1、CCITT RECMN*G.765 92 W 4862591 0575787 335 INTERNATIONAL TELECOMMUNICATION UNION CCITT THE INTERNATIONAL TELEGRAPH AND TELEPHONE CONSULTATIVE COMMITTEE GENERAL ASPECTS OF DIGITAL TRANSMISSION SYSTEMS; TERMINAL EQUIPM ENTS G.765 (09/92) PACKET CIRCUIT MULTIPLICATION EQUIPMENT i Recommendation G.765 C
2、CITT RECMN*G=765 92 4862591 0575788 271 FOREWORD The CCITT (the International Telegraph and Telephone Consultative Committee) is a permanent organ of the International Telecommunication Union (ITU). CCITT is responsible for studying technical, operating and tariff questions and issuing Recommendatio
3、ns on them with a view to standardizing telecommunications on a worldwide basis. The Plenary Assembly of CCITT which meets every four years, establishes the topics for study and approves Recommendations prepared by its Study Groups. The approval of Recommendations by the members of CCITT between Ple
4、nary Assemblies is covered by the procedure laid down in CCITT Resolution No. 2 (Melbourne, 1988). Recommendation G.765 was prepared by Study Group XV and was approved under the Resolution No. 2 procedure on the Ist of September 1992. CCIT NOTES 1) telecommunication administration and a recognized p
5、rivate operating agency. In this Recommendation, the expression “Administration” is used for conciseness to indicate both a 2) A list of abbreviations used in this Recommendation can be found in Annex A. o ITU 1993 All rights reserved. No part of this publication may be reproduced or utilized in any
6、 form or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the ITU. CCITT RECMN*G.765 92 48b259L 0575789 108 Recommendation G.765 PACKET CIRCUIT MULTIPLICATION EQUIPMENT (1992) 1 General considerations of packet circuit multiplication eq
7、uipment This Recommendation is intended as a base document for the specification and interconnection of packet circuit multiplication equipment (PCME) and packet circuit multiplication systems (PCMS) from various manufacturers. PCME provides for the compression and packetization of several types of
8、traffic. A PCME converts speech, voiceband data, facsimile, channel-associated (i.e. in-band) signalling, common channel signalling, video and digital data information rom primary rate channel formats or synchronous digital hierarchy (SDH) level 1 formats to link access procedure D-channel (LAPD)-li
9、ke frame format. The LAPD-like link layer protocol is used with unacknowledged operation to limit delay in the network. The LAPD-like frames are transported as packet streams in a wideband packet network over a full or fractional primary channel, or an SDH virtual tributary. Wideband packet technolo
10、gy, as used herein, refers to packet systems requiring transmission channels capable of supporting rates above 64 kbitls up to 150 Mbitls. Application-specific protocol at layers 3 and above are used to transport the various types of traffic. A functional representation of a PCME node is shown in Fi
11、gure 1/G.765. Originating endpoint Channel-oriented side L Packet side I- VI 1 I I A Terminating endpoint P HGURE 1/G.765 Endpoint node Recommendation G.765 (09/92) 1 I CCITT RECMN*G-7b5 92 = 4862593 0575790 92T On the channel-oriented (full-rate) side, 1544 kbii/s and/or 2048 kbits or SDH synchrono
12、us transport module (STM)-l interfaces are provided. A time slot interconnect function can connect any time slot or group of time slots to the proper processing function as required to packetize the incoming channel-oriented traffic. A frame cross-connect function transfers the layer 2 produced by t
13、he processing functions to the appropriate packetized side interface. On the packetized side, 1544 kbits and/or 2048 kbitfs or SDH STM-1 interfaces are provided to carry the bit-serial packet streams. PCME aliows networking in both the circuit and packet domains, offers bandwidth on demand, and achi
14、eves graceful degradation of voice quality during congestion. A reference model for a PCME network is shown in Figure 2/G.765. Originating endpoint/ terminating endpoint Packetized side Channel-oriented side FIGURE 2/G.765 Reference Model for a PCME network 1.1 Speech processing For speech traffic,
15、the input speech samples may be coded at the originating endpoint of the access node before packetization by one of several coding methods. The stream of coded speech is transformed into packets with the format specified in Recommendation G.764. The samples are collected over a period of 16 ms and d
16、ivided into blocks, as defined in Recommendation G.764. The blocks are arranged to facilitate block dropping, as explained below. Periods of activity and inactivity are respectively called ?bursts? and ?gaps?. It is not necessary to transmit packets during gaps and silent intervais may be removed. T
17、he terminating endpoint at the far side reconstructs a continuous stream of speech from the incoming packets using the information in the packet header: the time stamp (TS) value and the sequence number. The time stamp field stores tbe value of the total accumulated variable delay that a packet has
18、experienced. Each node adds to the time stamp value, the time it took to serve the packet using its local clock as a reference. 2 Recommendation 6.765 (091p2) CCITT RECMNxG-7b5 72 4Bb259L 0575793 Bbb Recommendation 6.765 (09D2) 3 I Build-out procedures compensate for the variable delay that packets
19、may experience within the network. These procedures apply for the first packet of a voiceband spurt, for ail signalling packets and for the first packet after a missing packet is detected. When any of the above packets arrive within the time stamp value less than the value of the build-out., they ar
