CN100471290C - Method for implementing half-duplex IP voice communication - Google Patents

Abstract

This invention provides a method realizing half duplex IP voice communication to solve the problem of incapable realizing pivotal half duplex control in PTT business according to RTP/RTCP in existing technology. The procedures are as follows a) voice data from calling subscriber is encoded to form speech packet after the successful call build. Identifier of indicating speech's beginning or finishing instruction is inserted into packet header of the voice packet. b) Voice packet with identifier is sent to control center through IP network, the speech of beginning of finishing of the subscriber is determined by control center according to the identifier. c) Voice packet is transmitted to receiving terminal by the center to accomplish the voice communication. PTT business that used in special swarm system or interphone is led in mobile or fixed IP communication network in this invention. It makes the network of using IP business to has new business features.

Description

Realize the method for half duplex IP voice communication

Technical field

The present invention relates to communication technique field, is that a kind of (InternetProtocol: internet protocol) network carries out the method for half-duplex voice communications by IP specifically.

Background technology

PTT (Push To Talk: button is promptly said) business is the communication service of a kind of point-to-point or multi-point speech, the user presses certain special keys (below be called the PTT button) back of certain terminal and speaks, its other party of conversation can only be listened this moment, and the user then can only listen other people to say after unclamping this special key.This business is widely used in various special-purpose trunked communication systems or intercom, its main business is characterized as can speak after the user presses the PTT button, network directly sends to these voice the member of whole group, communication mode adopts half-duplex mode, other users can only listen not talkative when the user spoke, features such as unclamp the PTT button after the user finishes, other users can press the request of PTT key and make a speech, and this PTT business is easy to operate because of it, use is flexible are subjected to users' welcome deeply.

Along with the communication technology, the particularly development of mobile communication (comprising CDMA, GSM, WCDMA, WLAN etc.) technology, the ability that provides IP to communicate by letter for terminal has been provided communication network.Gyp IP telephony network is mainly kept several parts such as (Gatekeeper), back-up system and telephone network and IP network by IP phone gateway (Gateway), pass and is constituted.It by gateway, the pass is kept and other back-up systems interconnect telephone network and IP network, utilize IP network in technology and advantage economically, use the IP network resource to greatest extent, realize the intercommunication of telephone service in two nets, make IP telephony network obtain using widely.

Be the transmission on IP network behind the guarantee speech information process gateway compressed encoding, RTP) and RTCP (Real-time Transport ControlProtocol: RTCP Real-time Transport Control Protocol) method of existing realization half duplex IP voice communication has adopted RTP (Real-time TransportProtocol:, this RTP/RTCP protocol application is in the point-to-point or the multi-point of IP network, its applied environment mainly considers to realize full duplex or single worker's media communication on local area network (LAN) or Internet, therefore the control procedure when but this method is not considered half-duplex operation can't realize half-duplex control crucial in the PTT business by RTP/RTCP.

Summary of the invention

At said circumstances, the present invention proposes a kind of method that realizes half duplex IP voice communication, to solve the problem that can't realize half-duplex control crucial in the PTT business that prior art exists by RTP/RTCP.

Solution of the present invention is such: a kind of method that realizes half duplex IP voice communication, and the method includes the steps of:

After a, the call setup success, calling party's voice data is encoded, form voice packet, and in the packet header of this voice packet, insert the identifier that shows speech beginning or END instruction;

B, this voice packet that has identifier is sent to control centre by IP network, this control centre judges according to this identifier, decision beginning or finish calling party's speech;

C, this control centre are transmitted to the side's talked about terminal with voice packet, finish voice communication.

Wherein, judge specifically according to this identifier among the described step b to be meant: if this identifier is 1, begin calling party's speech so, otherwise finish calling party's speech.

Beginning calling party's speech also further comprises among the described step b:

Judge whether to exist the current speaker, if there is no, begin calling party's speech so; Otherwise the priority that compares current speaker and calling party is if calling party's priority height begins calling party's speech so.

A kind of method of half duplex IP voice communication that realizes of the present invention is by the link in the encoded speech data packing, identifier with special implication will be inserted in the packet header of voice packet, it is the PTT sign, coming provides control information for control centre, and moving or fixedly realizing semiduplex voice communication on the IP network, thereby will originally only use professional introducing of PTT in special-purpose group system or intercom to move or fixing IP communication network, and make the communication network that utilizes IP increase new traffic performance.

Introduce the present invention in detail below in conjunction with description of drawings and specific implementation.

Description of drawings

Fig. 1 is the structural representation that the embodiment of the invention is formed system;

Fig. 2 is the method flow schematic diagram of the embodiment of the invention.

