WO2000005918A1 - Method for determining packets of data contained in a continuous data flow - Google Patents

Abstract

The data packets (SE) have a respective cell head with a constant length l, consisting of cell-head data (CID, LI, UUI) and a terminating cell-head control sum (HEC). A test unit (PE) divides the final data l from the continuous data flow (TDM) received by a generator polynome used as a basis for calculation of the cell-head control sum (HEC) from the cell-head data (CID, LI, UUI) and when the remainder of the division is zero, a timing mark (Z) indicating the beginning of a data packet (SE) is set.

Description

description

Method for determining data packets contained in a continuous data stream

Due to the increasing need for transmission of video information in modern communication technology, e.g. Fixed and moving images in videophone applications, or the display of high-resolution graphics on modern DV systems, the importance of transmission and switching techniques for high data transmission rates (greater than 100 Mbit / s) is increasing. A known data transmission method for high transmission bit rates is the so-called Asynchronous Transfer Mode (ATM). Data transmission based on the asynchronous transfer mode currently enables a variable transmission bit rate of up to 622 Mbit / s.

In the transmission technology known as asynchronous transfer mode (ATM), data packets of fixed length, so-called ATM cells, are used for the data transport. An ATM cell is composed of a five-byte long cell header containing the switching data relevant for the transport of an ATM cell, the so-called ^Λ header 'and a 48-byte long user data field, the so-called ^ payload'. In this case, in the user data field of an ATM cell, only one logical connection - often referred to in the literature as ^ Virtual Channel 'VC or ATM channel - is transmitted.

In the German patent application with the file number 198 187 76.9, a method has already been proposed by which a transmission of data assigned to different logical connections in the useful data area of one or more ATM cells is made possible. For this purpose, so-called substructure elements with a variable 0 to 64 byte useful data field are defined in the user data area of one or more ATM cells in accordance with the known ATM adaptation layer AAL type 2, using an address field in the Cell header of the substructure element is assigned to a logical connection. By means of the so-called ATM adaptation layer AAL - often referred to in the literature as ^λ ATM adaptation layer - the ATM data format is adapted to this - often referred to in the literature as ^Λ ATM layer (layer 2) Network layer (layer 3) according to the OSI reference model (Open Systems I_connection). In addition to the subdivision of the user data area of an ATM cell into substructure elements, the first byte of the user data area of an ATM cell is defined as a pointer that points to the first substructure element arranged in the user data area, ie it is the start address of the first in User data area located cell header of a substructure element stored. Using this pointer, it is possible to restore the synchronization between transmitter and receiver in the event that one or more ATM cells have been lost, for example due to a transmission error.

Furthermore, patent application GR 98 P 2136 (internal file number), which has the same priority, suggests switching ATM cells, which are subdivided into substructure elements according to the ATM adaptation layer AAL type 2, by means of an ISDN switching system by means of an arrangement arranged therein to implement time slot-oriented switching matrix module. For this purpose, the packet-oriented ATM data format in accordance with the ATM adaptation layer AAL type 2 is converted to a time slot-oriented data format in accordance with the TDM method (Time Division Multiplex). In the context of the time slot-oriented data format according to the TDM method, 32 user channels are available for data transmission, which are designed as ISDN-oriented B channels (Integrated Services Digital Network) with a transmission bit rate of 64 kbit / s each.

The object of the present invention is to specify a method by which the detection of boundaries between individual data packets for an extraction from into a continuous Inserted data packets is enabled.

The object is achieved according to the invention with the features of patent claim 1.

For a better understanding of the packet-oriented ATM data format according to the ATM adaptation layer AAL type 2, it seems necessary to first go into known principles again.

In the case of data transmission in the context of the packet-oriented ATM data format in accordance with the ATM adaptation layer AAL type 2, it is possible to subdivide the useful data area of an ATM cell into substructure elements.

A substructure element according to the ATM adaptation layer AAL type 2 is composed of a 3-byte long cell header and a user data area of variable length (0 to 64 bytes). The cell header of a substructure element is subdivided into an 8-bit channel identification CID (Channel I_dentifier), a 6-bit length identification LI (Length Indicator), and a 5-bit transmitter-receiver identification UUI (User-to -User Indication) and a 5 bit Ze11kopf checksum HEC (Header Error Control).

