DE19832997C2 - Network adaptation device and associated method - Google Patents

Description

The invention relates to a network adaptation device like an associated method for converting a first da tenstroms from a first network in at least one two data stream for a second network. In particular relates the invention relates to data streams that have structural elements contain compressed data.

Network adaptation devices (Interworking Units, IWU) connect two or more, usually different types Networks with each other. Examples of networks are local ones Networks (Local Area Network, LAN), networks for the area of a Metropolitan Area Network (MAN) or Weitver wide area network (WAN). Networks can Serve to transfer data between computer systems (e.g. Computer buses, ethernet, appletalk, internet). It can ever voice data can also be transmitted (e.g. analogue and di digital telecommunications, mobile communications, ISDN, digital broadcasting funk) and image information (e.g. video-on-demand, MPEG, digital television, videophone). Networks can be different have topologies, for example an irregular one meshed network, ring structure, bus topology or star arrangement nung. In addition, networks can have a variety of over use techniques such as analogue and digi tale, serial and parallel, synchronous and asynchronous transmission. Examples of transmission techniques represent the time division multiplex method TDM) and the asynchronous transfer mode (Asynchronous Transfer Mode, ATM). The transfer media can be the By air (e.g. terrestrial radio networks, satellite transmission cable), twisted cables (e.g. for conventional telephone networks),  Coaxial cable (e.g. for cable television, Ethernet) or glass fiber be used. In the information transfer sometimes also parallel cables with different Dielek trika and shielded cables are used.

Network adapters receive data from one Network and transfer this data to a second network ter. This requires that the network adapter be on both networks regarding the underlying technology fits. This includes, for example, in the case of digital data adjustment to those used in the networks Transmission protocols and bit rates. In addition to naming, Addressing and switching tasks necessary for a correct Delivery of the data to the appropriate addressee not Controllable control of the data must also be maneuverable flow through backflow and buffering if the networks have different bit rates. If the networks have different maximum packet sizes, then must the Network adaptation device also a segmentation and Perform packet reshaping.

Examples of well-known, frequently used network adaptation facilities are repeaters that are two networks in the first Connect OSI layer, bridges in the second layer, router in the third layer and gateways in higher layers. Other network adaptation facilities are in the field of telecommunications in switching systems Men used to connect different telephone networks to each other connect.

The better utilization required for the transmission links has led to the development of multiplexing, from two forms of which have emerged: frequency multiples  plex (FDM) and time division multiplex (TDM). Both methods are Summary of primary signals in a multiplexer, the single-channel transmission of the multiplex signal and the Rückge Primary signals are extracted by demultiplexers or time filters common. Each message signal occupies in frequency division multiplex a frequency band of a certain width, the basebands of the Primary signals through modulation techniques with staggered Carrier frequencies thus shifted to higher frequency ranges the one that they lie side by side on the frequency scale come. Time division multiplexing can be binary and continuous Signals are transmitted. Continuous Si signals, such as the Speech, before transmission by scanning into fast fol very short pulses (pulse code modulation, PCM).

One example is a network that uses PCM techniques the ISDN network (Integrated Services Digital Network). At ISDN, which was specified according to the OSI layer model, is a digital network that different Services integrated, such as B. Telephone, fax, Datex, Multi media (video, audio, data), and the signaling regardless of the form and content of communication leads. ISDN has separate channels for the transmission of User information (B channel) and for signaling (D- Channel). The international standardization of ISDN is through the ITU-T (International Telecommunication Union- Telecommunication Standardization Sector).

A data transport technology based on a very fast Pa chain switching is the asynchronous transfer mode (ATM), which is also the basis for broadband ISDN (B-ISDN) poses. ATM only uses packets for transmission a fixed length of 53 bytes. This smallest indivisible  Transmission unit is therefore referred to as an ATM cell. A ATM cell consists of a 5-byte header and a user data field of 48 bytes, the header under an which a channel identification (VPI), a path identification (VCI) and a PT field (Payload Type) that includes the type of Indicates user data.

