CN1097910C - Speech communication method - Google Patents
Speech communication method Download PDFInfo
- Publication number
- CN1097910C CN1097910C CN99117051A CN99117051A CN1097910C CN 1097910 C CN1097910 C CN 1097910C CN 99117051 A CN99117051 A CN 99117051A CN 99117051 A CN99117051 A CN 99117051A CN 1097910 C CN1097910 C CN 1097910C
- Authority
- CN
- China
- Prior art keywords
- speech
- label
- voice
- packet
- signaling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
- Telephonic Communication Services (AREA)
Abstract
The present invention discloses a speech communication method realized on the basis of connection-oriented grouping speech label switching technique. The present invention comprises the procedures that signallings are used for establishing connection, label distribution is obtained, label switching is used for transmitting digitized grouping speech on the basis of established connection, and labels in connection release are removed after conversation is completed. The present invention can also be used for circuit simulation and connecting type service which needs to guarantee a bandwidth, such as FR, except for speech service. The conversation method overcomes the defects of ATM speech transmission and IP net speech transfer. The present invention has advantages of low cost, high channel utilization and good conversation real time, and can share communication resources with IP net data transmission.
Description
The present invention relates to the communication technology, be specifically related to can with the speech transmissions technology of the Internet protocol data transmission technology compatibility, more particularly, relate to a kind of based on towards connecting and the speech communication method of speech packet technology.
With regard to the voice over packet present situation, two big mainstream technologys are arranged at present, first kind of speech that is atm forum VTOA (Voice Telephony Over ATM, i.e. ATM phone) group advocates fits on ATM (the Asynchronous Transfer Mode asynchronous transfer mode) network by AAL1 (ATM first kind adaptation layer protocol), AAL2 (the ATM second class adaptation layer protocol) and transmits.ITU-T also supports VTOA, and the signaling standard of having formulated AAL2 is (ITU-T is about the AAL2 signaling standard) Q.2630.1, and the little cell switching net that superposes again on atm network is realized the little cell switching of AAL2.Now also do not support the ATM commercial product of the little cell switching of AAL2 to release, let alone activate the service.Second kind is to be IPTEL (IP phone group) VOIP that tissue is advocated energetically of representative with IETF, and promptly speech carries by IP network.Now existing businessman releases product, and the operator activates the service one after another, attracts the telephone subscriber with low rate.The former transmits speech on atm network, and adopts the little cell switching of AAL2, and efficiency of transmission is all superior than IP phone with quality of service, and networking also can be equally flexible with second kind of VOIP that is adopted (Voice Over IP, i.e. IP phone).But along with popularizing of Internet, following global metadata traffic carrying capacity must surpass voice service.Therefore public network of future generation must be data-centered, and it is in the public Packet Based Network of feature that speech is added to IP (Internet protocol) business as auxiliary activities.As if so on the surface, as if it is quite reasonable to fuse speech with IP, feels to defeat VTOA, and second kind of scheme has its points of course.No matter in fact be to use atm network, still transmitting packet voice with IP network is not suitable method, and this is determined by the interactivity of the requirement of Speech Communication, real-time characteristics.Below this is analyzed.
To transmitting speech with ATM; at first; the cell length of ATM is oversize to packet voice; be difficult to reach the real-time requirement if directly exchange speech (particularly compression after packet voice) with ATM, little cell switching net of the AAL2 that has to superpose again on atm network is guaranteed Speech Communication interactivity and real-time requirement; efficiency of transmission is also higher simultaneously; but this stacked system has increased technical difficulty, and it is complicated that switching equipment also can become, unfavorable to cost.In addition, the cell length of ATM is too short again concerning data, and usually the IP long data packet can be greater than the payload field of cell, so cell tax has reduced efficiency of transmission.The most important thing is that connection-oriented service bearer technology of ATM and disconnected IP have any different in essence, transmitting the IP packet with ATM is not optimal design.Take a long view future based on the conditions of demand of IP data service under ATM can not become service bearer technology in the new public network for a long time.