The cell head checksum HEC - often referred to in the literature as a block check string - is determined in a so-called cyclic block check - often referred to in the literature as code-independent error protection and abbreviated as CRC (Cyclic Redundancy Check) - and is used to check the correctness of the first 19 bits transmitted in the cell header of a substructure element.

In the cyclic block check, the binary characters of the data block to be saved - i.e. the first 19 bits transmitted in the cell header of a substructure element - serve as a coefficient. of a so-called basic polynomial. For example, a basic polynomial "x ¹⁸ + x ¹² + x ⁷ + x ³ + 1" results from a data block "1000001000010001001". This basic polynomial is divided by a so-called generator polynomial "x ⁵ + x ² + 1" (standardized by ITU-T 1.362) modulo-2. The remainder remaining after the division corresponds to the cell head checksum HEC.

An essential advantage of the method according to the invention is that it enables data packets inserted into a continuous data stream to be recovered without transmitting additional information which marks the beginning of a data packet.

Advantageous developments of the invention are specified in the subclaims.

An advantage of embodiments of the invention defined in the subclaims is, inter alia, that by inserting filler cells or filler data into a substructure element when converting a packet-oriented data format into a time slot-oriented data format, a mediation of compressed data (with a variable transmission bit rate) is possible without previous decompression. This prevents a loss of quality when transmitting compressed data.

An embodiment of the invention is explained below with reference to the drawing.

Show:

1: a structural diagram for the schematic representation of the conversion of the packet-oriented ATM data format into the time slot-oriented TDM data format according to a first conversion mode; 2 shows a structural diagram for the schematic representation of the conversion of the packet-oriented ATM data format into the time slot-oriented TDM data format according to a second conversion mode; 3: a structural diagram for the schematic representation of a TDM channel with associated time stamps.

1 shows a schematic illustration of a conversion of the packet-oriented ATM data format according to the ATM adaptation layer AAL type 2 (ATM adaptation layer) into the time slot-oriented data format according to the TDM method (Time Division Multiplex) according to a first conversion mode of a conversion unit UE. Data is transmitted in the context of the packet-oriented ATM data format via ATM cells ATM-Zl, ATM-Z2. An ATM cell ATM-Zl, ATM-Z2 is composed of a cell header H which contains the switching data relevant for the transport of an ATM cell ATM-Zl, ATM-Z2 and a 48-byte useful data field. In the case of data transmission in the context of the packet-oriented ATM data format in accordance with the ATM adaptation layer AAL type 2, it is possible to subdivide the useful data area of an ATM cell ATM-Zl, ATM-Z2 into substructure elements SE.

A substructure element SE according to the ATM adaptation layer AAL type 2 is composed of a 3-byte long cell header and a user data area I of variable length (0 to 64 bytes). The cell header of a substructure element SE in turn is subdivided into an 8-bit channel identification CID (Channel Identifier), a 6-bit length identification LI (Length Indicator), a 5-bit transmitter-receiver identification UUI ( User-to-User Indication) and a 5 bit long cell header checksum HEC (Header Error Control).

By dividing an ATM connection with the aid of substructure elements SE into individual data streams that are independent of one another, as in the example of the ATM cells ATM-Zl, ATM-Z2 in of the figure, several can within an ATM connection based on the 8-bit channel identification CID

(256) logical connections are addressed, all with the same ATM address - consisting of a VPI value

(Virtual Path Identifier) and a VCI value (Virtual Channel I_dentifier) - can be addressed. There is also the possibility of defining a substructure element SE for the transmission of signaling information assigned to the logical connections.

In the figure, for example, four different substructure elements SE are shown, which are defined on the basis of different channel identifications CID in the cell header 0, ..., 3 of the substructure elements SE. The 6-bit length identification LI in the cell header of a substructure

Elements SE a user data field I of variable length (0 to 2 ° byte) can be defined, so that data transmission with variable transmission bit rate can be realized for the different logical connections.

