KR100428683B1 - Apparatus and method for boundary identification of received frame of GFP - Google Patents

Abstract

An object of the present invention is to provide a device for identifying a boundary of a GFP received frame and a method thereof. The present invention provides a counter for outputting GFP (Generic Frame Procedure) frame data by 4 bytes, and each of the 4 bytes divided by the counter. In this case, a plurality of HEC (Header Error Check) calculation units respectively performing header error control, an OR operation unit for generating a motor call signal by performing an OR operation on the outputs of the plurality of HEC calculation units, and the OR operation unit from the OR operation unit. A frame counter that outputs a length field value to determine the boundary position of the next frame when the motor call signal is input and the synchronization is correct, and identifies the position as the boundary of the next frame when the length field value reaches a specific value; Connected to the frame counter, and always fix the position of the frame check sequence (FCS) field of the GFP to the same value To provide the necessary signal to the frame counter and the border consisting of the decrementing counter and outputting it to the frame counter when reaching a specified value to count down the length field value received from the frame counter identifying unit; By including a descrambler that performs descrambling on the GFP frame data, in consideration of the variable frame length of the GFP, in identifying the boundary, the boundary of the frame is identified using the length field and four headers are identified. By performing the error control calculation, it is possible to minimize the time required to identify the boundary of the frame.

Description

Apparatus and method for boundary identification of received frame of GFP}

The present invention relates to a device for identifying a boundary of a Generic Frame Procedure (GFP) received frame and a method thereof. In particular, in consideration of a variable frame length of a GFP, a frame boundary is identified using the length field in identifying the boundary. The present invention relates to an apparatus and method for identifying a boundary of a GFP received frame, which is suitable for minimizing a time required for identifying a frame boundary by performing four header error control (HEC) calculations.

In general, GFP is a technology that transmits the signal of the data link layer, which is the second layer of the OSI-7 (Open System Interconnection) layer, to a physical layer by mapping it to a specific format regardless of its type. The corresponding payload is Ethernet ( It is a signal of a data link layer such as Ethernet, High-Level Data Link Control (HDLC), and Packet Over SONET (POS).

The biggest advantage of this GFP is its low overhead. In particular, there is no special overhead for identifying the frame boundary at the receiving side.

1 is a block diagram of a general GFP processing apparatus.

Here, reference numeral 1 denotes a multiplexer that receives an encapsulated signal and an idle insertion signal and multiplexes the converted signal into a GFP frame, and 2 denotes a GFP frame output from the multiplexer 1. Scrambling is a scrambler that performs scrambling, and 3 is an HEC generation unit that performs HEC generation on the GFP frame output from the multiplexer 1.

Also, reference numeral 4 is a descrambler that descrambles the Synchronous Payload Envelope (SPE) mapped octet stream, and 5 is a HEC check on the SPE mapped octet stream. 6 is a HEC check unit for performing a decapsulation operation by demultiplexing the descrambled signal in the descrambler 4 and removing an idle signal at an Idle termination. It is a demultiplexer.

Thus, as shown in FIG. 1, all payloads of the second layer in the OSI-7 layer are converted into a GFP frame having one format through encapsulation. After that, the scrambling process is performed. In this case, the header error checking (HEC) is excluded from the scrambling.

Since the physical layer requires a continuous flow of data, when there is no payload to send, the frame in the form of an idle is sent.

The receiver checks the unscrambled HEC to identify the frame boundary. If the boundary identification is successful and synchronized, the payload portion is descrambled and then decapsulated to generate and transmit a second layer signal again.

In the case of GFP, there is no separate field for identifying a frame boundary. In the case of HDLC, boundary identification is performed by leaving a flag field, but in the case of GFP, the boundary identification method described in ITU-T I.432 is basically used without leaving a flag-like field.

FIG. 2 is a block diagram of a general asynchronous transmission cell, FIG. 3 is a block diagram of a general Ethernet frame, and FIG. 4 is a block diagram of a general HDLC frame.

Thus, the ITU-T I.432-based method is used for cell-based Asynchronous Transfer Mode (ATM). The interface structure consists of a continuous flow of 53-byte cells.

Cell-based asynchronous transmission (ATM) requires some form of synchronization because no external frames are added.

Synchronization is based on the Cell Header Error Control (HEC) field.

4, in the case of an Ethernet frame, it can be seen that there are 7 bytes of preamble and 1 byte of boundary identification byte. 5, it can be seen that in the case of HDLC, a flag byte is used for boundary identification of a frame.