Concerning transmit speech on IP network, the multi layer protocol stack that transmits according to data encapsulates speech packet step by step, and the expense all too is big, and efficiency of transmission is extremely low.Voice service and data service are just the opposite on service request, and be if carry all business with IP in the new public network, although be optimal design to the data business, opposite to voice service.The disconnected attribute of IP is not suitable for the interactivity and the real-time requirement of Speech Communication in essence, and the quality of IP phone is difficult to reach the carrier class requirement.
Therefore, single say that current connection-oriented TDMA (time division multiplexing) exchange and transmission technology are considered comprehensively, have been optimal design, are difficult to find better substitute technology from cost, on the speech quality with regard to voice service.But because the variation of business demand, the position of data service will be higher than the position of voice service, and voice service has to submit to the packet switching network with the data service optimal design.Tool prospect be that IP optimizes optical-fiber network, it will be the optimal design to the data business, if adopt VOIP to transmit speech, to voice service but not so.Development of technology is wasted transfer resource in a large number without any reason, more has no reason to reduce the quality level of Speech Communication.Must seek another approach on new public network can both be optimal design to the data business but also to voice service not only.
The objective of the invention is to seek a kind of new speech communication method, this method can overcome the shortcoming of ATM speech transmissions and the transmission of IP network speech, the desired real-time of Speech Communication, interactivity requirement had both been considered, consider with the compatible of transfer of data exchange again and can utilize existing communication protocol and equipment and can expanding on the following new public network uses, specifically, this speech communication method expense is lower, the channel usage height, the conversation real-time good, and can with IP network transfer of data common communications resources.
The object of the present invention is achieved like this, constructs a kind of speech communication method, it is characterized in that, comprises connecting, utilize on the basis that connects the label exchange to transmit packet voice, removes step of connecting behind end of conversation.
According to speech communication method of the present invention, it is characterized in that described connecting may further comprise the steps:
Make a call with signaling by source terminal;
Network switch response is called out and to the terminal distribution label,
The destination address of network based calling is selected route,
Each switch uses signaling replay call on selected path, and distributing labels on the way, and generates Label Forwarding Information Base on every switch,
Call proceeding is to the purpose terminal, if the purpose terminal is agreed call accepted, label distribution finishes and finishes connection end to end, then begins conversation.
According to method provided by the invention, it is characterized in that the described label exchange transmission packet voice that utilizes may further comprise the steps on the basis that connects:
With voiceband user digitlization and compression, export a frame of digital speech packet at regular intervals at interval,
Speech to framing carries out the stream protocol processing, promptly adds the stream protocol head,
It is multiplexing to carry out label in terminal, promptly adds distributed labels, carries out link layer process, adds PPP/HDLC head and verification tail,
By physical circuit the network switch sent in this frame of digital speech,
According to transmitting the switch label value, finally deliver to destination in every switch on the way,
Solve the digital voice frame according to the anti-process of source end eventually in purpose, finally be reduced into voice signal.
According to speech communication method provided by the invention, it is characterized in that described connection is removed and be may further comprise the steps:
Any end in the conversation is initiated the request of removing with signaling, and the network switch accepts request, and discharges distributed labels;
Remove information through each switch with the signaling transmission on the way, and discharge distributed labels, refresh and transmit;
Dismounting information is delivered to the other end, discharges whole labels, end of conversation.
According to method provided by the invention, it is characterized in that, set up in the process in described connection, make a call with signaling H.323+ by described source terminal, H.323+ each switch also uses or other network and internetwork signaling replay call on selected path.
According to method provided by the invention, it is characterized in that it is to make a call with signaling H.323+ that described source terminal sends connection request to network.
According to method provided by the invention, it is characterized in that, described can be to utilize OSPF (Open Shortest Path First according to call intent address decision route, OSPF), BGP (Boader Gateway Protocol, Border Gateway Protocol) the IP dynamic routing table of Routing Protocol generation also can be the configurations shown route.