For a conversion of the packet-oriented data format according to the ATM adaptation layer AAL type 2 to the time slot-oriented data format according to the TDM method, each substructure element SE of an ATM cell ATM-Zl defined for the transmission of user data is ATM-Z2 a TDM channel

K0, ..., K3 assigned to the time slot-oriented data format according to the TDM method. An assignment of a substructure element SE to a TDM channel K0, ..., K3 takes place in a signaling phase preceding the user data transmission. There are generally 32 user channels available for data transmission in the context of the time slot-oriented data format in accordance with the TDM method, which are designed as ISDN-oriented B channels with a constant transmission bit rate of 64 kbit / s each.

As part of the conversion of the packet-oriented data format according to the ATM adaptation layer AAL type 2 to the time slot-oriented data format in accordance with the TDM method must additionally adapt the - possibly variable - transmission bit rate of the packet-oriented data format resulting from the size and arrival of substructure elements SE to the constant transmission bit rate of 64 kbit / s of the time slot oriented data format. This is achieved in the context of the first conversion mode of a conversion unit UE by inserting so-called filler cells FZ of variable length into the continuous TDM data stream.

In the conversion unit UE, substructure elements SE received and packaged in ATM cells ATM-Zl, ATM-Z2 are unpacked via a packet-oriented data transmission link. Subsequently, so-called filler cells FZ are converted to the substructure containing the useful data for the conversion of the - possibly variable - transmission bit rate resulting from the size and arrival of the substructure elements SE to the constant transmission bit rate of 64 kbit / s of the time slot-oriented data format. SE elements added. The length of a fill cell FZ is determined by a so-called fill cell header FZH. The length of a fill cell FZ is chosen so that the total transmission bit rate of a substructure element SE and a fill cell FZ results in an integer multiple of 64 kbit / s. If the transmission bit rate of a substructure element SE is greater than 64 kbit / s - that is, greater than the transmission bit rate of a TDM channel K1, ..., K4 - the user data transmitted in a substructure element SE are transferred to several TDM channels K1,. .., K4 divided.

Finally, these data (substructure elements SE and fill cells FZ together) are assigned to a TDM channel K0,..., K1 agreed in the signaling phase of a time slot-oriented data transmission link and forwarded on. 2 shows a schematic illustration of a conversion of the packet-oriented ATM data format according to the ATM adaptation layer AAL type 2 (ATM adaptation layer) into the time slot-oriented data format according to the TDM method (Time Division Multiplex) according to a second conversion mode of the conversion unit UE.

In contrast to the first conversion mode, no separate filler cells FZ are inserted in the continuous TDM data stream in the second conversion mode. An adaptation of the - possibly variable - transmission bit rate of the packet-oriented data format resulting from the size and arrival of the substructure elements SE at the conversion unit UE according to the ATM adaptation layer AAL type 2 to the constant transmission bit rate of 64 kbit / s of the time slot-oriented data format according to the TDM method is carried out by filling the substructure elements SE with filler data FD, so that the total transmission bit rate of a substructure element SE - consisting of user data and filler data FD - is an integer multiple of 64 kbit / s results.

3 shows a schematic illustration of a TDM channel TDM with substructure elements SE packed therein. The octet-oriented TDM data stream of the TDM channel passes through a test unit PE, in which a test - in the following with

Designated cell header check - the received data is carried out for a specific bit sequence which characterizes the start of a substructure element SE. This cell head check is carried out on the basis of the cell head checksum HEC transmitted in the cell head SEH of a substructure element SE.

As already described, the binary characters of a data block to be secured - in this case the first 19 bits transmitted in the cell header SEH of a substructure element SE - are used as coefficients of a so-called basic polynomial as part of a cyclic block check. This basic polynomial is called a generator polynomial "X ⁵ + x ² + 1" (standardized by ITU-T 1.362) divided modulo-2. The remainder remaining after the division corresponds to the cell head checksum HEC.

For the cell head test, the last three of the

Test unit PE received bytes in a shift register - often also referred to as a FIFO (First In First Out) memory

- cached and divided by the generator polynomial modulo-2. If the division results in the remainder of zero, the 3 temporarily stored bytes are the characteristic bit sequence searched for and a time mark Z which is synchronous with the beginning of the cell header is set. Then the byte first stored in the shift register is deleted from the shift register and a new byte - the last one received at the test unit PE - is written into the shift register for a new division.