5 is a state transition diagram for cell boundary identification in the conventional asynchronous transmission method.

It consists of a hunting state, a pre-sync state and a sync state.

Here, the basic state of cell boundary identification is a hunting state. That is, the hunting state is a state in which cell boundaries are not identified. In this state, the boundary identification algorithm compares the header error control (HEC) with the calculated header error control (HEC). If the correct header error control (HEC) is found, it is assumed that the header has been found to enter the motor state.

In the motor state, the cell structure is estimated. In other words, the cell of the Asynchronous Transfer Mode (ATM) has a fixed size (53 bytes), so the position of the next cell can be known. Then, the header error control (HEC) of the cell is checked again. If the same process is repeated n times, and the header error control (HEC) is correct in the process, the boundary identification state enters a synchronous state.

At this time, if the header error control (HEC) is wrong even in one of the n inspection processes, the boundary identification state is transferred to the hunting state.

In the synchronous state, it is assumed that cell boundary identification has been performed, and the header error control (HEC) at this time is used for error detection of the cell header, not for boundary identification. In addition, if the header error control (HEC) is continuously wrong more than m times, it is determined that the boundary identification information is lost and enters the hunting state again.

Where n and m are design parameters. Large m values take more time to synchronize, but can reduce false cell boundary identification, and large n values take more time to recognize that they are not sorted, but can reduce false misalignment recognition. .

Therefore, if the cell of the asynchronous transmission method (ATM) is transmitted in the physical layer based on the synchronous transmission method (SDH, Synchronous Digital Hierarchy), the above parameters have values of N = 7 and M = 6. If the Asynchronous Cell is transmitted to the physical layer based on the cell, the above parameters have values of n = 7 and m = 8.

In addition, during the boundary identification process, the cell with the wrong header error control (HEC) is discarded.

However, this conventional technology is a technique applied to the asynchronous transmission method (ATM) to the last. In order to apply the conventional technology applied to the asynchronous transmission method (ATM) to the GFP as it is, various new technical elements are required.

First, the cell length of the asynchronous transmission method (ATM) is a fixed length. This means that once you identify a cell boundary, you can easily determine the location of the next cell. However, the frame length of GFP is 10 to 65539 bytes. This means that a variable length frame is used. This means that once you find the boundary of a frame, you don't know the length of the next frame. After all, it indicates that a new field is needed to know the starting point of the next frame.

In addition, an asynchronous transmission (ATM) cell is 53 bytes long and contains relatively short data.

However, the maximum length of one frame in GFP is 65539 bytes. In other words, in the process of boundary identification, seven frames must be discarded in order to move from the motor state to the synchronous state, which causes too much data loss. That is, CRC check is performed when the motor is moved from the motor state to the synchronous state. At this time, it must be corrected seven times in a row and performed up to eight times. At this time, the CRC check is performed seven times and seven frames are discarded.

As such, when the conventional technology is applied to the GFP as it is, there are many problems as described above.

Accordingly, the present invention has been proposed to solve the above conventional problems, and an object of the present invention is to consider the boundary of a frame using the length field in identifying the boundary in consideration of the variable frame length of the GFP. An apparatus and method for identifying a boundary of a GFP received frame capable of minimizing time required for identifying and identifying a frame boundary by performing four header error control calculations.

In order to achieve the above object, a boundary identification apparatus for a GFP reception frame according to an embodiment of the present invention includes a counter for outputting GFP frame data by 4 bytes, and the counter by the counter. A plurality of HEC (Header Error Check) calculation units which respectively perform header error control for each case divided by 4 bytes, and an OR operation unit which receives an output of the plurality of HEC calculation units and performs an OR operation to generate a motor call signal. And, when the motor call signal is inputted from the logical sum operation unit, when the synchronization is corrected, the length field value is output to determine the boundary position of the next frame, and when the length field value reaches a specific value, this position is the boundary of the next frame. The frame counter to identify and the frame counter connected to the frame counter always have the same position in the frame check sequence (FCS) field of the GFP. A boundary identifier configured to provide a signal necessary to fix the value to the frame counter, down-count the length field value input from the frame counter, and output a signal to the frame counter when a specific value is reached; The technical configuration is characterized by including a descrambler that performs descrambling on the frame data.