According to method provided by the invention, it is characterized in that H.323+ described all have (NNI) signaling capability between signaling or other network and network interface through each switch on the way.
According to method provided by the invention, it is characterized in that described H.323+ signaling is transmitted by TCP (transmission control protocol) agreement.
According to method provided by the invention, it is characterized in that describedly speech is carried out the step that digitlization and compression handle adopting G.729 (ITU-T is about the standard of compress speech) standard logarithmic word voice compression.
According to method provided by the invention, it is characterized in that describedly at interval exporting a frame of digital speech packet at regular intervals and comprising G.729 packet voice of 2-3 frame.
According to method provided by the invention, it is characterized in that the described stream protocol head that adds is meant according to (ITU-T about in the online speech junction traffic of ATM standard) requirements definition form I.366.2: the packet header that adds the 1-2 byte in the bag front on the digital voice frame, this packet header comprises the packet type position, grouping serial number and other purposes position, wherein, the packet type position point out be pass on a message sound first kind grouping or pass second class grouping of non-voice information, to described first kind grouping, described grouping serial number adopts mould 8 countings.
According to method provided by the invention, it is characterized in that, the definition of described label is identical with tag format among the MPLS (multiprotocol label switching), each label comprises the label value that 20 bits are represented, 3 grades of service that bit is represented, at the bottom of the stack that remaining 1 bit in MPLS is represented and depositing the words time of representing of 8 bits, packet switching does not have purposes for packet voice, keeps temporarily.
According to method provided by the invention, it is characterized in that before carrying out described HDLC frame envelope step, comprising also frame internal information 0X7E is carried out the step that escape is handled that being about to frame internal information 0X7E escape is 0X7D 0X5E, is 0X7D 0X5D with frame internal information 0X7E escape.
Implement provided by the invention based on speech packet label exchange speech communication method, owing to adopt connection-oriented machine-processed VOLS, can be used for except that voice service circuit simulation (CES:CircuitEmulation Service/ circuit emulation service), FR (Frame Relay, frame relay) these need guarantee the connecting-type business of bandwidth.Because connection of every application, the bandwidth of a label representative is not particularly limited, and bag is long unfixing yet, and this just provides flexibility for adapting to the variety classes business.This also is one of new public network ability that need possess, and promptly contains all business in unified network.
In conjunction with the accompanying drawings and embodiments, further specify characteristics of the present invention, in the accompanying drawing:
Fig. 1 is a speech packet label switching protocol layer schematic diagram;
Fig. 2 is a packet voice stream protocol header structure schematic diagram;
Fig. 3 is the structural representation of Shim Header;
Fig. 4 is the simple expression of packet voice information process;
Fig. 5 implements the simple presentation graphs that the inventive method is carried out a voice communication process.
1, packet voice label exchange (VOLS, i.e. Voice Over Label Switch, the exchange of speech label)
IP-based Packet Data Network, to IP packet and speech packet can both optimization approaches be two classes to be divided into groups to treat respectively in network, be that speech packet is not to be carried in the IP bag, but directly enter transmission and exchange in the consolidated network, and keep connection-oriented feature, to satisfy the requirement of Speech Communication interactivity, real-time.The basis that can converge the grouping of two classes in consolidated network simultaneously is label exchange (Label Switch), the protocol stack of terminal and network switching node user face and chain of command when Fig. 1 shows voice over packet.
Routing and forwarding are separated in the label exchange, and this technology has been used for IP exchange (MPLS), and it also can be used for the packet voice exchange simultaneously.Accord with if label is considered as the gap marker of voice user, just can be multiplexing to a plurality of users' mark in the realization of transmission link level, subscriber channel can be exchanged to another link from a link by the label exchange and get on.Label switching plane and transmission link are shared in packet voice and IP packet in same network like this.
