation procedure executed by the wireless terminal 1, specifically, the details of step S111 in FIG.

  In step S201, the emulator unit 157 receives the RTP test packet transmitted from the RTP test packet transmission unit 155c.

  In step S202, the measurement control unit 156e updates (Sa ++) the number of received packets Sa.

  In step S203, the measurement control unit 156e calculates which RTP packet (Na) in the evaluation target period corresponds to the RTP test packet received by the emulator unit 157. Specifically, the measurement control unit 156e calculates the remainder from (Sa−S0 + 1) to (N1−N0 + 1), and adds N0 too.

  In step S204, the measurement control unit 156e reads the transmission delay information and the packet loss information of the Na packet and applies them to the RTP test packet received by the emulator unit 157.

(3.3) Details of Evaluation Value Calculation Procedure Next, details of the evaluation value calculation procedure executed by the wireless terminal 1 will be described. FIG. 12 is a flowchart showing the evaluation value calculation procedure executed by the wireless terminal 1, specifically, the details of step S113 in FIG.

  In step S301, the statistic calculation unit 156b and the evaluation value calculation unit 158 calculate the evaluation value Md without decimation by the buffer control unit 159b.

  In step S302, the statistic calculation unit 156b and the evaluation value calculation unit 158 calculate the evaluation value Ma with decimation by the buffer control unit 159b.

  In step S303, the buffer control unit 159b reads the Nth packet from the jitter buffer unit 159a. Here, the sequence number of the RTP test packet read from the jitter buffer unit 159a is SN (N), and the read time is T (N).

  In step S304, the playback speed calculation unit 159c calculates the sequence number difference Ssub.

  In step S305, the playback speed calculation unit 159c calculates the packet interval Tsub.

  In step S306, the playback speed calculation unit 159c determines whether the sequence number difference Ssub calculated in step S304 is larger than the packet interval Tsub calculated in step S305.

  If the sequence number difference Ssub is larger than the packet interval Tsub, the process proceeds to step S307. In step S307, the playback speed calculation unit 159c calculates the playback speed Vspd according to Expression (3).

  On the other hand, if the sequence number difference Ssub is equal to or smaller than the packet interval Tsub, the process proceeds to step S308. In step S308, the playback speed calculation unit 159c sets the playback speed Vspd to 0 according to Expression (4).

  In step S309, the evaluation value calculation unit 158 calculates an equation based on the evaluation value Md calculated in step S301, the evaluation value Ma calculated in step S302, and the reproduction speed Vspd calculated in step S307 or step S308. The final evaluation value M is calculated according to (4).

(4) Specific Example Next, the effect obtained by the present embodiment will be described with a specific example. In the following, a MOS value that is subjective speech quality is used as the speech quality evaluation value. In the MOS value, the human actually evaluates the voice quality in the following five stages 1 to 5.

  5: I can easily understand the content of the story. It requires no effort and can be heard in a relaxed state.

  4: Can be understood with little effort. If you are careful, no effort is required.

  3: Requires moderate effort.

  2: It takes considerable effort.

  1: No connection, call disconnection, silence. Even if you try hard, you can't hear anything.

  In the period when the discard due to thinning occurred, although the voice can be heard, the conventional method had an evaluation value of about 1 (very bad (no effort at all)) An evaluation value was calculated for the location using the above-described evaluation value calculation procedure. As a result, an evaluation value of about 2.0 to 2.5 (bad (requires considerable effort)) was calculated. Therefore, it was confirmed that an evaluation value close to subjective evaluation can be calculated.

(5) Operation / Effect According to the present embodiment, the wireless terminal 1 stores the RTP test packet receiver 156a that receives the RTP test packet and the RTP test packet received by the RTP test packet receiver 156a. A jitter buffer unit 159a that outputs an RTP test packet at an interval Tsub, and an evaluation value calculation unit 158 that calculates a voice evaluation value (R value) based on the RTP test packet accumulated by the jitter buffer unit 159a.

  Further, the wireless terminal 1 calculates the sequence number difference Ssub between the buffer control unit 159b that thins out the RTP test packets accumulated by the jitter buffer unit 159a at regular intervals, and the RTP test packet that is thinned out by the buffer control unit 159b, A playback speed calculation unit 159c that calculates a ratio (playback speed Vspd) of the sequence number difference Ssub to the interval Tsub.

  The evaluation value calculation unit 158 calculates an evaluation value Md based on the RTP test packet before thinning by the buffer control unit 159b and an evaluation value Ma based on the RTP test packet after thinning by the buffer control unit 159b. Further, the evaluation value calculation unit 158 calculates a final evaluation value M based on the evaluation value Md, the evaluation value Ma, and the reproduction speed Vspd.