In order to achieve the above object, the boundary identification method of the GFP reception frame according to an embodiment of the present invention checks header error control every 4 bytes for all input data when the state of the GFP frame boundary identification is a hunting state. A first step of determining whether the correct header error control is found; when the correct header error control is found, the state of frame boundary identification is set to the motor state, and the length field is decreased by one for each byte to determine the value of the length field. A second step of determining whether the value is 0; a third step of determining whether the header error control value is correct after calculating the header error control value of the next frame using the position of the value if the length field value is 0; If the header error control value is correct, the state of frame boundary identification is set to a synchronous state, and the frame boundary and the length field are used to identify the frame boundary. Is characterized by its technical configuration, including performing a fourth step.

1 is a block diagram of a general GFP processing apparatus;

2 is a block diagram of a general asynchronous transmission cell,

3 is a configuration diagram of a general Ethernet frame,

4 is a configuration diagram of a typical HDLC frame,

5 is a state transition diagram for cell boundary identification in the conventional asynchronous transmission method,

6 is a block diagram of an apparatus for identifying a boundary of a GFP received frame according to an embodiment of the present invention.

7 is a detailed configuration diagram of the boundary identification unit in FIG. 6,

8 is a flowchart illustrating a boundary identification method of a GFP received frame according to an embodiment of the present invention.

9 is a configuration diagram of a GFP frame used in the present invention.

Explanation of symbols on the main parts of the drawings

10: descrambler 20: boundary identification unit

21: counter 22 to 25: first to fourth HEC calculator

26; Logic Unit 27: Frame Counter

28: decrement counter 30: demultiplexer

Hereinafter, an embodiment according to the present invention configured as described above, an apparatus for identifying a boundary of a GFP reception frame, and a method thereof will be described in detail with reference to the accompanying drawings.

6 is a block diagram of an apparatus for identifying a boundary of a GFP received frame according to an embodiment of the present invention.

As shown therein, a descrambler 10 which performs descrambling on the GFP frame data; The descrambler 10 performs a header error control (HEC) calculation using a length field of GFP frame data, identifies a frame boundary, and transmits the identified result to the descrambler 10 so that the delimiter 10 succeeds. It is configured to include a boundary identification unit 20 to perform descrambling in.

Here, the reference numeral 30 is a demultiplexer which decapsulates the descrambled signal by the descrambler 30 and performs decapsulation and removes an idle signal at an Idle termination. .

7 is a detailed configuration diagram of the boundary identification unit in FIG. 6.

As shown therein, a counter 21 for generating a counter necessary to perform header error control by dividing the input GFP frame data into a plurality of cases; A plurality of HEC calculators 22 to 25 which respectively perform header error control for each case divided by the counter; A logic sum calculator 26 that receives the outputs of the plurality of HEC calculators 22 to 25 to generate a logical sum operation to generate a motor call signal PSYN_LD; When the motor call signal PSYN_LD is input from the logical sum operation unit 26 and the synchronization is correct, the length field necessary to determine the position of the next frame is extracted, and when the final synchronization is determined, the frame is valid. A frame counter 27 for informing; A decrement counter 28 which is connected to the frame counter 27 and provides the frame counter 27 with a signal END_DNCNT necessary to always fix the position of a frame check sequence (FCS) field of a GFP to the same value. It is configured to include.

In the above, the plurality of HEC calculators 22 to 25 perform calculation of the header error control (HEC) corresponding to the first in the GFP frame data divided by the counter 21, and in the motor state or the synchronous state, A first HEC calculator 22 for detecting an error in the data length field; A second HEC calculator 23 for calculating a second header error control (HEC) corresponding to the second from the GFP frame data divided by the counter 21; A third HEC calculator 24 for calculating a third header error control (HEC) in the GFP frame data divided by the counter 21; And a fourth HEC calculator 25 for calculating a header error control (HEC) corresponding to the fourth in the GFP frame data divided by the counter 21.

8 is a flowchart illustrating a boundary identification method of a GFP received frame according to an embodiment of the present invention.

As shown in the drawing, when the frame boundary identification state is a hunting state, a first step (ST11) of checking the header error control (HEC) for every input data by a predetermined byte and determining whether the correct header error control is found ( ST12); A second step (ST13 to ST15) of determining whether the value of the length field is zero by reducing the length field by one for each byte when the correct header error control is found; A third step (ST16 to ST18) of starting a next frame, calculating a header error control (HEC) value and determining whether the header error control (HEC) value is correct when the length field value is 0; If the header error control value is correct, the frame boundary identification state is set as a synchronous state, and a fourth step (ST19) (ST20) of identifying a frame boundary using a frame counter and a length field is performed.

In the first step, the header error control is performed by 4 bytes when the header error control is performed.

Operation of the boundary identification device and method of the GFP reception frame according to the present invention configured as described above will be described in detail with reference to the accompanying drawings.