2, the controlling unit Speech Communication in the packet voice label exchange (VOLS) need be guaranteed quality by connection-oriented technology, and network is in case connect the contract bandwidth that just must guarantee the user during connecting for the user has set up.Signaling is the product that connects control, should comprise signaling between terminal use and network switching node and the signaling between switching node and switching node in the exchange of packet voice label.If consider to utilize the work of having done as far as possible, signaling between terminal use and network switching node can increase or revise (H.323+) on basis H.323, connect the needs that allow control (CAC:Connection Admission Control/ connection allows control) as satisfying.Perhaps, H.323+ signaling between switching node and switching node can be used, and perhaps formulates new signaling.At first send out connection request by signaling H.323+ to network when the user need converse like this, and to the network applying label, can the Network Check whether resource (as bandwidth) of the end-to-end connection of guaranteed grouping user be confirmed and be connected.
The routing that connects when setting up both can utilize the dynamic routing result (as OSPF, the routing table that the BGP Routing Protocol produces) of IP network also can dispose explicit route.The service traffics on the heavy route of load are dredged in the polarization of loading on the network transmission line, and explicit route is perhaps essential.Packet voice and IP packet can be shared label fully in same net, set up H.323+ united and coordinating of control signaling as long as the LDP (Label Distribution Protocol/ Label Distribution Protocol) when resource management can exchange with the IP label is connected with packet voice, just can guarantee that miscellaneous service dynamically uses label.Because packet voice needs strictly to set up end to end to connect, so each switching node in the network all needs signaling capability, as H.323+.This is equivalent to be the RSVP of end-to-end traffic bandwidth reserved (ReSource reSerVation Protocol/ RSVP) in the exchange of IP label.Both differences are H.323+ static each call distribution bandwidth resources that are, in case the connection that has been certain call setup, connect the bandwidth resources that all keep distribution before up to removing this, and RSVP is each data flow dynamic assignment resource, the i.e. bandwidth resources of wanting continuation application to need every the regular hour, otherwise apparatus for network node (LSR) is in case find overtimely will discharge the bandwidth resources of being reserved.
Chain of command information can transmit in consolidated network in independent IP control net or with user profile, and transmission wherever all requires signaling is transmitted reliably, and it is proper therefore to transmit signaling with Transmission Control Protocol.If in consolidated network, transmit signaling and user profile, can on the transmission link between node device, distribute a label special use, fixing, and distribute suitable bandwidth to transmit signaling.
3, the user data link in the packet voice label exchange (VOLS)
For voice service, be speech information on user's face.In order to reduce the bandwidth of transmission bandwidth, particularly toll transmission line that every road speech consumes, generally all to compress to handle and form packet voice digitized speech.But the packet voice after the compression should be not less than the online speech quality of existing public telephone for the user.Can this be the key point that replace existing telephone network to new public network.3.1) VDSP (speech Digital Signal Processing)
The online digital voice of public telephone adopts the 64Kbps pcm encoder to transmit at present, because the maturation of Digital Signal Processing (DSP) technology and device is hopeful fully to voice compression, saves transmission bandwidth.G.729 the compression standard that can reach the online PCM speech quality of public telephone (MOS=4) (MOS:Mean OpinionScore/ Mean Opinion Score) is, this algorithm can be with every road voice compression to 8Kbps, compression ratio is at least 8 times (considering that quiet inhibition compression ratio is higher), and MOS=4.G.729 only with regard to algorithm, every 15ms exports the long frame packet voice of 80 bits (10 byte).If described in G.114, should be within 150ms for the receptible end-to-end One Way Delay of most of users, therefore at most can be with 2～3 frames after G.729 packet voice collects together the group bag transmit.3.2) the packet voice stream protocol
Behind packet voice group bag, need handle through stream protocol.The packet voice stream protocol mainly is to guarantee to wrap end to end recovery, when promptly each packet voice wraps in and sends sequence number should be arranged, and when through Network Packet Loss or time-delay when too big, receiving end is handled accordingly according to sequence number.The packet voice stream protocol adds a packet header in the bag front, is about 1～2 byte, and formal definition can satisfy requirement I.366.2.Fig. 2 is a relevant function of finishing the stream protocol head with a byte:
I.366.2, reference defines two kinds of packet types: first kind grouping is used to the sound of passing on a message, and second class is used to pass non-voice voice band information, as DTMF specific informations such as (dual-tone multifrequencies).
