CN114143268B - TCP-based message receiving method and device, electronic equipment and storage medium - Google Patents

Abstract

The application discloses a message receiving method and device based on TCP, electronic equipment and a storage medium, wherein the message receiving method based on TCP can comprise the following steps: if the current message segment is not the s-th sequence message segment to be received, determining that the current message segment is a current disordered message segment; wherein s is a positive integer; if two cached message segments exist, determining whether a preset condition is met between the current disordered message segment and the cached message segment, and obtaining a determination result; wherein, the cached message segment is: historical disordered message segments; the meeting of the preset conditions between the current disordered message segment and the cached message segment comprises the following steps: the current disordered message segment and the cached message segment content are continuous or have overlapped content; the N is smaller than or equal to the preset maximum number M of the message segments allowed to be cached; and determining whether to cache the current disordered message segment according to the determination result.

Description

TCP-based message receiving method and device, electronic equipment and storage medium

Technical field

The present application relates to the field of network technology, and in particular to a message receiving method and device, electronic equipment and storage medium based on Transmission Control Protocol (Transmission Control Protocol, TCP).

Background technique

The receiving functions of the TCP layer generally include:

Receive message segments from the Internet Protocol (IP) layer;

Parse header information;

According to the correspondence between serial number and address Cache data in the cache area;

Execute window control and submit the data received in the positive sequence to the application layer in a timely manner;

Maintain out-of-order packets.

Maintenance methods for out-of-order packets include: caching all packets that fall within the window and using linked lists or data identifiers to identify them. However, this caching method will cause huge storage overhead.

Contents of the invention

Embodiments of the present application provide a TCP-based message receiving method and device, electronic equipment, and storage media.

The first aspect of the embodiment of the present application provides a message receiving method based on the transmission control protocol TCP, including:

If the current message segment is not the s-th sequential message segment to be received, determine that the current message segment is the current out-of-order message segment;

If there are N cached message segments, determine whether the preset conditions are met between the current out-of-order message segment and the cached message segment, and obtain a determination result; wherein the cached message segment is: Historical out-of-order message segments; satisfying the preset conditions between the current out-of-order message segment and the cached message segment includes: the contents of the current out-of-order message segment and the cached message segment Continuous or overlapping content exists; the N is less than or equal to the preset maximum number of segments M that are allowed to be cached;

According to the determination result, it is determined whether to cache the current out-of-order message segment.

Based on the above solution, determining whether to cache the current out-of-order message segment according to the determination result includes:

When the determination result indicates that the current out-of-order message segment and the cached message segment do not meet the preset condition, discard the current out-of-order message segment;

or,

When the N is less than the M and the current message segment and the cached message segment do not meet the preset condition, the current message segment is cached.

Based on the above solution, determining whether to cache the current out-of-order message segment based on the determination result also includes:

If the current out-of-order message segment and the cached message segment meet the preset condition, the operation to be performed on the first message segment is determined based on the current out-of-order message segment, wherein the first The message segment is any one of the N cached message segments.

Based on the above solution, if the current out-of-order message segment and the cached message segment meet the preset condition, determine the operation to be performed on the first message segment based on the current out-of-order message segment, Include at least one of the following:

If the message contents of the current out-of-order message segment are all included in the first message segment, determine to maintain the first message segment;

If part of the content of the current out-of-order message segment is included in the first message segment or the current out-of-order message segment is continuous with the content of the first message segment, according to the current out-of-order message segment segment, update the first message segment to obtain a first update message segment, and cache the first update message segment; wherein the first update message segment includes: within the current out-of-order message segment Content currently not included in the first message segment.

Based on the above solution, the method also includes:

If at least part of the content of the first update message segment is located outside the receiving window, the current out-of-order message segment is discarded.

Based on the above solution, the cache of the first update message segment includes:

If the first update message segment is located within the receiving window, cache the first update message segment.

Based on the above solution, the method also includes:

If the preset condition is met between the first update message segment and the second message segment, merge the first update message segment and the second message segment to obtain a second update message segment; Wherein, the second message segment is a cached message segment other than the first message segment among the N cached message segments;

Cache the second update message segment.

Based on the above solution, the cache of the second update message segment includes:

If the second update message segment is located within the receiving window, cache the second update message segment.

Based on the above solution, the method also includes:

When at least part of the second update message segment is located outside the receiving window, the current out-of-order message segment is discarded.

Based on the above solution, the method also includes:

If the current message segment is the s-th sequence message segment;

Submit the s-th sequence message segment to the application layer.

Based on the above solution, the method also includes:

When the current message segment is the s-th sequential message segment, determine whether the preset condition is satisfied between the cached message segment and the s-th sequential message segment;

Submit the cached message segments that satisfy the preset condition between the s-th sequential message segment and the s-th sequence message segment to the application layer;

Delete the cached message segment that has been submitted to the application layer.

Based on the above solution, the method also includes:

When a message segment is submitted to the application layer, the information of the s-th sequence message segment to be received and the receiving window are updated according to the message segment submitted to the application layer.

Based on the above solution, the method also includes:

At least one of the N cached message segments is updated. According to the updated cached message segment, the boundary information of the cached message segment is updated, wherein the boundary information includes: indicating the corresponding The information and indication on the left boundary of the cached message segment correspond to the information on the right boundary of the cached message segment.