9 is a configuration diagram of a GFP frame used in the present invention.

Thus, in the case of an Ethernet frame, there are 7 bytes of preambles and 1 byte of boundary identification bytes. In case of HDLC, a flag byte is used to identify a frame boundary.

On the other hand, in the GFP used in the present invention, such a separate boundary identification byte does not exist, and frame boundary identification is performed only by HEC. HEC is also used for frame boundary identification, but also for error checking of coreHeader. Therefore, boundary identification and error checking are performed with only two bytes.

As a result, doing so can reduce unnecessary overhead. Also, unlike the case of writing a boundary pattern byte of a specific pattern, the same boundary identification algorithm can be applied to any signal loaded in the payload.

6 is a frame receiving unit of the GFP. In this case, it receives only the payload part of the SONET frame, which is a signal of the physical layer, and extracts only the signal necessary for the GFP among the payload, and sends it to the second layer.

At this time, the descrambler 10 operates only in a synchronous state. The boundary identification unit 20 checks the HEC so that descrambling is performed only when the boundary identification is successful, and the descrambling data is generated as the original two-layer signal after removing the header. The frame of the second layer generated here is transmitted to the upper layer through each port. And the idle terminal (idle termination) serves to remove the idle frame inserted in the transmitter for continuity of the physical layer signal.

In FIG. 7, the counter 21 generates a counter necessary for performing header error control by dividing the input GFP frame into a plurality of cases. That is, the incoming data is divided into four cases, and the counter signal HEC_RST for header error control is generated and output.

The first HEC calculator 22 calculates the header error control (HEC) for the first data in the GFP frame divided by the counter 21 and outputs the first motor call signal PSYN_LD1. . In particular, it performs header error control in the first case in the hunting state, but is actually used to detect an error in the data length field in the motor state or the synchronous state.

The second HEC calculator 23 calculates a second header error control HEC from the GFP frame data divided by the counter 21 and outputs a second motor call signal PSYN_LD2.

The third HEC calculator 24 calculates a third header error control HEC from the GFP frame data divided by the counter 21 and outputs a third motor call signal PSYN_LD3.

The fourth HEC calculator 25 calculates the header error control HEC corresponding to the fourth from the GFP frame data divided by the counter 21 and outputs the fourth motor call signal PSYN_LD4.

The OR operation unit 26 receives the motor call signals PSYN_LD1 to PSYN_LD4 output from the first to fourth HEC calculators 22 to 25, and performs an OR operation to generate and output the motor call signal PSYN_LD. Thus, the logical sum operation unit 26 outputs an electric motor call signal when any one of the first to fourth HEC calculation units 22 to 25 is matched with the header error control calculation.

The frame counter 27 receives the motor call signal PSYN_LD from the logical sum operation unit 26 and extracts a length field necessary for determining the position of the next frame when the synchronization is correct. It also plays a role in notifying that the frame is valid when it is finally determined that the synchronization is correct.

The decrement counter 28 serves to provide the frame counter 27 with the signal END_DNCNT necessary to always fix the position of the FCS field, which is to be selectively adopted by the GFP, to the same value.

Referring to the operation of the present invention in more detail as follows.

First, input to the GFP from the physical layer is made in bytes. And at that time, only the data stream comes into the input, not the other signals that distinguish the frames.

Therefore, in order to process the input data, an algorithm that can distinguish the frame from the input data is required.

As shown in FIG. 9, in the frame structure of the GFP, there is a header error control field, which is a code for detecting an error for the first length field and the length field. These two fields can be used to identify frame boundaries.

Therefore, in the hunting state, the starting point of the frame is unknown, so the header error control is checked every 4 bytes for all input data. However, while checking the first 4 bytes, data continues to be input. In other words, if you count from the first byte to count four bytes, the four bytes starting from the second byte cannot be counted while calculating the second and third bytes.

If any one of the four header error control calculation units 22 to 25 finds the correct header error control, the motor call signal PSYN_LD # is immediately generated. Here, # represents the number of each header error control calculation unit 22 to 25.

The PSYN_LD signal generated by each HEC control unit 22 to 25 is generated and transmitted to the frame counter 27.

Upon receiving this signal, the state of frame boundary identification becomes the motor state, and the frame counter 27 is reset to "0". The frame counter 27 then sends the two bytes immediately after the PSYN_LD signal, the value of the length field, to the decrement counter 28, which is a 16-bit decrement counter.