  The evaluation value calculation unit 158 outputs the evaluation value Md as the evaluation value M when the reproduction speed Vspd is equal to or lower than the standard reproduction speed TH. Further, when the reproduction speed Vspd exceeds the standard reproduction speed TH, the evaluation value calculation unit 158 brings the evaluation value M closer to the evaluation value Ma as the reproduction speed Vspd increases.

  Therefore, an error between the voice evaluation value M calculated by the wireless terminal 1 and the voice evaluation value based on the subjective evaluation can be reduced.

  According to this embodiment, the transmission characteristic measurement unit 15a expires the transmission characteristic information of the voice packet received in the evaluation target period (1 second) shorter than the predetermined period (60 seconds). Repeat until output. The RTP tester unit 15b calculates an evaluation value (R value) based on the transmission characteristic information output by the transmission characteristic measurement unit 15a.

  Therefore, it is possible to provide a voice evaluation device that realizes voice quality evaluation in a short period even when a voice quality evaluation value is calculated based on transmission characteristic information measured in a specified period.

  This makes it possible to evaluate the evaluation target period in the wireless communication environment. Therefore, in the wireless terminal 1, the speech encoding method can be switched between the low rate and the high rate in response to the decrease in the speech evaluation value. Furthermore, voice evaluation at the time of handover between different wireless communication systems such as EVDO, WiMAX, and W-LAN becomes possible.

  According to the present embodiment, the transmission characteristic measurement unit 15a measures the transmission delay according to the equation (2). That is, the transmission delay is measured without using a time stamp or RTT. For this reason, even when the wireless terminal 1 and the IP telephone terminal 4 are asynchronous or when the transmission delay is different between the upstream packet transmission path and the downstream packet transmission path, the transmission delay of the received RTP packet is reduced. It becomes possible to measure.

(6) Other Embodiments As described above, the present invention has been described according to the embodiment. However, it should not be understood that the description and drawings constituting a part of this disclosure limit the present invention. From this disclosure, various alternative embodiments, examples and operational techniques will be apparent to those skilled in the art.

  In the embodiment described above, the wireless communication system specifying unit 153b specifies the wireless communication system in the evaluation target period. However, the wireless communication system specifying unit 153b may specify the wireless base station or access point that was connected in the evaluation target period, according to the base station identifier. In this case, a reference transmission delay may be set for each radio base station or access point.

  In the embodiment described above, the communication system 10 includes a plurality of wireless communication systems, but there may be only one. In the above-described embodiment, RTP is used as a packetization and transfer protocol for voice data. However, other protocols may be used.

  In the embodiment described above, the case where the evaluation target period is about 1 second and the specified period is at least about 60 seconds has been described. However, the present invention is not limited thereto, and the evaluation target period may be a period shorter than the specified period. .

  In the above-described embodiment, the configuration in which voice evaluation is performed in the wireless terminal 1 has been described. Furthermore, in the case where only voice evaluation is performed, the above-described transmission characteristic measurement unit 15a and RTP tester unit 15b may be provided outside the communication terminal (wireless terminal 1).

  In the RTP tester unit 15b, an actual network may be used instead of the emulator unit 157. In this case, the control message transmission / reception unit 155a, the transmission control unit 155b, the RTP test packet transmission unit 155c, and the control packet transmission / reception unit 155d are arranged on the packet transmission side. The RTP test packet receiving unit 156a receives the RTP test packet via the network.

  Thus, it should be understood that the present invention includes various embodiments and the like not described herein. Therefore, the present invention is limited only by the invention specifying matters in the scope of claims reasonable from this disclosure.