20、e stored for the following duration: Time before play-out = Build-out - Time stamp value If they arrive with a time stamp value that exceeds the build-out, they are discarded. Other packets are placed in the play-out queue in the order of their 3) an LAPD specific procedure; 4) a V. 120 specific pro
21、cedure. Interconnection with digital data on 64 kbit/s clear channels may involve ways to increase the compression gain, such as removing HDEC flags and/or network idle codes. 1.5 Facsimile For group 3 facsimile transport, the V.21 handshake signals may be transported as voiceband data in voice pack
22、ets defined in Recommendation G.764. The coded page information is demodulated to extract the baseband signais, which are transmitted in the facsimile frames described in P 12. There are three types of facsimile frames: 1) facsimile capability indication frames; 2) spurt header frames that contain m
23、odem control information; 3) facsimile page information frames that contain the T.4 coded image information UN unscrambled format. The PCME at the terminating end recombines the facsimile frames to reconstruct the original facsimile signal. 1.6 Video trunspolt Recommendation H.221 defines how video,
24、 audio signals of Recommendation H.261 at the rate of 64 kbit/s or less, and various data rates are framed, then submultiplexed, and combined with audiovisual control and indication signals for transmission over the multiple By multiple HO, H11 or H12 channels. In particular, the “information transf
25、er rate command“ defined in Recommendation H.221 ailows the transmitting terminal to vary the number of B or HO channels carrying the combined audiovisual information at 20 ms intervals. Thus, if the format of Recommen- dation H.221 is converted to a packetized version, variable rate video could be
26、transmitted over the wideband packetized system so as to improve video quality during motion without requiring the continuous allocation of the peak bit rate. Recommendation H.26 1 specifies coding of video for visual telephony and video teleconferencing. Recom- mendation H.261 can encode video at r
27、ates suitable for operation over one or six B (64 kbit/s) channels, one to five HO (384 kbit/s), or an HI 1 (1536 kbids) or H12 (1920 kbit/s) channel. The actual source coding rate can be controlled to accommadate fixed bit-rate transmission with suitable transmit buffering and coding control feedba
28、ck. Alternatively, the source coding rate of the video signal can be ailowed to vary with picture complexity and motion if a variable rate transmission facility is available. The video processing frames on the H.221 signai contain the bit rate allocation signal of the H.221 framing format. These fra
29、mes constitute one or more packets each of 20 ms of H.221 signai. 4 Recommendation 6.765 (0992) CCITT RECMN*G-765 92 48b259L 0575793 639 2 2.1 2.1.1 2.1.2 Interfaces I544 kbit/s interfaces Physical interface The physical interface conforms to the specifications in Q 2 of Recommendation G.703. Frame
30、structure The basic frame structure is shown in 4 2.1 of Recommendation G.704. The characteristics of the frame structure carrying channels at various bit rates in 1544 kbit/s is given in 4 3 of Recommendation G.704. 2.2 2048 kbit/s interface 2.2.1 Physical interface The physical interface conforms
31、to the specifications in 8 6 of Recommendation G.703. 2.2.2 Frame structure The basic frame structure is shown in Q 2.3 of Recommendation G.704. The characteristics of the frame structure carrying channels at various bit rates in 2048 kbit/s is given in 4 5 of Recommendation G.704. Bit 1 of the fram
32、e can be used in accordance with Q 2.3.3 of Recommendation G.704 for a cycle redundancy check (CRC) procedure. 2.3 SDH SW- I SDH STM-1 may be supported. The primary level signals shall be mapped as virtuai containers (VCs) (VC-11 for 1544 kbit/s or VC-12 for 2048 kbit/s) into an STM-1 (155 520 kbit/
33、s) as described in Recommenda- tion G.709. The electrical interface characteristics for STM-1 shall conform to Q 12 of Recommendation G.703. 2.4 Channel-oriented side interconnection with the packetized side By service Administration, it shall be possible to treat any time slot, or contiguous group
34、of time slots, on the channel-oriented interface as a single circuit, and to connect that circuit to functions that provide packetization for: 1) speech and voiceband data; 2) signalling; 3) facsimile; 4) video; 5) data. 2.5 Packetized-side interconnection It shall be possible to direct any flow of
35、packets on a channel-oriented side circuit to any packet stream on the packetized side. Any packet stream from the packetized si voiceband data at less than 4800 bids; voiceband data at 1200 to 2400 bids; voiceband data at less than 1200 bids; 3.3 Speech packetization Speech packetization shaii be d
36、one according to Recommendation G.764. As an example in a representative application, voiceband data at rates exceeding 9600 bids shall be carried in pulse-code modulation (PCM). Voiceband data at rates greater than 7200 bids and less than or equal to 9600 bit/s shall be encoded using the 40 kbit/s
37、fixed rate algorithm of Recommendation G.726. Voiceband data at 1200 to 4800 bids shall be encoded using the 32 kbit/s fixed rate algorithm of Recommendation G.726. Voiceband data at less than 1200 bids shall be encoded as for speech, using the 24 kbit/s fixed rate algorithm of Recommendation G.726.