The second aspect of the embodiment of the present application provides a message receiving device based on Transmission Control Protocol TCP. The device includes:

A first determination module, configured to determine that the current message segment is the current out-of-order message segment if the current message segment is not the s-th sequential message segment to be received;

The second determination module is used to determine whether the current out-of-order message segment and the cached message segment satisfy the preset conditions if there are N cached message segments, and obtain a determination result; wherein, the The cached message segment is: a historical out-of-order message segment; the preset condition between the current out-of-order message segment and the cached message segment includes: the current out-of-order message segment and all The content of the cached message segments is continuous or there is overlapping content; the N is less than or equal to the preset maximum number of segments M that is allowed to be cached;

A cache module, configured to determine whether to cache the current out-of-order message segment according to the determination result.

A third aspect of the embodiment of the present disclosure provides an electronic device, including:

memory;

A processor, connected to the memory, is configured to implement the TCP-based message receiving method provided by any of the foregoing technical solutions by executing computer-executable instructions stored on the memory.

The fourth aspect of the embodiment of the present disclosure provides a computer storage medium, which is characterized in that the computer storage medium stores computer executable instructions; after the computer executable instructions are executed by the processor, the method provided by any of the foregoing technical solutions is implemented. TCP message receiving method.

The technical solution provided by the embodiment of the present disclosure first allows out-of-order messages to be cached, but the number of cached out-of-order messages is N. On the one hand, compared to out-of-order messages, subsequent requests are directly discarded, which can reduce repeated requests and Repeated transmission, on the other hand, limits the number of out-of-order messages, which can greatly reduce the storage overhead occupied by out-of-order messages, thereby achieving a balance between repeated transmission and storage overhead.

Description of drawings

Figure 1 is a schematic flow chart of a TCP-based message receiving method provided by an embodiment of the present application;

Figure 2 is a schematic flow chart of a TCP-based message receiving method provided by an embodiment of the present application;

Figure 3 is a schematic diagram of message reception provided by the embodiment of the present application;

Figure 4 is a schematic structural diagram of a TCP-based message receiving device provided by an embodiment of the present application;

Figure 5 is a schematic diagram corresponding to fields stored in a register provided by an embodiment of the present disclosure;

Figure 6 is a schematic flow chart of a TCP-based message receiving method provided by an embodiment of the present application;

Figure 7 is a schematic flow chart of a TCP-based message receiving method provided by an embodiment of the present application;

Figure 8 is a schematic flowchart of a TCP-based message receiving method provided by an embodiment of the present application. .

Detailed ways

In order to understand the characteristics and technical content of the present application in more detail, the implementation of the present application will be described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the present application.

As shown in Figure 1, an embodiment of the present disclosure provides a TCP-based message receiving method, including:

S110: If the current message segment is not the s-th sequential message segment to be received, determine that the current message segment is the current out-of-order message segment;

S120: If there are N cached message segments, determine whether the current out-of-order message segment and the cached message segment meet the preset conditions, and obtain a determination result; wherein, the cached message segment is: a historical out-of-order message segment; the preset condition satisfied between the current out-of-order message segment and the cached message segment includes: the current out-of-order message segment and the cached message segment The segment content is continuous or there is overlapping content; the N is less than or equal to the preset maximum number M of segments allowed to be cached;

S130: According to the determination result, determine whether to cache the current out-of-order message segment.

The TCP-based message receiving method provided by the embodiment of the present disclosure is applied to the message receiving end. The receiving end can be any network device. The network device can be: a terminal and/or a server. The terminals include but are not limited to: fixed terminals and/or mobile terminals. The fixed terminal includes: desktop computer, smart TV or advertising playback equipment, etc. The mobile terminals include but are not limited to: mobile phones, tablets, wearable devices, smart home devices, smart office devices, and/or various smart robots. The intelligent robot includes but is not limited to: a ground robot including a mobile chassis and/or a low-flying aerial robot.

The receiving end may include: a Field Programmable Gate Array (FPGA) chip. The FPGA chip is used to submit the message segments received by the IP layer to the TCP layer. The TCP layer then submits the message segments to the application layer. , to realize the sending and receiving of messages before the application layer.

If the currently received message segment is the first message segment, if the number of the message segment starts counting from 0, then the sth sequential message segment is message segment 0; if the number of the message segment starts counting from 1, then the sth sequence message segment starts from 1. The s sequence segment is segment 1. If the currently received message segment is not the first message segment, the currently received message segment is the message segment m+1 that has been submitted to the application layer; m is any natural number.

If the sequence number of the currently received message segment does not match the sequence number of the s-th sequence message segment, it means that the current message segment is an out-of-order message segment, and because it is received at the current time, the current message segment is the current message segment. Out-of-order message segments.

If N out-of-order message segments have been cached, it is necessary to determine whether to cache the current out-of-order message segment based on whether the current message segment and the cached message segment are continuous or have duplicate content.

If the value of N is less than or equal to M; M can be any positive integer. For example, the value of M can be 2, 3, or 4. In this case, the value of N can be 0, 1, 2, or any other natural number smaller than M. If N is 0, it means that there are currently no cached segments.