Similar to the OSI reference model, the ATM model is also verti Kal divided into several layers, for example the Physical layer, the ATM layer, the ATM Adaptation layer (AAL) and higher layers for control and for user data. It also becomes a horizontal limb according to aspects of data exchange between users (User Plane), aspects of communication control (Control Plane) and management aspects (Management Plane).

The quality of a (virtual) ATM channel is guaranteed by the rendered service class. To define the four Service classes A to D cover only the aspects of Timeliness QoS is taken into account when implementing Parame ter are provided which, for example, the cell loss rate, the cell delay and the fluctuation of the cell delay, so describe the jitter. The requirement profiles possible Applications are based on their bit rate, the time relationship hung between source and sink and the type of connection distinguished. For example, the voice transmission is a typical application of service class A. Examples of Class B applications are video transmissions in MPEG-2 format and voice transmission in mobile networks. The Service classes C and D are used to implement conventional ones Data services such as file transfer.

The ATM adaptation layer (AAL) forms the interface the layers based on ATM. Structure your tasks themselves in the implementation of those required by higher classes  Service parameters according to service classes A to D and the Segmentation on the transmitter side and reassembly on the receiver side user data. According to the recommendations of the ITU, ver different AAL types for the different service classes and Applications provided.

In AAL type 2 , data transmission according to service class B was originally to be implemented. According to ITU-T I.363.2, AAL type 2 is now a multiplexing method for several sources that are transmitted via a common virtual channel connection (VCC). The new AAL Type 2 can be used for bandwidth-efficient transmission of low-rate, short packets for delay-sensitive applications. An important example of such traffic is the transmission of compressed speech, as used in mobile networks. Since even the packetization delay for a normal 64 kbps voice channel can lead to echo and delay problems, this situation becomes even more critical when the speech is compressed with a lower bit rate (e.g. with 8 kbps).

Filling an 8 kbit per second ATM cell requires about 48 ms. One way to reduce this fill time it would be to only partially fill the ATM cell. This would however make very inefficient use of networks represent resources.

For example, is from the German published application DE 196 81 138 T1 a method for transmitting kompri mated voice data known, in which a bidirectional order Conversion between an ATM data stream and a TDM data current occurs. Contain the data packets of the ATM data stream only data from one channel of the TDM data stream, see above that the filling time for such a package is relatively long,  which results in poor quality of the speech data to be transmitted results.

The AAL type 2 now enables the multiplexing of several sources of low bandwidth in ATM cells of a single virtual connection. This considerably reduces the packetization delay, since, for example, filling an ATM cell with two 8 kbit / s streams requires only 24 ms.

The AAL type 2 is subdivided into a CPS (common part sublayer) and an SSCS (service specific con vergence sublayer). The multiplexing of several channels according to AAL type 2 takes place in the CPS sub-layer and is carried out on the basis of so-called CPS packets. In FIG. 2, the CPS packet format is shown for the AAL type 2. The CPS package consists of a header (CPS-PH) and a section for the user data (CPS-PP). The header includes a channel identifier (CID) 50 to identify the various packets. The length of the user data part (CPS-PP) is contained in a length indicator (LI) 52 . The headers also include further fields 54 , 56 . The payload contained in the CPS package can have a length of up to 45 octets or up to 64 octets, depending on the information in LI field 52 .

An important application for AAL type 2 is the transmission of compressed voice data, for example in mobile networks. Speech compression is described, for example, in ITU recommendation G.728. Such methods use prediction algorithms such as the LD-CELP algorithm (Low-Delay Code Excited Linear Prediction). Encoders and decoders using these algorithms use a codebook and only transfer an index to this codebook. In order to reduce the delay time, the algorithms apply a backward adaptation of synthesis filter coefficients and gain factors. The use of speech vectors and codebooks in particular implies that the original data cannot be restored in an identical manner by decoding a coded signal. The deviations can be so large that they can be heard and perceived as disturbing.