Embodiment

In order to understand present embodiment better, at first simply introduce the related IP telephony network of present embodiment.

Gyp IP telephony network is mainly kept several parts such as (Gatekeeper), back-up system and telephone network and IP network by IP phone gateway (Gateway), pass and is constituted.It by gateway, the pass is kept and other back-up systems interconnect telephone network and IP network, utilizes IP network in technology and economically advantage, uses the IP network resource to greatest extent, realizes the intercommunication of telephone service in telephone network and IP network.And gateway, the pass is kept and back-up system is being undertaken work such as communication Protocol Conversion, address transition, calling connection, authentication, charging collection and network management jointly, wherein introduces and the closely-related gateway of the present invention, voice compression coding techniques and its Real-time Transmission.

Gateway, gateway device are between telephone network and the IP network, are special machines, and its major function has: provide and the interface of telephone network interconnection and the interface that interconnects with IP network; Finish the compressed encoding of speech; The conversion of communication protocol; Carry out Route Selection; The generation of metering data; For the caller phone user provides voice prompt function or the like.

Gateway is cheated stored-program control exchange (PBX:Private Branch eXchange) by simulating a typical telephone network.When the user had dialled number and begins to call out, information was sent to PBX, and PBX delivers to local gateway with start information then; Afterwards, the IP address that local gateway is tabled look-up and obtained remote gateways according to called number, two gateways are set up a connection, and the gateway of receiving terminal is asked local PBX for instructions and finished this calling, phone ringing then, the side's talked about off-hook.Along with closing of the circuit, the gateway that begins to give orders or instructions wraps the gateway that sends to reception by IP network with these after the calling flow is sent into the coded system coding, the gateway that receives is carried out reverse process, rearranges also decompress(ion), package, transfers them to the recipient there again.

Wherein coded system coding just needs to adopt the voice compression coding techniques, and the voice compression technology is owing to be subjected to the restriction of the network bandwidth, and IP phone is always wished the coding method of adopting compression ratio high more.G.771 ITU-T speech coding standard mainly contain, G.726, G.728, G.729, G.723.1 wait.G.723.1 IP phone generally adopts at present, this G.723.1 the speech coding standard be 6.3K/5.3K dual rate speech coding standard, mechanism with silence detection, noise filling and lost frame recovering etc., and speech quality is relatively right, can be to outer other sound of speech, effectively compressing as music etc., is the coding standard that the network phone product of many maturations is supported.Certainly, now the lower compression method that compares of encoding rate has a lot, the compression method that has in addition reach 1.2K lower, but because the influence of composite factors such as their speech quality, coding rate, environmental suitability, G.723.1 its degree of popularizing still is not so good as.The G.723.1 codec that IP phone is used need encapsulate voice packets, general encapsulation is to add IP packet header, User Datagram Protocol (UDP:User Data Protocol) packet header and real-time transport protocol (rtp) packet header before voice packets, and the packet header total length of adding up is 40 bytes.For the quality that ensures speech, the time delay of control speech, the unitary code stream that coding packing back forms is normally about 20kbit/s, and bandwidth reduction is about 3 times.After adopting technical finesses such as quiet inhibition and multichannel, the average code stream that transmits on the network is about 12kbit/s, and bandwidth availability ratio is about 5 times.Like this, if can run 5 road speeches under the shared bandwidth of former No. one phone, 1/5 before so long-distance cost will drop to, therefore user's long-distance telephone expenses also descend.

Voice packet is sent the employing Real-time Transmission by IP network, be the transmission on IP network behind the guarantee speech information process gateway compressed encoding, IP telephony system has adopted real-time transport protocol (rtp) and RTCP Real-time Transport Control Protocol (RTCP), these two kinds of agreements are prior art, do not repeat them here.

As shown in Figure 1 and Figure 2, it is the described a kind of composition structural representation that carries out voice communication system by IP network of the embodiment of the invention, suppose that user A, user B, user C three people set up a PTT group call (so-called group call, be meant two people or two above callings of participating in of people, as conference telephone etc., two people's calling also can be thought a kind of special group call), in this system, be provided with a PTT controller 21, it is control centre, it is a gateway, the forwarding of the voice packet during being used for being responsible for the PTT call establishment and conversing; And on user A, user B, user C three people's terminal, be equipped with the PTT button, this PTT button is used for the PTT business, voice just can send other people to during the user pressed the PTT button, unclamp can only listen behind the PTT button not talkative.