In one embodiment, if the maximum number of historical out-of-order packets has been cached, and the current packet segment is also an out-of-order packet segment, the method determines whether there is a gap between the current out-of-order packet segment and the cached packet segment. Continuous content or overlapping content determines whether the current message segment needs to be cached, thereby reducing the storage overhead caused by the unlimited number of cached out-of-order message segments, and compared with not caching out-of-order message segments, it can be reduced Repeat the transfer.

In some embodiments, the S130 may include:

S131: When the determination result indicates that the current out-of-order message segment and the cached message segment do not meet the preset condition and the N is equal to the M, discard the current out-of-order message segment. .

If the current out-of-order message segment and the cached message segment are neither continuous segments nor overlapping content, it means that if the current out-of-order message segment is cached, the cached out-of-order message segment will exceed the size The number of segments is N, so the current out-of-order message segment will be discarded, thereby reducing the problem of excessive storage overhead.

If the current out-of-order message segment and the cached message segment are consecutive message segments, one of the following conditions is met:

The last byte of the cached message segment is adjacent to the first byte of the current out-of-order message segment;

The first byte of the cached message segment and the last byte of the current out-of-order message segment are adjacent bytes.

The current out-of-order message segment and the cached message segment have overlapping contents, including:

At least one byte contained in the current out-of-order segment is within a cached segment.

According to the unpacking of the TCP protocol, multiple bytes within a message segment are continuously distributed.

In other embodiments, the method further includes:

When the determination result indicates that the current out-of-order message segment and the cached message segment do not meet the preset condition, then the current out-of-order message segment and the cached message segment are discarded. S-sequential message segments are the segments with the closest distance between them.

If an out-of-order message segment is received, a request for the s-th sequence message segment will be sent again based on TCP. In view of this, based on the sequence request of the receiving end, the closer the current s-th sequence message segment is Segments will be requested earlier, thereby reducing the average cache duration of a cached message segment, thereby reducing the cache capacity occupied by reducing the average cache duration of a single cached message segment, thereby reducing cache overhead again.

In other embodiments, S130 may also include:

S132: When the N is less than the M and the current message segment and the cached message segment do not meet the preset condition, cache the current message segment.

If N is less than M, it means that the currently cached message segments have not reached the maximum number of segments. At this time, if the currently received message segment is an out-of-order message segment, the current out-of-order message segment can continue to be cached.

In other embodiments, the S130 further includes:

S133A: If the current out-of-order message segment and the cached message segment meet the preset condition, obtain the first updated message segment based on the current out-of-order message segment and the first message segment, , the first message segment is any one of the N cached message segments;

S133B: Cache the first update message segment.

At this time, regardless of whether N is less than or equal to M, the number of cached message segments will be reduced by merging message segments based on the preset conditions between the current out-of-order message segment and each cached message segment. Also reduce caching of duplicate content.

In one embodiment, the S130 further includes:

If the current out-of-order message segment and the cached message segment meet the preset condition, the operation to be performed on the first message segment is determined based on the current out-of-order message segment, wherein the first The message segment is any one of the N cached message segments.

The current out-of-order message segment may be caused by the retransmission of the message segment or the loss or duplication of bytes carried due to forwarding or packet loss during the transmission process.

In this case, the preset conditions between the current out-of-order message segment and the cached first message segment can be divided into the following situations:

The content between the current out-of-order message segment and the first message segment is continuous and has no duplication;

All contents of the current out-of-order message segment are included in the first message segment;

Part of the content of the current out-of-order message segment is included in the first message segment, and some new content is also included.

In some embodiments, the S130 may include at least one of the following:

If the current out-of-order message segment and the cached message segment meet the preset condition, a first updated message segment is obtained according to the current out-of-order message segment and the first message segment, where, The first message segment is any one of the N cached message segments;

Cache the first update message.

That is to say, the operation on the first message segment is different for different situations. The following is an analysis based on the above situations:

Case 1: All contents of the current out-of-order message segment are included in the first message segment, then the first message segment will not be updated and the current cache status of the first message segment will be maintained; that is, the current first message segment The text segment is the aforementioned first update message segment;

Case 2: Part of the content of the current out-of-order message segment is included in the first message segment. Add the content of the current message segment that is not included in the first message segment to the first message segment. After the content is added, The first update message segment;

Case 3: All the contents of the current out-of-order message segment are not included in the first message segment, then the current message segment and the first message segment are directly merged to obtain the merged first updated message segment.

The subsequent first updated message segment will be used as the cached message segment at the next moment.

According to the different situations listed above, the first updated message segment is obtained based on the current out-of-order message segment and the first message segment, including at least one of the following:

If the message contents of the current out-of-order message segment are all included in the first message segment, determine that the first message segment is the first update message segment;

If part of the content of the current out-of-order message segment is included in the first message segment, add the content of the current out-of-order message segment that is not included in the first message segment to the first message segment. message segment to obtain the first update message segment;

If the contents of the current out-of-order message segment and the first message segment are continuous, merge the first message segment and the current out-of-order message segment to obtain a first updated message segment.

In some embodiments, the method further includes:

A receive window is configured, which limits the maximum number of buffered bytes P.

The window length of the receiving window shown in Figure 3 is 8 bytes, that is, P=8. The first byte of the receive window is the k-th byte.