This characteristic bit sequence is repeated with each cell head SEH of a substructure element SE. There is also the possibility that the characteristic bit sequence occurs anywhere in the continuous data stream due to a random bit configuration. In order to be able to suppress such randomly occurring bit constellations when recognizing the start of a substructure element SE - that is, the cell header SEH of a substructure element SE - an additional check is carried out in addition to the cell header check

- hereinafter referred to as distance check - carried out.

As in the case of a conversion of the packet-oriented ATM data format into the time slot-oriented TDM data format in accordance with the second conversion mode of the conversion unit UE, only substructure elements SE with the same length occur - the length of a substructure element SE is filled up with Filling data FD adapted to the required constant transmission bit rate of a TDM channel - so the scope of the distance check merely checks whether the characteristic bit sequence occurs at the same distance. If the character teristic bit sequence three times in a row at the same distance, the detection is considered synchronous. If a characteristic bit sequence occurs at a distance that deviates from the distance interval, the detection becomes asynchronous. The recognition is synchronized again after recognizing the characteristic bit sequence three times at the same distance. With the aid of the length identification LI transmitted in the cell header SEH of such a substructure element SE, the transmitted user data can be separated from the filler data FD.

As in the case of a conversion of the packet-oriented ATM data format into the time slot-oriented TDM data format according to the first conversion mode of the conversion unit UE, substructure elements SE with different lengths occur - the lengths of the respective substructure elements SE and Lengths of the respective filler cells FZ can be different - the distance check is carried out taking into account the length identification LI transmitted in the cell header SEH of a substructure element SE. If the characteristic bit sequence occurs three times in succession at the distance specified by the length identification LI, the detection is considered to be synchronous. If a characteristic bit sequence occurs at a distance that is not predetermined by the length identification LI, the detection becomes asynchronous. The recognition becomes synchronous again after recognizing the characteristic bit sequence three times at the distance specified by the length identification LI.

Claims

claims 1. Method for recognizing data packets (SE) transmitted successively in a continuous data stream (TDM), the data packets (SE) each having an m-bit long cell header (SEH), which is composed of an n-bit long cell header data field (CID, LI, UUI) and a p-bit long cell header checksum (HEC), whereby the last m bits received from the continuous data stream (TDM) are checked by a test unit (PE) for a predefinable correlation between the first n bits and the subsequent p Bits are checked, and in cases in which this correlation is recognized, synchronization information (Z), which marks the beginning of a data packet (SE) and is assigned to the last m bits, is generated. 2. The method according to claim 1, characterized in that for a subsequent check, one of the last m bits first received at the test unit (PE) q-bit group separated from the m bits and one last received at the test unit (PE) q -Bit group is newly appended. 3. The method according to any one of the preceding claims, characterized in that for a check of the last m bits that were received from the continuous data stream (TDM) at the test unit (PE), these m bits by one, the creation of the cell head checksum ( HEC) from the generator polynomial on which the cell header data (CID, LI, UUI) are based, and in cases in which the division remainder is zero, the synchronization information (Z) is generated. 4. The method according to claim 3, characterized in that the last m bits received from the continuous data stream (TDM) are temporarily stored in a shift register, and that for a subsequent division, the first of the stored last m bits received at the test unit (PE) q-bit group is shifted out of the shift register and the q-bit group last received at the test unit (PE) is reinserted. 5. The method of claim 2 or 4, d a d u r c h g e k e n n z e i c h n t that in cases where the length of the cell header (SEH) is an integer multiple of 8 bits, the q-bit group has a length of 8 bits. 6. The method according to any one of the preceding claims, characterized in that in cases in which the data packets (SE) have a constant length, the last generated synchronization information (Z) is only evaluated as valid if the last x synchronization information (Z) in the same time interval was generated. 7. The method according to any one of the preceding claims 1 to 5, characterized in that in cases in which the data packets (SE) have a variable length predetermined by a length identification (LI) in the cell header data, the last generated synchronization information ( Z) is only evaluated as valid if the last x synchronization information (Z) was generated in the time intervals specified by the respective length identifications (LI). 8. The method of claim 6 or 7, d a d u r c h g e k e n n z e i c h n e t that the last 3 synchronization information (Z) are provided as the last x synchronization information (Z).