At first the embodiment of the invention is set up a calling before beginning between user A, user B, user C, and call establishment belongs to known technology, and how the embodiment of the invention sets up a calling if not relating to.Along with call establishment, user A, user B, user C three people can converse, and the embodiment of the invention specifically may further comprise the steps:

The first, after the call setup success, calling party's voice data is encoded, form voice packet, and in the packet header of this voice packet, insert the identifier that shows speech beginning or END instruction, as shown in Figure 3;

If user A wishes speech, it presses the PTT button request speech of terminal, its terminal session sound bag encapsulates, before voice packets except adding IP packet header, outside User Datagram Protocol (UDP) packet header and the real-time transport protocol (rtp) packet header, also insert the identifier that shows speech beginning or END instruction, it is the PTT identification field, present embodiment is assumed to be " 1 ", in this voice packets packet header except this PTT identification field, can also comprise user ID, other information such as sequence number of bag, as shown in Figure 1, wherein, background color is that the square frame of black is meant voice packet packet header among the figure, and background color is meant voice content for the square frame of white.First voice packet that user A sends can comprise voice content, also can not comprise voice content and has only packet header, so that allow system and other users know that A begins speech, has reduced the conflict that a plurality of users fight for speech simultaneously as early as possible.

The second, the voice packet that this is had an identifier is sent to control centre by IP network, and this control centre judges according to this identifier, decision beginning or finish calling party's speech;

The terminal of user A is sent to control centre with the voice packet that this has identifier by IP network, and promptly the PTT controller 21, and this PTT controller 21 judges whether to exist the current speaker after receiving this voice packet, and if there is no, user A begins speech so; Otherwise the priority that compares current speaker and user A is if calling party's priority height begins calling party's speech so.

Three, this control centre is transmitted to the side's talked about terminal with voice packet, finishes voice communication.

Suppose that not having other people at that time makes a speech, perhaps current spokesman's priority is lower than user A, and the voice packet that user A is sended over is transmitted to other users (user B and user C) in the current PTT group by IP network.

After the terminal of user B and user C received that PTT that user A sends is designated 1 voice packet, prompting user A made a speech, and this moment, user B and user C can not make a speech, unless the priority of user B and user C is higher than user A.

When user A finished speech, user A unclamped the PTT button, and the terminal of user A sends one or several PTT to the PTT controller and is designated 0 voice packet.

The PTT controller identifies the unmanned speech of current group after receiving that PTT that user A sends is designated 0 voice packet, allows other users' speeches in the group, and the voice packet that A is sent is transmitted to B and C simultaneously.

As shown in Figure 2, after the terminal of user B and user C received that PTT is designated 0 voice packet, prompting user B and user C: the speech of user A finished, if at this moment user B or user C wish that speech can press the speech of PTT button.

When the situation that a plurality of users request to make a speech simultaneously occurring, the PTT controller decides that user to have the right to make a speech according to the order of user speech bag arrival and user's priority, be that voice packet elder generation arrives the PTT controller or the higher user of priority will be allowed to speech, system is transmitted to other users with this user's voice, and arrival PTT controller or the lower user's voice bag of priority are abandoned by the PTT controller behind the voice packet.The prompting user failure of making a speech when the user terminal that does not obtain right to speak receives that PTT that other users send is designated 1 voice packet.

The embodiment of the invention provides a kind of method that realizes semiduplex voice communication on mobile communication or fixed communication IP network, and making originally to become possibility by PTT professional introducing mobile communication or the fixed communication IP network that special-purpose group system or intercom provide; It has defined a kind of packing manner that is used for the media stream of half-duplex operation in the ip voice communication, its applied environment is not limited only to public mobile network (as CDMA, GSM, WCDMA etc.), also can be that WLAN (WLAN (wireless local area network)) waits other mobile communications networks and fixed communication network (as fixed LAN, Internet etc.).

Claims (3)

1, a kind of method that realizes half duplex IP voice communication is characterized in that the method includes the steps of:

After a, the call setup success, calling party's voice data is encoded, form voice packet, and in the packet header of this voice packet, insert the identifier that shows speech beginning or END instruction;

B, this voice packet that has identifier is sent to control centre by IP network, this control centre judges according to this identifier, decision beginning or finish calling party's speech;

C, this control centre are transmitted to the side's talked about terminal with voice packet, finish voice communication.

2, a kind of method that realizes half duplex IP voice communication as claimed in claim 1, it is characterized in that, judge specifically according to this identifier among the described step b and be meant: if this identifier is 1, begin calling party's speech so, otherwise finish calling party's speech.

3, a kind of method that realizes half duplex IP voice communication as claimed in claim 1 is characterized in that, beginning calling party's speech also further comprises among the described step b:

Judge whether to exist the current speaker, if there is no, begin calling party's speech so; Otherwise the priority that compares current speaker and calling party is if calling party's priority height begins calling party's speech so.

Images