The starting point of the specific receiving window is: the first byte of the s-th sequence message segment to be received;

The termination byte of the receiving window is: P bytes. P is the maximum number of cache bytes contained in the window. P can be any positive integer.

The end byte of the receive window minus the first byte of the receive window equals P, and P can be greater than the maximum number of bytes contained in a message segment.

In the embodiment of the present disclosure, in order to suppress the depositor overhead caused by too many bytes contained in a single cached message segment, a receiving window is introduced, and how to cache the message segment is further determined through the receiving window.

Exemplarily, the caching of the first update message segment includes:

If the first update message segment is located within the receiving window, cache the first update message segment.

The first update message segment is located within the receiving window, that is, the number of the last byte of the first update message segment is less than or equal to the end byte of the current receiving window.

If the first update message segment is within the receiving window, the first update message segment is cached; otherwise, the first update message segment is not cached.

After not caching the first update message segment, the method further includes:

Discard the current out-of-order segment;

or,

The first updated message segment is cached after discarding one or more message segments with the smallest distance between the current random message segment and the current s-th sequential message segment. The distance here can be reflected in the number of bytes of the interval.

In some embodiments, the method further includes:

If the preset condition is met between the first update message segment and the second message segment, a second update message segment is obtained based on the first update message segment and the second message segment; Wherein, the second message segment is a cached message segment other than the first message segment among the N cached message segments;

Cache the second update message segment.

If the content between the generated update message segment and the second message segment is continuous or has duplicate content, the first update message segment and the second message segment will be merged to obtain the second update message segment.

The first message segment can be a message segment that meets the preset conditions with the current out-of-order message segment; if the current message segment meets the preset conditions with two cached message segments at the same time, the current message segment After the segment is merged with one of the cached segments to obtain the first update segment, the first update segment will be fully preset with another cached segment (i.e., the second segment). relation.

In some embodiments, if the preset condition is met between the first update message segment and the second message segment, the second message segment is determined based on the first update message segment. operations, including one of the following:

If the entire content of the second message segment is included in the first update message segment, delete the second message segment;

If part of the content of the second message segment is included in the first update message segment, add the content of the second message segment that is not included in the first update message segment to the third update message segment. An update message segment is obtained to obtain the second update message segment;

If the contents of the second message segment and the first update message segment are continuous, merge the first update message segment and the second message segment to obtain the second update message segment.

The second message segment here is: in addition to the first message segment, any one of a plurality of cached message segments that have a preset relationship with the current message segment.

In one embodiment, caching the second update message segment includes:

If the second update message segment is located within the receiving window, cache the second update message segment.

In another embodiment, the method further includes:

When at least part of the second update message segment is located outside the receiving window, the current out-of-order message segment is discarded.

By setting the receive window, it is equivalent to setting the maximum bytes contained in M message segments, thereby controlling the maximum buffer amount.

In some embodiments, the method further includes:

If the current message segment is the s-th sequence message segment;

Submit the s-th sequence message segment to the application layer.

When the TCP layer determines that the current message segment received is the s-th sequence message segment, it directly reports the s-th sequence message segment to the application layer, that is, it does not cache the currently received sequence message segment and directly submits it to the application layer. layer, thereby reducing the cache occupied by the current message segment.

In some embodiments, the method further includes:

After the s-th sequence message segment is submitted to the application layer, update the sequence number of the s-th sequence message segment and determine the s+1-th sequence message segment;

Determine whether the s+1th sequential message segment exists in the cached message segment;

When the s+1th sequential message segment exists in the cached message segment, submit the s+1th positive sequence in the cached message segment to the application layer;

Delete the cached message segment submitted to the application layer.

For example, after the s-th sequence message segment is submitted to the application layer, the s+1-th sequence message segment will be determined. The first byte of the s+1th sequential message segment is the next byte of the last byte of the sth sequential message segment.

If there is a cached message segment containing the first byte of the s+1th sequential message segment in the cached message segment, then the cached message segment is the s+1th sequential message segment. paragraph.

Submit the s+1 sequential message segment of the cached message segment to the application layer, and delete the cached message segment that has been submitted to the application layer from the cache area. On the one hand, it reduces repeated submissions to the application layer, and on the other hand, it reduces repeated submissions to the application layer. On the one hand, cached message segments can be submitted to the application layer without repeated transmission.

In some embodiments, after receiving the s-th sequence message segment, the application layer will transmit a request for the s+1-th sequence message segment to the sending end. If the currently cached message segment contains the s+1-th sequence message segment, Sequential message segment, the request can be intercepted to reduce the request triggering the sender to repeatedly transmit the s+1 sequential message segment.

In some embodiments, the method further includes:

When a message segment is submitted to the application layer, update the sequence number of the sequence message segment to be received based on the message segment that has been submitted to the application layer;

and / or,

Update the receiving window, wherein the N cached message segments are located in the receiving window.

The sequence number of the updated sequential message segment here is the next byte of the last byte of the message segment that has been submitted to the application layer. The first byte of the receive window here is the first byte of the updated sequence message segment + P bytes.