For first kind grouping, grouping serial number adopts mould 8 countings, is used for receiving terminal and detects packet loss or overtime, thereby take corresponding recovery measure.Remain 3 bits and can be used as other needs.3.3) the multiplexing and exchange of label
Share switching plane in order to exchange, so the definition of label must be identical with the tag format among the MPLS with the IP label.Universal tag is called Shim Header among the MPLS, its formal definition such as Fig. 3.
Shim Header is totally 4 byte longs, and wherein label value is 20 bits, can identify greater than 10 simultaneously
6Individual subscriber channel or data flow, and label value has only local significance, and after subscriber channel exchanged on another section link from one section link, label value had also just changed.For the packet voice business, connect distributing labels when setting up, dismounting discharges label again after connecting, and label resources can reuse.Not only can transmit packet voice but also can transmit the IP packet on every section link like this, only need just can realize that with different label value signs label is multiplexing.
When the packet voice user made a call, network based called address was determined route; And then be that by signaling this calls out the preassignment label value on each section link along this paths; After the label value on every section link all distributed, transmitting also of each switching node on this paths all generated.User's passage is identified by label value fully during conversation, and promptly basis is transmitted and can be realized the surface speed forwarding voiceband user with hardware in switching node equipment.When talk-through was removed this paths, shared label discharged on each section link on the way, simultaneously transmitting also and will refresh in each switching node.From the mechanism of transmitting, with as broad as long to the forwarding of IP grouping among the MPLS, so both can share the switching plane in the node device fully.
The grade of service among the Shim Header (CoS:Class of Service/ service type) is very important, the queue scheduling mechanism of label switching equipment inside will be distinguished priority with the grade of service, give high priority for real time business such as speeches, at first from formation, dispatch away.
For the voice service that connects towards strictness, particularly use explicit route, last two do not have purposes among the Shim Header.3.4)PPP/HDLC
Be that speech packet or IP packet all need be carried out just sending on the physical circuit after the link layer framing encapsulation and transmit, receiving terminal carries out deciding frame, separating frame according to this encapsulation.Be generally the point-to-point contact at the transmission line between node device on the public network, so only link layer protocol is that (PPP is Point to Point Protocol abbreviation to PPP/HDLC, refer to point-to-point protocol, HDLC is Highlevel DataLink Control abbreviation, the Optical synchronization digital transmission network), leased line (special line, dialing) goes up and use refer to high-level data link layer control protocol), and at IP Over SDH (SDH:.Flag of frame symbol in the encapsulation of HDLC frame is 0x7E, transmits and uses 0 bit insertion/delet method to get rid of the possibility of the 0x7E that occurs in the out of Memory in the frame in the rules.This regulation is suitable for Bit Oriented transmission, but these be that the transmission system of unit of transfer is restricted with 8 hytes (Octet) for resembling SDH, so the escape that comes achieve frame internal information 0x7E in the rules of IP over SDH with escape character, stipulates as follows:
0x7E
0x7D 0x5E
0x7D
0x7D 0x5D
It is unit of transfer that this method both had been suitable for 8 hytes, is suitable for the Bit Oriented system for transmitting again, and excellent adaptability is arranged, and should promote the use of in new public network.4. illustrative examples