1 is an overall schematic configuration diagram of a communication system according to an embodiment of the present invention. It is a schematic block diagram of the radio | wireless terminal which concerns on embodiment of this invention. It is explanatory drawing explaining operation | movement of the jitter buffer provided in the radio | wireless terminal which concerns on embodiment of this invention. It is a functional block diagram of each function performed by the control part and memory | storage part of the radio | wireless terminal which concerns on embodiment of this invention. It is explanatory drawing for demonstrating the transmission delay measurement process performed with the radio | wireless information which concerns on embodiment of this invention. It is a schematic diagram which shows an example of the radio | wireless information which concerns on embodiment of this invention. It is explanatory drawing for demonstrating the emulation process performed by the transmission characteristic measurement part and RTP tester part which concern on embodiment of this invention. It is a figure which shows the relationship between the packet interval in the evaluation value calculation process which concerns on embodiment of this invention, and a sequence number difference. It is explanatory drawing for demonstrating the reproduction speed in the evaluation value calculation process which concerns on embodiment of this invention. It is a flowchart which shows the outline | summary of the evaluation value calculation procedure performed by the radio | wireless terminal which concerns on embodiment of this invention. It is a flowchart which shows the emulation procedure performed by the radio | wireless terminal which concerns on embodiment of this invention. It is a flowchart which shows the detail of the evaluation value calculation procedure performed by the radio | wireless terminal which concerns on embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Wireless terminal, 3 ... Communication network, 4 ... IP telephone terminal, 10 ... Communication system, 11a-11c ... Antenna, 12a ... EV-DO communication part, 12b ... WiMAX communication part, 12c ... LAN communication part, 13 ... Operation 14, audio codec unit, 15 control unit, 15 a transmission characteristic measurement unit, 15 b RTP tester unit, 15 c encoding control unit, 16 a microphone, 16 b speaker, 17 display unit, 18 storage unit , 18a ... wireless information storage unit, 18b ... received packet information storage unit, 18c ... reference transmission delay storage unit, 19 ... jitter buffer, 31 ... wireless base station, 32 ... wireless base station, 33 ... access point, 151a ... wireless information Acquiring unit 151b ... Packet receiving unit 152a ... Stable period specifying unit 152b ... Transmission interval calculating unit 152c ... Relative delay calculating unit 152d ... Transmission Characteristic information generation unit, 153a ... evaluation target period selection unit, 153b ... wireless communication system identification unit, 154a ... timing unit, 154b ... time extension control unit, 155a ... control message transmission / reception unit, 155b ... transmission control unit, 155c ... RTP test Packet transmission unit, 155d ... control packet transmission / reception unit, 156a ... RTP test packet reception unit, 156b ... statistic calculation unit, 156c ... control packet transmission / reception unit, 156d ... RTT measurement unit, 156e ... measurement control unit, 156f ... control message transmission / reception Unit, 157... Emulator unit, 158... Evaluation value calculation unit, 159a... Jitter buffer unit, 159b... Buffer control unit, 159c.

Claims (4)

A packet receiver for receiving a voice packet including a sequence number;
A buffer unit that accumulates the voice packets received by the packet receiver and outputs the accumulated voice packets at a predetermined interval;
A voice evaluation device comprising: an evaluation value calculation unit that calculates a voice evaluation value indicating the quality of a voice call based on the voice packet accumulated by the buffer unit;
A thinning processing unit that thins out the voice packets accumulated by the buffer unit at regular intervals;
Calculating a difference between the sequence numbers between the voice packets after thinning by the thinning processing unit, and a ratio calculating unit calculating a ratio of the difference with respect to the predetermined interval;
The evaluation value calculation unit
A first calculation unit that calculates a first evaluation value that is the voice evaluation value based on the voice packet before thinning by the thinning processing unit;
A second calculation unit that calculates a second evaluation value that is the voice evaluation value based on the voice packet after the thinning by the thinning processing unit;
A third calculator that calculates a third evaluation value that is the voice evaluation value based on the first evaluation value, the second evaluation value, and the ratio;
The voice evaluation device that outputs the first evaluation value when the ratio is equal to or less than a threshold value, and outputs the third evaluation value when the ratio exceeds the threshold value. The third calculator is
A subtractor for subtracting the second evaluation value from the first evaluation value;
A division unit for dividing the subtraction result by the subtraction unit by the ratio;
The speech evaluation apparatus according to claim 1, further comprising: an addition unit that adds the second evaluation value and a division result by the division unit and uses the addition result as the third evaluation value. A transmission characteristic measurement unit for measuring transmission characteristic information including information on transmission delay and packet loss in a packet transmission path between the packet transmission apparatus and the packet transmission apparatus in a first measurement period;
A packet transmitter for transmitting the voice packet;
A voice packet processing unit that delays or discards the voice packet transmitted by the packet transmission unit based on the transmission characteristic information measured by the transmission characteristic information output unit;
The packet receiving unit receives the voice packet output from the voice packet processing unit,
The transmission characteristic measuring unit is
A short period measurement unit that measures the transmission characteristic information in a second measurement period shorter than the first measurement period;
The voice evaluation apparatus according to claim 1, further comprising: a transmission characteristic information output unit that repeatedly outputs the transmission characteristic information measured by the short period measurement unit to the evaluation value calculation unit until the first measurement period expires. . A packet receiver for receiving a voice packet including a sequence number;
A voice evaluation method using the buffer unit that accumulates the voice packet received by the packet receiver and outputs the accumulated voice packet at a predetermined interval,
Thinning out the voice packets stored by the buffer unit at regular intervals;
Calculating a difference of the sequence numbers between the voice packets after the thinning;
Calculating a ratio of the difference with respect to the predetermined interval;
Calculating a first evaluation value that is a voice evaluation value based on the voice packet before thinning;
Calculating a second evaluation value which is a voice evaluation value based on the voice packet after thinning;
Calculating a third evaluation value that is a voice evaluation value based on the first evaluation value, the second evaluation value, and the ratio;
A voice evaluation method comprising: outputting the first evaluation value when the ratio is equal to or less than a threshold; and outputting the third evaluation value when the ratio exceeds the threshold.

Images