In one embodiment, the method further includes:

At least one of the N cached message segments is updated. According to the updated cached message segment, the information corresponding to the boundary of the cached message segment is updated, wherein the boundary information includes : Indicates information corresponding to the left boundary of the cached message segment and information indicating corresponding to the right boundary of the cached message segment.

As shown in Figure 4, an embodiment of the present disclosure provides a TCP-based message receiving device. The device includes:

The first determination module 110 is configured to determine that the current message segment is the current out-of-order message segment if the current message segment is not the s-th sequential message segment to be received; wherein, the s is a positive integer;

The second determination module 120 is used to determine whether the preset conditions are met between the current out-of-order message segment and the cached message segment if there are N cached message segments, and obtain a determination result; wherein, The cached message segment is: a historical out-of-order message segment; the preset condition between the current out-of-order message segment and the cached message segment includes: the current out-of-order message segment and The contents of the cached message segments are continuous or have overlapping contents;

The cache module 130 is configured to determine whether to cache the current out-of-order message segment according to the determination result.

In some embodiments, the first determination module 110, the second determination module 120 and the cache module 130 may be program modules; after the program modules are executed by the processor, the above functions can be implemented. In some embodiments, the first determination module 110, the second determination module 120 and the cache module 130 may be soft and hard combined modules; the soft and hard combined module board includes but is not limited to a programmable array; the programmable array Including but not limited to: field programmable arrays and/or complex programmable arrays.

In some embodiments, the first determination module 110, the second determination module 120 and the cache module 130 may be pure hardware modules; the pure hardware modules include but are not limited to application specific integrated circuits.

In some embodiments, the caching module 130 is specifically used when the determination result indicates that the current out-of-order message segment and the cached message segment do not meet the preset condition and the N is equal to the When M is stated, discard the current out-of-order message segment; or, when N is less than M and the current message segment and the cached message segment do not meet the preset conditions, cache the The current message segment.

In some embodiments, the caching module 130 is specifically used to cache the current out-of-order message segment and the first cached message segment according to the current out-of-order message segment and the first cached message segment if the current out-of-order message segment and the cached message segment satisfy the preset condition. message segment to obtain a first update message segment, where the first message segment is any one of the N cached message segments; and the first update message segment is cached.

In some embodiments, the cache module 130 is specifically configured to perform at least one of the following: if all the message contents of the current out-of-order message segment are included in the first message segment, determine that the first message segment The text segment is the first update message segment;

If part of the content of the current out-of-order message segment is included in the first message segment, add the content of the current out-of-order message segment that is not included in the first message segment to the first message segment. message segment to obtain the first update message segment;

If the contents of the current out-of-order message segment and the first message segment are continuous, merge the first message segment and the current out-of-order message segment to obtain a first updated message segment.

In some embodiments, the device further includes:

A discarding module configured to discard the current out-of-order message segment if at least part of the content of the first update message segment is located outside the receiving window.

In some embodiments, the cache module 130 is specifically configured to cache the first update message segment if the first update message segment is located within the receiving window.

In some embodiments, the device further includes:

Obtaining module, configured to obtain a second update message segment according to the first update message segment and the second message segment if the preset condition is satisfied between the first update message segment and the second message segment. Update message segments; wherein the second message segment is a cached message segment other than the first message segment among the N cached message segments;

The cache module 130 is also used to cache the second update message segment.

In some embodiments, the obtaining module is specifically configured to perform one of the following:

If the entire content of the second message segment is included in the first update message segment, delete the second message segment;

If part of the content of the second message segment is included in the first update message segment, add the content of the second message segment that is not included in the first update message segment to the third update message segment. An update message segment is obtained to obtain the second update message segment;

If the contents of the second message segment and the first update message segment are continuous, merge the first update message segment and the second message segment to obtain the second update message segment.

In some embodiments, the cache module 130 is also configured to cache the second update message segment if the second update message segment is located within the receiving window.

In some embodiments, the device further includes:

A discarding module configured to discard the current out-of-order message segment when at least part of the second update message segment is located outside the receiving window.

In some embodiments, the device further includes:

A reporting module is configured to submit the s-th sequence message segment to the application layer if the current message segment is the s-th sequence message segment.

In some embodiments, the device further includes:

An update module, configured to update the sequence number of the s-th sequence message segment and determine the s+1-th sequence message segment after the s-th sequence message segment is submitted to the application layer;

An existence module is used to determine whether the s+1th sequential message segment exists in the cached message segment;

The reporting module is also configured to submit the s+1th positive sequence in the cached message segment to the application when the s+1th sequential message segment exists in the cached message segment. layer;

The caching module 130 is configured to delete the cached message segments submitted to the application layer.

In some embodiments, the device further includes:

A sequence number module, used to update the sequence number of the sequence message segment to be received according to the message segment submitted to the application layer when a message segment is submitted to the application layer;

and / or,

A window module is used to update the receiving window, wherein the N cached message segments are located in the receiving window.

In some embodiments, the device further includes:

The boundary information module is used to update at least one of the N cached message segments, and update the information corresponding to the boundary of the cached message segment according to the updated cached message segment, wherein , the boundary information includes: information indicating the left boundary corresponding to the cached message segment and information indicating the right boundary corresponding to the cached message segment.

The boundary information may be: the sequence number of the first byte and the sequence number of the last byte of the cached message segment.