Conventional network adapters that compress When data is received from a network, the data is decompressed before being fed into another network, to the data to adapt to the transmission rate required there. At a reverse conversion in another network adaptation the data is then compressed again in the direction. This is in any case at least before handover to the final participant always necessary. With this procedure, the Sprachqua lity with any additional Decompression and compression worsened. In addition the transmission delay times increase.

The invention is therefore based on the object of a network adaptation device and an associated method ben, the heightened quality of the transmitted language possible.

This object is achieved by the method according to claim 1 and the network adapters according to the patents sayings 13 and 14 solved.

Preferred embodiments are the subject of the subclaims.

In particular, the invention provides network adaption and a related procedure in which a Decompression and compression is avoided. The compress Data that contain the user data remain unaffected, as long as they do not have to be processed by yourself (e.g. at Conferences). This has the advantage that during the Transmission delay times are kept as low as possible the.

Preferred embodiments of the invention are next explained. Show it:  

Fig. 1 is a data flow flowchart to show how he inventive operation of a network adaptation device; and

Fig. 2 shows the structure of a CPS packet.

In Fig. 1, the operation of a preferred exemplary embodiment of the invention is shown. The network adapter (IWU) 2 receives a data stream 4 , which is shown as VCC. In the preferred exemplary embodiment of FIG. 1, the data stream 4 is composed of a sequence of ATM cells 16 , each of which is led by an ATM cell header 18 . The respective ATM cell header 18 is followed by a series of AAL2 structure elements 20 , each of which comprises a header 22-28 and a user data field 30-44 . Since different AAL2 structure elements 20 of an ATM cell form 16 different subchannels. An AAL2 structure element 20 is preferably assigned to a subchannel by means of a corresponding specification in the CID field 50 of the respective header 22-28 .

The network adapter 2 assigns each subchannel its own channel 6-14 on the output side. A header 22-28 followed by the respective compressed data 30-44 is transmitted in each of these channels 6-14 . For the adaptation to the transmission rate required in the output network, leo octets 46 are also inserted into the user data fields. As a result, the structural element on the output side is correspondingly lengthened.

When inserting empty information into the cells, the header 22-28 is preferably modified accordingly, so that the original information can be easily retrieved from the extended cells at a later point in time.

The changes necessary for this conversion and recovery tion of the header is preferably in the course of a usual Header exchange when passing through a switch, e.g. B. in egg nem switching network of a switching center.

When changing the header, the LI field 52 and the UUI field 54 are preferably taken over, the separation of the compressed useful data 30-44 from the inserted empty octets being made possible in particular by changing the LI field 52 .

In a preferred embodiment of the invention, two mobile radio subscribers communicate, including the ISDN networks. Compressed voice data are transferred from a mobile radio subscriber to the network adaptation device 2 in a subchannel. The user data transmission rate is 80 kbps, for example. If there are about five subchannels, each subchannel takes up 16 kbps. On the output side of the network adaptation device 2 is the ISDN network with a row of B channels, each of which has 64 kbps. Since these channels are TDM channels in this example, a constant data stream is necessary in them.

On the output side, a se is preferably used for each data stream Parater TDM channel used. It can, for example Video transmission, but also several TDM channels are used the.

A network adaptation according to the invention is preferred device can be implemented as a hardware solution. A network  However, the adaptation device can also be implemented as software be mented.

Furthermore, the network adaptation according to the invention can be towards a single, separate and independent Ge advises, but it can preferably also in a ver system or a private branch exchange can be integrated.

In addition to the conversion of a data stream shown in FIG. 1 into several data streams, wherein leo octets are inserted into the transmitted cells, a network adaptation device for the reverse conversion is also provided according to the invention.