38、 Decoding shall be done using the algorithm indicated in Recommendation G.764 header. Note I - In a national network, the time stamp (TS) and the build-out procedures of Recommendation G.764 may be replaced by a fixed-delay for the fmt packet. The fixed additional delay for the first packet of a spe
39、ech burst shall be administerable to accommodate transit times through networks or through interconnection nodes not implementing TS updating. At the originating endpoint, the TS will be set to zero. In an intermediate node, the TS field will not be updated. However, the build-out procedure will be
40、used always on network-network and user-network interfaces. Note 2 - To meet the allowable transmission delay given in Recommendation G.114, the build-out shall be selected such that the total delay (propagation + build-out) shall not exceed the 400 ms allowance for voice traffic. The build-out shou
41、ld not exceed 70 ms in normal conditions. 3.3.1 ADPCMPCM transcoding A specific incoming channel may be defined by administration to correspond to information containing p-law or A-law voiceband samples. 3.3.2 PCM operation Recommendation G.764 specifies that the ADPCM coder on both sides shall be r
42、eset at the beginning of every speech burst. This is the case when all samples for a speech burst are coded with the same ADPCM algorithm specified in the coding type (CT) field. However, Recommendation G.764 does not specify the action of the ADPCM coders when the part of the traffic is coded accor
43、ding to Recommendation G.711. When an originating PCME transports the incoming channel-oriented traffic in PCM format, it shall arrange the PCM coded speech as described in Recommendation G.764. The resultant voice packets shall have the format described in Figure 2/G.764, and shall be transmitted o
44、n the packetized side of the endpoint. 6 Recommendation 6.765 (09D2) CCITT RECMN*G-765 92 m 4862593 0575795 403 m The ADPCM encoder at the originating endpoint shall operate on the incoming PCM signal using the 40 kbit/s fixed rate ADPCM algorithm of Recommendation G.726, even though its APCM output
45、 is not transmitted. The ADPCM codec at the terminating endpoint shall operate as an encoder in an identical manner as the ADPCM codec of the originating endpoint, and its ADPCM output shall not be transmitted. Thus, when there are no line errors, the state variables of both ADPCM coders that are de
46、scribed in Tables 6/G.726 and 7/G.727 shall be identical for both endpoints. Furthermore, both ADPCM codecs shall be tracking the input PCM signal and would be ready to switch to their respective normal operations without introducing adaptation gaps. When the originating endpoint PCME begins coding
47、the incoming traffic, using the ADPCM algorithm indicated in the coding type field of the packet header, the ADPCM coded speech shall be arranged as described in Figure 7/G.764 and then transmitted in voice packets with the formats shown in Figure 2/G.764. At the terminating endpoint, the ADPCM code
48、c shall decode the speech, using the ADPCM algorithm indicated in the coding type field to reconstruct the original signal, and then convert it to PCM, as indicated in Recommendations G.726 and G.727. In the case of the embedded ADPCM algorithms of Recommendation G.727, the decoder at the terminatin
49、g endpoint shall also use the information coded in the block dropping indicator field to select the decoding algorithm to use. 4 Video processing The PCME shall implement packetization of fixed-rate and variable-rate video for teleconferencing and videotelephony. The specific procedures are for further study. 5 Interface with cellular speech packetization networks This item is for further study. 6 Digital data interface The PCME shall be equipped to receive, packetize and transmit the following digital data traffic arriving on the channel-oriented side as: 1) special traffic, using the D