If there are N cached segments in the current cache, there are N left boundary information and N right boundary information. The information indication on the left boundary corresponds to the first byte of the cached segment; the information indication on the right boundary corresponds to the last byte of the cached segment.

The embodiment of the present disclosure provides an FPGA-based method for supporting TCP protocol out-of-order reception. The processing capabilities of this solution are described as follows:

1) Can receive two out-of-order message segments with no overlapping content;

2) If the third message segment is received sequentially, that is, the third message segment currently received is a sequential message segment, the sequential message segment can be processed normally (normal processing here includes but is not limited to : Directly submitted to the application layer), if it is an out-of-order segment and there is no overlap with the previous two out-of-order segments, the segment is discarded. The first two out-of-order message segments here are the aforementioned cached message segments. In this embodiment, M=2 is used as an example.

If the currently received out-of-order message segment overlaps with the first two buffered out-of-order message segments, the out-of-order message segment needs to be processed on a case-by-case basis based on the overlap.

The method based on the FPGA chip that supports out-of-order reception of the TCP protocol is suitable for the field of data interconnection communication between devices through the Ethernet TCP protocol. The TCP protocol is implemented by the FPGA chip, which releases the load of software processing on the general processor. Its characteristics are: it supports a certain degree of out-of-order processing of TCP protocol message segments; it adopts pipeline processing and occupies very little storage resources of the FPGA chip.

The so-called support for a certain degree of out-of-order processing: two out-of-order message segments with no overlapping relationship can be received; if the third message segment is received sequentially, it can be processed normally. If it is an out-of-order message segment, and it is the same as the first two If the out-of-sequence segment does not overlap, the segment is discarded. In practical applications, this solution can cover more than 95% of the out-of-order situations of the TCP protocol.

The so-called pipeline processing method, on the premise of ensuring the FPGA timing, uses five cycles to receive an out-of-sequence message segment, occupying very little FPGA storage resources, low implementation complexity, and high protocol reliability. The receiving window control logic includes two parts:

Maintenance of receiving window pointer;

Maintenance of out-of-order message segments.

(1) Maintenance of receiving window pointer

The schematic diagram of the receiving window is shown in Figure 5. The specific meanings of the identifiers appearing in the figure are shown in Table 5. rx_dat points to the first byte of sequential data, rx_nxt points to the next sequential data expected to be received, and rx_wnd identifies the size of the receiving window. This value is not affected by out-of-order data.

Among them, the change of rx_dat is implemented in the storage management module. This module uses this variable when calculating the right boundary of the receiving window. The change of rx_nxt is implemented by this module and notified to the storage management module. The storage management module uses this variable to determine the end position of the sequential data for reading by the application layer.

Table 1 Register management module interface description

Table 1

(2) Out-of-order message segment maintenance

As shown in Figure 6 and Figure 7, the receiving end maintains a register with two entries, and each entry sets a valid bit. For a newly received message segment, the processing process is as follows:

Kth cycle (data arrival): Calculate the information on the left boundary of the message segment (the sequence number of the first byte) and the information on the right boundary of the message segment (the sequence number of the last byte) based on the sequence number and length of the current message segment. ); at the same time, the information about the left boundary of the window, the information about the left and right boundaries of the out-of-order message segment 1, and the information about the left and right boundaries of the out-of-order message segment 2 are known.

The K+1th cycle: the left boundary of the window minus six boundary values (the left and right boundaries of the sequential message segment, the out-of-order message segment 1, and the out-of-order message segment 2). The boundary value may include: as a sequence number of the edited byte, the boundary value serves as a type of boundary information.

K+2 cycle: Boundary comparison is performed based on the difference result calculated in the first cycle, including: comparison of the left boundary of the message segment and the left boundary of the window, comparison of the left boundary of the message segment and the four out-of-order boundaries, comparison of the right boundary of the message segment The boundary is compared with four out-of-order boundaries, the left boundary of out-of-order segment 1 is compared with the right boundary of out-of-order segment 2, and the left boundary of out-of-order segment 2 is compared with the right boundary of out-of-order segment 1.

The operation of the k+2th cycle can be as follows:

If it is a sequential message segment, modify rx_nxt to the last byte sequence number plus 1. This situation includes message segments whose sequence number is less than rx_nxt, but the sum of the sequence number and length is greater than or equal to rx_nxt.

If it is an out-of-order message segment, compare the first and last byte sequence numbers of the current message segment with the sequence numbers in the valid entry, and modify the entry.

The operation of the K+3th cycle can be as follows:

If it is a sequential message segment, determine the relationship between rx_nxt and out-of-sequence message segment 1. If there is an entry connected to rx_nxt, modify the value of rx_nxt and set the entry to be invalid;

If it is an out-of-order message segment, if the out-of-order message segment 1 is invalid and the out-of-order message segment 2 has not been updated, it will be stored in the out-of-order message segment 1; if the out-of-order message segment 1 is valid, the out-of-order message segment 1 will be stored in the out-of-order message segment 1. If segment 2 is invalid and out-of-order segment 1 has not been updated, it is stored in out-of-order segment 2.