Be provided with two terminals of A, B and link to each other through two switches, A makes a call to B, and switch 1 is given its value of A distributing labels L=15, switch 1 is called to switch 2, distributing labels value L=202, switch 2 find that purpose terminal B can arrive distributing labels value L=87.The digital voice PPP/HDLC frame that Fig. 5 transmits when being conversation on the line, every through a switch, its label value changes.The packet voice of sending into switch 1 as terminal A uses distributed labels value 15, and behind the arrival switch 1, label value is exchanged for 202, sends into switch 2, and last label value is exchanged for 87 and delivers to purpose terminal B.In case after packet voice was sent into switching network, switch just to the label value exchange, can not done any processing to high level.Wherein, used following prior protocols, do not change: 1) label: be used for and IP packet shared link resource, realize the multiplexing and exchange of label, be taken from " LDP Specification " IETF (Internet Engineering Task Force/ Internet engineering duty action tissue) draft-feldman-ldp-spec-00.txt; 2) PPP/HDLC agreement is taken from " PPP over SONET/SDH " IETF draft-ietf-pppext-sonet-ds-01.txt.5. transmission efficiency relatively
Below existing VTOA, VOIP mode transporting speech method and speech transfer approach based on VOLS of the present invention are done one relatively in transmission efficiency, analyze the expense of variety of way bandwidth resources.
As adopt the G.729 voice compression algorithm of MOS=4, consider the restriction of end-to-end One Way Delay 150ms, G.729 per two frames export group bag transmission together, and the packing time delay is 30ms, add the processing delay of DSP, the speech packet time delay of transmitting terminal is expected to be controlled in the 35ms.5.1)VTOA
For the little cell switching of VTOA/AAL2, each ATM cell can be adorned the two-way compressed voice, and transmission information is composed as follows:
2 (the little header of 3 bytes+20 byte compressed voice)+1 byte offsets indication+1 byte is filled+5 byte ATM cell heads
Expense totally 13 bytes in the ATM cell of 53 byte longs, overhead rate is 13/53=24.5%.5.2)VOIP
For the application VOIP on the IP V4, each IP packs one road speech, and transmission information is composed as follows: 20 byte compressed voice+12 byte RTP (real time transport protocol) head+8 byte UDP (UserDatagram Protocol/ User Datagram Protoco (UDP)) head+20 byte IP heads+7 byte PPP/HDLC heads
Expense 47 bytes in the transmit frame of 67 bytes, overhead rate is 47/67=70.1%.
When adopting MPLS IP exchange, expense also will increase by 4 bytes, and overhead rate increases and causes 71.8%.If IP evolves to the next generation, promptly IP V6 adopts 64/128 bit IP address, and the IP head will be increased to 24/40 byte, and overhead rate further increases.5.3)VOLS
For VOLS of the present invention, each packs 1 road speech, and referring to Fig. 4, transmission information is composed as follows:
20 byte compressed voice+1 byte packet speech stream protocol header+4 byte Shim Header+7 byte PPP/HDLC heads
Expense 12 bytes in the transmit frame of 32 bytes altogether, overhead rate is 12/32=37.5%
Comparative result is summarised in following table.
| Mode | Speech channel number/voice packets long (byte) | Transmit message length (byte) | Transmission efficiency (%) |
| VTOA/ AA2 | 2/40 | 53 | 75.5 |
| | 1/20 | 32 | 62.5 |
| | 1/20 | 67 | 29.9 |
This shows, though transmitting expense, the speech of VOLS of the present invention exceeds 13 percentage points than VTOA/AAL2, but the expense than VOIP has reduced nearly one times, and comparative result is that VOIP transmits the required bandwidth resources of one road speech more intuitively, if can pass the two-way speech with VOLS.More crucial is that VOLS can fuse in one network by high IP with transmitting data efficiency, and particularly Wei Lai new public network will be optimized optical-fiber network based on IP, and VOLS is more meaningful.6. service application is promoted
Speech based on VOLS of the present invention transmits just towards one of business that connects, this connection-oriented mechanism, can be used for except that voice service circuit simulation (CES), FR these need guarantee the connecting-type business of bandwidth.Because connection of every application, the bandwidth of a label representative is not particularly limited, and bag is long unfixing yet, and this just provides flexibility for adapting to the variety classes business.This also is one of new public network ability that need possess, and promptly contains all business in unified network.If attempt Every Thing over IP, since the disconnected feature of IP, the very difficult quality that effectively guarantees these connecting-type business thereon.