The 16-bit decrement counter 28 decrements this length field by one backward for each byte, and informs the frame counter 27 of the time END_DNCNT when the value is " 0 ". The time END_DNCNT at which the value of the 16-bit decrement counter 28 becomes "0" is a signal indicating the start of the next frame.

The frame counter 27 transmits a signal NEXT_LOAD indicating the start of the next frame to the first HEC calculator 22. At this time, the second to fourth HEC controllers 23 to 25, which are other HEC controllers, do not operate, and only the first HEC calculator 22 performs the calculation.

If the HEC value calculated immediately after the first HEC calculator 22 receives the NEXT_LOAD signal is a correct value, the first HEC calculator 22 generates a SYN_LOAD signal. After that, the state of the frame boundary identification circuit is synchronized, and the frame boundary and the length field are used to identify the frame boundary.

In addition, if the HEC calculation value calculated immediately after receiving the NEXT_LOAD signal is wrong, the state of frame boundary identification is hunting, and the first process is repeated.

Likewise, if the HEC calculation value is wrong even in the synchronous state, the frame boundary identification state becomes hunting state, and all processes are repeated from the beginning.

Thus, the present invention, unlike the cell boundary identification in the conventional asynchronous transmission method (ATM), all the process from the hunting state to the motor state and the motor state to the hunting state and the motor state to the synchronous state and the synchronous state to the hunting state This is done in one run. In other words, in the case of the asynchronous transmission method, the HEC calculation value must be corrected seven times in order to move from the motor state to the synchronous state. It becomes possible.

This is because the size of one frame of the GFP is much larger than the cells of the asynchronous transmission method, thereby minimizing the number of dropped frames.

This is because, if the state is not synchronized in the boundary identification unit 20, the frame is considered to be useless value.

It is also possible to use a single HEC calculation circuit and a buffer instead of using several HEC calculation circuits, but using several HEC calculation circuits can find frame boundaries faster.

As described above, the present invention considers that the frame length of the GFP is variable, and thus, the time required to identify the frame boundary by using the length field and perform four header error control calculations to identify the frame boundary is determined. Will be minimized.

Although the preferred embodiment of the present invention has been described above, the present invention may use various changes, modifications, and equivalents. It is clear that the present invention can be applied in the same manner by appropriately modifying the above embodiments. Accordingly, the above description does not limit the scope of the invention as defined by the limitations of the following claims.

As described above, the apparatus and method for identifying a boundary of a GFP received frame according to the present invention identify the boundary of the frame using the length field in identifying the boundary in consideration of the variable frame length of the GFP. It is possible to minimize the time required to identify the frame boundary by performing two header error control calculations, and the effect of identifying the frame boundary more accurately.

Claims (5)

A counter that outputs GFP (Generic Frame Procedure) frame data by 4 bytes each, and a plurality of HECs (Header Error Check) each of which performs header error control for each case divided by 4 bytes by the counter. A logic sum operation unit configured to generate a motor call signal by performing a logical sum operation based on a calculation unit, the outputs of the plurality of HEC calculation units, and to determine a boundary position of a next frame when the motor call signal is inputted from the logical sum calculation unit to be synchronized Outputs a length field value and when the length field value reaches a specific value, a frame counter identifying this position as a boundary of the next frame, connected with the frame counter, and determining the position of a frame check sequence (FCS) field of a GFP. To provide the necessary signal to the frame counter at all times A boundary identification unit comprising a decrement counter which down counts the received length field value and outputs it to the frame counter when a specific value is reached; And a descrambler for descrambling the GFP frame data. delete The method of claim 1, wherein the plurality of HEC calculation unit, A first HEC calculator configured to calculate a header error control corresponding to the first in the GFP frame data divided by the counter, and detect an error of a data length field in an electric motor state or a synchronous state; A second HEC calculator configured to calculate a second header error control in the GFP frame data divided by the counter; A third HEC calculator configured to calculate a header error control corresponding to the third in the GFP frame data divided by the counter; And a fourth HEC calculator configured to calculate a header error control corresponding to the fourth in the GFP frame data divided by the counter. A first step of checking header error control every 4 bytes for all input data when the state of the GFP frame boundary identification is hunting, and determining whether the correct header error control is found; A second step of determining whether the value of the length field is zero by reducing the length field by one for each byte when the correct header error control is found; If the length field value is 0, calculating a header error control value of a next frame using the position of the value and determining whether the header error control value is correct; If the header error control value is correct, the frame boundary identification state is set to a synchronous state, and a fourth step of identifying a frame boundary using a frame counter and a length field is performed. Identification method. delete