The operation of the K+4th cycle can be as follows:

If it is a sequential message segment, determine the relationship between rx_nxt and out-of-order message segment 2. If there is an entry connected to rx_nxt, modify the value of rx_nxt and set the entry to be invalid;

If it is an out-of-order message segment, the left boundary of the message segment is not equal to the left boundary of the window and both the out-of-order message segment 1 and the out-of-order message segment 2 are valid (before updating): If the left boundary of the message segment is less than or equal to The right boundary of out-of-order message segment 1 and the right boundary of the message segment are greater than or equal to the left boundary of out-of-order message segment 2 and the left boundary of out-of-order message segment 2 is greater than the right boundary of out-of-order message segment 1: the out-of-order message segment The right boundary of segment 2 is assigned to the right boundary of out-of-order message segment 1, and out-of-order message segment 2 is invalid. If the left edge of the message segment is less than or equal to the right edge of out-of-order message segment 2 and the right edge of the message segment is greater than or equal to the left edge of out-of-order message segment 1 and the left edge of out-of-order message segment 1 is greater than the left edge of out-of-order message segment 2 The right boundary of: assign the right boundary of out-of-order message segment 1 to the right boundary of out-of-order message segment 2, and out-of-order message segment 1 is invalid.

A processing example of this solution can be shown in Figure 8. The value of rx_nxt (the sequence number of the first byte of the next sequential message expected to be received) is 1. The three message segments arrive sequentially. The left and right boundaries of the message segments are bytes (2, 3) and (5, 7) respectively. , (3,8).

The sequence number (2, 3) of the bytes contained in the first message segment is determined to be an out-of-order message segment, then the left and right boundaries are stored, and the out-of-order message segment 1 entry is valid (T indicates that the entry is valid, F indicates The entry is invalid).

The byte sequence number (5, 7) contained in the second message segment is determined to be an out-of-order message segment, and there is no intersection with the out-of-order message segment 1 entry. The left and right boundaries are stored, and the out-of-order message segment 2 is stored. The entry is valid; the byte sequence number (3, 8) contained in the third message segment is an out-of-order message segment, but it overlaps with the entries of out-of-order message segment 1 and out-of-order message segment 2 at the same time. And after updating out-of-order message segment 1 and out-of-order message segment 2, there is an inclusive relationship between the two. Finally, the entry for out-of-order message segment 1 is valid, and the entry for out-of-order message segment 2 is invalid.

It has been verified that in practical applications, this solution can cover more than 95% of the out-of-order reception of message segments of the TCP protocol. At the same time, it takes up very little storage resources of the FPGA chip. It has low implementation complexity and high protocol reliability, and can be widely used in various applications. In the application environment of out-of-order reception. The disadvantage is that this solution cannot handle the situation where the third out-of-order message segment is not related to the first two out-of-order message segments. However, the TCP protocol has a retransmission mechanism, and the retransmission of a small number of discarded messages has little impact on the protocol.

An embodiment of the present disclosure provides an electronic device, characterized in that the electronic device includes:

memory;

A processor, connected to the memory, is configured to implement the TCP-based message receiving method provided in any of the foregoing embodiments by executing computer-executable instructions stored on the memory.

The electronic device may be any receiving device that receives messages, and the receiving device may include an FPGA chip that receives messages. The processor may include the FPGA chip.

Embodiments of the present disclosure provide a computer storage medium that stores computer-executable instructions; after the computer-executable instructions are executed by a processor, the TCP-based message receiving method provided by any of the foregoing embodiments can be implemented .

The technical solutions described in the embodiments of this application can be combined arbitrarily as long as there is no conflict.

In the several embodiments provided in this application, it should be understood that the disclosed methods and smart devices can be implemented in other ways. The device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods, such as: multiple units or components may be combined, or can be integrated into another system, or some features can be ignored, or not implemented. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the devices or units may be electrical, mechanical, or other forms. of.

The units described above as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units; Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, all the functional units in the embodiments of the present application can be integrated into one second processing unit, or each unit can be separately used as a unit, or N or more units can be integrated into one unit; The above integrated unit can be implemented in the form of hardware or in the form of hardware plus software functional units.

The above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application. should be covered by the protection scope of this application.

Claims (17)

1. A message receiving method based on a transmission control protocol TCP, comprising:

If the current message segment is not the s-th sequence message segment to be received, determining that the current message segment is a current disordered message segment; wherein s is a positive integer;

if N cached message segments exist, determining whether a preset condition is met between the current disordered message segment and the cached message segment, and obtaining a determination result; wherein, the cached message segment is: historical disordered message segments; the meeting of the preset conditions between the current disordered message segment and the cached message segment comprises the following steps: the current disordered message segment and the cached message segment content are continuous or have overlapped content; the N is smaller than or equal to the preset maximum number M of the message segments allowed to be cached;

determining whether to cache the current disordered message segment according to the determination result; and determining whether to cache the current disordered message segment according to the determination result, wherein the determining comprises the following steps: and when the determined result shows that the current disordered message segment and the cached message segment do not meet the preset condition, discarding the message segment with the closest interval distance between the current disordered message segment and the cached message segment and the s-th sequence message segment.

2. The method according to claim 1, wherein determining whether to cache the current out-of-order segment according to the determination result comprises:

discarding the current disordered message segment when the determined result shows that the current disordered message segment and the cached message segment do not meet the preset condition and the N is equal to the M;

or,

and caching the current message segment when the N is smaller than the M and the current message segment and the cached message segment do not meet the preset condition.