Claims (11)
1, a kind of speech communication method is characterized in that, comprises following three steps:
A, connect, wherein saidly connect and comprise:
A1, make a call with signaling by source terminal;
The response of A2, the network switch is called out and to the terminal distribution label;
The destination address of A3, network based calling is selected route;
A4, each switch uses signaling replay call on selected path, and distributing labels on the way, and generates Label Forwarding Information Base on every switch;
A5, call proceeding are to the purpose terminal, if the purpose terminal is agreed call accepted, label distribution finishes and finishes connection end to end, then begins conversation;
B, on the basis that connects, utilize label exchange to transmit packet voice, wherein saidly on the basis that connects, utilize the label exchange to transmit packet voice to comprise again:
B1, with voiceband user digitlization and compression, export at interval a frame of digital speech packet at regular intervals;
B2, the speech of framing is carried out stream protocol handle, promptly add the stream protocol head;
B3, to carry out label in terminal multiplexing, promptly adds distributed labels, carries out link layer process, adds PPP/HDLC head and verification tail;
B4, the network switch sent in this frame of digital speech by physical circuit;
According to transmitting the switch label value, finally deliver to the purpose terminal in B5, every the switch on the way;
B6, solve the digital voice frame according to the anti-process of source terminal, finally be reduced into voice signal in the purpose terminal;
C, behind end of conversation, remove to connect, wherein said behind end of conversation, remove to connect comprise again:
Any end in C1, the conversation is initiated the request of removing with signaling, and the network switch accepts request, and discharges distributed labels;
C2, remove information through each switch with the signaling transmission on the way, and discharge distributed labels, refresh and transmit;
C3, dismounting information are delivered to the other end, discharge whole labels, end of conversation.
2, method according to claim 1, it is characterized in that, set up in the process in described connection, made a call with signaling H.323+ by described source terminal, H.323+ each switch also uses or other network and internetwork signaling replay call on selected path.
3, method according to claim 2 is characterized in that, it is to make a call with signaling H.323+ that described source terminal sends connection request to network.
4, method according to claim 3 is characterized in that, described can be the IP dynamic routing table that utilizes OSPF, BGP Routing Protocol to produce according to call intent address decision route, also configurations shown route.
5, method according to claim 1 is characterized in that, H.323+ described all have (NNI) signaling capability between signaling or other network and network interface through each switch on the way.
According to claim 2,3 or 5 described methods, it is characterized in that 6, described H.323+ signaling is transmitted by Transmission Control Protocol.
7, method according to claim 1 is characterized in that, describedly speech is carried out the step that digitlization and compression handle adopts G.729 standard logarithmic word voice compression.
8, according to claim 1 according to method provided by the invention, it is characterized in that describedly at interval exporting a frame of digital speech packet at regular intervals and comprising G.729 packet voice of 2-3 frame.
9, according to claim 7 or 8 described methods, it is characterized in that the described stream protocol head that adds is meant according to requirements definition form I.366.2: the packet header that adds the 1-2 byte in the bag front on the digital voice frame, this packet header comprises the packet type position, grouping serial number and other purposes position, wherein, the packet type position point out be pass on a message sound first kind grouping or pass second class grouping of non-voice information, to described first kind grouping, described grouping serial number adopts mould 8 countings.
10, according to claim 7 or 8 described methods, it is characterized in that, described label definition is identical with the tag format of MPLS, each label comprises the label value that 20 bits are represented, 3 grades of service that bit is represented are at the bottom of the stack that remaining 1 bit in MPLS is represented and depositing the words time of representing of 8 bits.