Claims (27)

1. Method for converting a first data stream ( 4 ) formed from a sequence of data packets from a first network into at least one second data stream (6-14) formed from a periodic sequence of second structural elements for a second network and / or for converting back the second data stream into a data stream corresponding to the first data stream, characterized in that
that a data packet of the first data stream ( 4 ) consists of a sequence of first structural elements ( 20 ) assigned to different second structural elements, and
that the second structural elements assigned to a period are each assigned data of the first structural elements ( 20 ) in unchanged form and data carrying empty information. 2. The method according to claim 1, wherein the first structural elements ( 20 ) are embedded in ATM cells ( 16 ) AAL2 structural elements. 3. The method according to claim 1 or 2, wherein the first data stream ( 4 ) carries a plurality of subchannels and data in first structure elements ( 20 ) of different subchannels correspond to those in various second data streams ( 6-14 ). 4. The method according to any one of claims 1 to 3, wherein each of the second data streams ( 6-14 ) is transmitted in a separate TDM channel. 5. The method according to any one of claims 1 to 4, wherein the second network is an ISDN network. 6. The method according to any one of claims 1 to 5, wherein the first data stream ( 4 ) and the second data stream ( 6-14 ) have different bit rates. 7. The method according to any one of claims 1 to 6, wherein he Most and second structure elements have headers that be changed when converting. 8. The method of claim 7, wherein the change in the header is a change in length information ( 52 ). 9. The method according to any one of claims 1 to 8, wherein the data contained in the first data stream ( 4 ) are speech signals. 10. The method according to any one of claims 1 to 9, wherein it network is a mobile network. 11. The method according to any one of claims 1 to 10, wherein the Procedure is software controlled. 12. The method according to any one of claims 1 to 10, wherein the Procedure is carried out under hardware control.   13. Network adapter ( 2 ) with connections to a first and a second network, with
Means for receiving a first data stream ( 4 ) from the first network formed from a sequence of data packets, and
Means for sending into the second network a second data stream (6-14) formed from a periodic sequence of second structural elements,
characterized,
that a data packet of the first data stream ( 4 ) consists of a sequence of first structural elements ( 20 ) assigned to different second structural elements, and
that means for assigning first structural elements ( 20 ) in unchanged form and data carrying empty information to the second structural elements assigned to a period are provided. 14. Network adaptation device ( 2 ) with connections to a first and a second network, with
Means for receiving a second data stream ( 6-14 ) formed from a periodic sequence of second structure elements from the second network, the second structure elements each consisting of useful data and empty data, and
Means for sending a first data stream ( 4 ) formed from a sequence of data packets into the first network,
characterized,
that a data packet of the first data stream (4) follows from a From consists of first, different second structural elements associated structure elements (20), wherein the first structural elements (20), the Dutzdaten the second Strukturele elements in unchanged form included. 15. The network adapter according to claim 13 or 14, wherein the first structural elements ( 20 ) are embedded in ATM cells ( 16 ) AAL2 structural elements. 16. Network adaptation device according to one of claims 13 to 15, wherein the first data stream ( 4 ) has a plurality of subchannels and the data contained in first structure elements ( 20 ) of different subchannels correspond to those in different second data streams ( 6-14 ). 17. Network adapter according to one of claims 13 to 16, wherein each of the second data streams ( 6-14 ) is transmitted in a separate TDM channel. 18. Network adaptation device according to one of claims 13 to 17, the second network being an ISDN network. 19. Network adaptation device according to one of claims 13 to 18, wherein the first data stream ( 4 ) and the second data stream ( 6-14 ) have different bit rates. 20. Network adaptation device according to one of claims 13 to 19, the first and second structural elements over Have headers and the means to create means to change the header include. 21. Network adaptation device according to claim 20, wherein the means for changing the header means for changing a length information ( 52 ). 22. Network adaptation device according to one of claims 13 to 21, wherein the data contained in the first data stream ( 4 ) are speech signals. 23. Network adaptation device according to one of claims 13 to 22, the first network being a mobile network. 24. Network adaptation device according to one of claims 13 to 23, the means for receiving, generating and transmitting are realized by software. 25. Network adaptation device according to one of claims 13 to 23, the means for receiving, generating and transmitting are realized by hardware. 26. Network adaptation device according to one of claims 13 to 25, wherein the network adaptation device ( 2 ) is designed as an independent device. 27. Network adaptation device according to one of claims 13 to 25, wherein the network adaptation device ( 2 ) is part of a switching system.