3. The method according to claim 1 or 2, wherein determining whether to cache the current out-of-order segment according to the determination result further comprises:

if the current disordered message segment and the cached message segment meet the preset conditions, a first updated message segment is obtained according to the current disordered message segment and a first message segment, wherein the first message segment is any one of N cached message segments;

and caching the first updated message segment.

4. The method of claim 3, wherein the obtaining a first updated segment from the current out-of-order segment and the first segment comprises at least one of:

If the message content of the current disordered message segment is contained in the first message segment, determining that the first message segment is the first updated message segment;

if part of the content of the current disordered message section is contained in the first message section, adding the content of the current disordered message section which is not contained in the first message section to obtain the first updated message section;

and if the current disordered message segment is continuous with the first message segment, merging the first message segment and the current disordered message segment to obtain a first updated message segment.

5. The method according to claim 4, wherein the method further comprises:

and discarding the current disordered message segment if at least part of the content of the first updated message segment is positioned outside the receiving window.

6. The method of claim 5, wherein the buffering the first update segment comprises:

and if the first updating message segment is positioned in the receiving window, caching the first updating message segment.

7. The method of claim 5, wherein the method further comprises:

if the preset condition is met between the first updating message segment and the second message segment, obtaining a second updating message segment according to the first updating message segment and the second message segment; wherein the second message segment is a cached message segment other than the first message segment in the N cached message segments;

And caching the second updated message segment.

8. The method of claim 7, wherein if the preset condition is satisfied between the first updated segment and the second segment, determining the operation on the second segment according to the first updated segment comprises one of:

deleting the second message segment if the whole content of the second message segment is contained in the first updated message segment;

if part of the content of the second message segment is contained in the first updated message segment, adding the content of the second message segment, which is not contained in the first updated message segment, to the first updated message segment to obtain the second updated message segment;

and if the content of the second message segment is continuous with that of the first updating message segment, combining the first updating message segment and the second message segment to obtain the second updating message segment.

9. The method of claim 7, wherein the buffering the second update segment comprises:

and if the second updating message segment is positioned in the receiving window, caching the second updating message segment.

10. The method according to claim 9, wherein the method further comprises:

And discarding the current out-of-order message segment when the second updated message segment is at least partially outside the receiving window.

11. The method according to claim 1, wherein the method further comprises:

and if the current message segment is the s-th sequence message segment, submitting the s-th sequence message segment to an application layer.

12. The method of claim 11, wherein the method further comprises:

after the s-th sequence message segment is submitted to the application layer, updating the sequence number of the s-th sequence message segment, and determining the s+1-th sequence message segment;

determining whether the (s+1) -th sequential message segment exists in the cached message segment;

when the (s+1) th sequence message segment exists in the cached message segment, submitting the (s+1) th positive sequence in the cached message segment to an application layer;

and deleting the cached message segment submitted to the application layer.

13. The method according to claim 11 or 12, characterized in that the method further comprises:

when a message segment is submitted to the application layer, updating the sequence number of the sequence message segment to be received according to the submitted message segment to the application layer;

and/or the number of the groups of groups,

And updating a receiving window, wherein N cached message segments are positioned in the receiving window.

14. The method according to any one of claims 1 to 2 or 4 to 12, further comprising:

at least one of the N cached message segments is updated, and according to the updated cached message segment, updating the information of the boundary corresponding to the cached message segment, wherein the information of the boundary comprises: information indicating a left boundary corresponding to the cached segment and information indicating a right boundary corresponding to the cached segment;

the information of the boundary may be: the sequence number of the first byte and the sequence number of the last byte of the cached message segment;

if N cached message segments are cached currently, the message segments have N left boundary information and N right boundary information; the information of the left boundary indicates the first byte of the corresponding cached message segment; the information of the right boundary indicates the last byte of the corresponding buffered segment.

15. A message receiving apparatus based on a transmission control protocol TCP, the apparatus comprising:

the first determining module is used for determining that the current message segment is a current disordered message segment if the current message segment is not the s-th sequence message segment to be received; wherein s is a positive integer;

The second determining module is used for determining whether the current disordered message segment and the cached message segment meet the preset condition or not if N cached message segments exist, and obtaining a determining result; wherein, the cached message segment is: historical disordered message segments; the meeting of the preset conditions between the current disordered message segment and the cached message segment comprises the following steps: the current disordered message segment and the cached message segment content are continuous or have overlapped content;

the caching module is used for determining whether to cache the current disordered message segment according to the determination result; the cache module is specifically configured to: and when the determined result shows that the current disordered message segment and the cached message segment do not meet the preset condition, discarding the message segment with the closest interval distance between the current disordered message segment and the cached message segment and the s-th sequence message segment.

16. An electronic device, the electronic device comprising:

a memory;

a processor, coupled to the memory, for enabling the implementation of the method provided in any of claims 1 to 14 by executing computer-executable instructions stored on the memory.

17. A computer storage medium having stored thereon computer executable instructions; the computer-executable instructions, when executed by a processor, are capable of carrying out the method provided in any one of claims 1 to 14.