11, method according to claim 1, it is characterized in that before carrying out described HDLC frame envelope step, also comprise frame internal information 0X7E is carried out the step that escape is handled, being about to frame internal information 0X7E escape is 0X7D 0X5E, is 0X7D 0X5D with frame internal information 0X7E escape.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN99117051A CN1097910C (en) | 1999-08-27 | 1999-08-27 | Speech communication method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN99117051A CN1097910C (en) | 1999-08-27 | 1999-08-27 | Speech communication method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1286556A CN1286556A (en) | 2001-03-07 |
| CN1097910C true CN1097910C (en) | 2003-01-01 |
Family
ID=5279698
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN99117051A Expired - Fee Related CN1097910C (en) | 1999-08-27 | 1999-08-27 | Speech communication method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1097910C (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101222477B (en) * | 2007-01-10 | 2011-08-10 | 华为技术有限公司 | Method, equipment and system for initiator and receiver cancelling call |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5553068A (en) * | 1990-03-19 | 1996-09-03 | Fujitsu Limited | ATM cell broadcasting system |
| GB2320159A (en) * | 1996-10-02 | 1998-06-10 | Ibm | Aggregate Inernet route switching |
-
1999
- 1999-08-27 CN CN99117051A patent/CN1097910C/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5553068A (en) * | 1990-03-19 | 1996-09-03 | Fujitsu Limited | ATM cell broadcasting system |
| GB2320159A (en) * | 1996-10-02 | 1998-06-10 | Ibm | Aggregate Inernet route switching |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1286556A (en) | 2001-03-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7486684B2 (en) | Method and apparatus for establishment and management of voice-over IP virtual private networks in IP-based communication systems | |
| US6829254B1 (en) | Method and apparatus for providing efficient application-level switching for multiplexed internet protocol media streams | |
| CN1316797C (en) | Method and apparatus for creating a packet using a digital signal processor | |
| US20060133359A1 (en) | Multiprotocol convergence switch (MPCS) and method for use thereof | |
| KR100776408B1 (en) | System and method for enhancing a voice channel in voice over internet protocol | |
| US20020176403A1 (en) | Method and system for communicating voice over IP access networks | |
| US7012922B1 (en) | Packet communications system and method | |
| CN1217862A (en) | System supporting variable bandwidth asynchronous transfer mode network access for wireline and wireless communications | |
| CN1859077A (en) | Time division multiplex data transmitting method based on packet exchanging | |
| CN1391760A (en) | Packet pipe architecture for access networks | |
| CN100481808C (en) | Tunneling TDM traffic over MPLS | |
| CN1812355A (en) | Method for guaranteeing end-to-end business service quality and switching in network | |
| CN1097910C (en) | Speech communication method | |
| CN102238164A (en) | Internet protocol (IP) telecommunication network-oriented multi-protocol message adaptation method | |
| US7292546B2 (en) | Bandwidth reduction within packet switched networks by not sending idle timeslots | |
| EP1346599B1 (en) | Method of optimizing equipment utilization in telecommunication access network | |
| CN101043382A (en) | Method for guarantying NGN end to QoS end radically | |
| CA2379437A1 (en) | Communications using hybrid circuit-switched and packet-switched networks | |
| CN1275432C (en) | Analog local net realizing method based on digital phonetic exchanging net | |
| CN100341293C (en) | Time-delay channel distributing method | |
| CN1081419C (en) | Wide band multi-business digital switching system | |
| EP1429516A1 (en) | Method and gateway for transportation of stream traffic | |
| CN1368814A (en) | Method for configuring network by overlapping ATM with Internet protocols | |
| Chan et al. | Voice over quantum IP | |
| CN1625175A (en) | Method for transmitting multi-protocol tag exchange protocol data unit |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C19 | Lapse of patent right due to non-payment of the annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |