JP2013077931A - Network transmitter and program therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a network transmitter and a program therefor, capable of transmission with good response even to communication requiring immediacy, while maintaining communication efficiency by aggregation in network data transmission.SOLUTION: A separation of transmission data set by an application is determined in a MAC layer, and if there is a separation, transmission is performed immediately. Consequently, a transmission data not requiring immediacy is transmitted using ordinary aggregation, whereas a data requiring immediacy can be transmitted immediately. As a result, it is possible to suppress a deteriorated response to the application, which is a drawback of aggregation, while maintaining a merit by the aggregation.

Description

  The present invention relates to a transmitter in a computer network (hereinafter simply referred to as a network), and more particularly to an apparatus capable of performing efficient transmission.

  Needless to say, improving the efficiency of data transmission and reception over a network is an important technical theme. Naturally, the approach to this theme has been made since the beginning of the network. In recent years, wireless networks with a narrower bandwidth than the wired network have been rapidly spreading, and the demand for higher efficiency has increased compared to the past. In view of such a situation, what is called aggregation (or frame aggregation) is used as means for improving the efficiency of data transmission and reception in a wireless network (Non-Patent Document 1).

Cisco Systems G.K. <http://www.cisco.com/web/JP/solution/netsol/mobility/ngw/literature/pdf/wlpfmc_wp.pdf>

  In simple terms, aggregation is the transmission of data in a wireless network by bundling (aggregating) a plurality of transmission frames, which are the smallest unit of data, to control communications associated with data transmission. Try to reduce. This realizes efficient data transmission.

  FIG. 1 shows a specific example of the above-described aggregation operation, particularly in the transmitter. The applications (TCP application and UDP application) in the figure are, for example, TCP (Transmission Control Protocol) and UDP (User Datagram Protocol) applications, and specifically, HTTP (Hyper Text Transfer Protocol) and SNMP (Simple Network Management). Protocol).

  FIG. 1 illustrates a case where the application 1 and the application 2 transmit data. As shown in the figure, application 1 has two transmission frames and application 2 has three transmission frames. In the case of a normal transmitter, a transmission frame is transmitted in units such as application 1-1 and application 2-1. However, when aggregation is used, the arrival of the transmission start timing every predetermined time is used as a trigger. A plurality of transmission frames are transmitted together. In the case of FIG. 1, transmission is not performed until five transmission frames are collected. In this way, communication associated with data transmission can be reduced as much as possible. In addition, the aggregation is usually performed in the MAC layer (data link layer), and is performed independently of the upper layers (transport layer and network layer) to which TCP or UDP belongs.

  On the other hand, data transmission using aggregation sometimes causes a decrease in communication response. In the case of FIG. 1 described above, the application 1 is not transmitted until the transmission start timing arrives even though there are only two transmission frames. In other words, the transmission frame before that frame is continuously held in the transmitter. If the transmission frame from the application 1 requires immediacy, such an operation may cause a decrease in the response of the application 1, and possibly cause a serious situation.

  The present invention solves this problem. That is, it is to provide a transmitter capable of transmitting data with high response to a communication requiring immediacy while maintaining efficiency of communication by aggregation in data transmission in a network, and a program thereof.

  A first form according to the present invention is a transmitter having a processing structure of at least two layers of an upper layer processing unit and a lower layer processing unit in a computer network, and transmitting data based on the processing structure, The upper layer processing unit, in the receiver that is the other party of communication, sets the communication settings (communication settings) that are intended to be used by the upper layer processing unit of the receiver at the same layer as the transmission data. Communication setting means for performing transmission, and transmission data transmitting means for passing the transmission data subjected to communication setting to a lower layer than itself, and the lower layer processing unit receives transmission data passed from a layer higher than itself. Transmission data receiving means, transmission data holding means for holding the received transmission data in a predetermined area, and transmission for transmitting the transmission data held in the predetermined area to the receiver. Means, a first judgment means for judging whether or not a predetermined condition managed only in the lower layer processing unit which is itself is satisfied, and if the predetermined condition is satisfied, transmission data to the transmission means First transmission instruction means for instructing transmission, second determination means for determining whether or not the communication setting performed by the communication setting means is a predetermined communication setting, and the communication setting is a predetermined communication setting A second transmission instruction unit that instructs the transmission unit to transmit the transmission data.

  Preferably, the lower layer processing unit is a processing unit that executes processing in the data link layer, the predetermined condition is arrival of a predetermined time, and the predetermined communication setting indicates a delimiter of transmission data in the upper layer processing unit Is.

  A second form according to the present invention is a program for causing a transmitter in a computer network to function as an upper layer processing means and a lower layer processing means, and at least the receiving party that is the other party of communication in the upper layer processing means. In the machine, the transmission data is set for communication (communication setting) that is intended to be used by the upper layer processing means of the receiver that is at the same level as the transmission data. A step of passing to a lower layer is executed, and the lower layer processing means receives at least the transmission data passed from the higher layer than itself and holds the received transmission data in a predetermined area. When it is judged whether or not a predetermined condition managed only by a certain lower layer processing means is satisfied, and the predetermined condition is satisfied When transmission data held in a predetermined area is transmitted to a receiver, it is determined whether the communication setting applied to the transmission data is a predetermined communication setting, and the communication setting is a predetermined communication setting. If determined, the program executes a step of transmitting transmission data held in a predetermined area to a receiver.

  Preferably, the lower layer processing means is means for executing processing in the data link layer, the predetermined condition is arrival of a predetermined time, and the predetermined communication setting indicates a delimiter of transmission data in the upper layer processing means It is.

  In the present invention, the transmission data set by the application is judged at the MAC layer, and if there is a break, it is transmitted immediately. As a result, transmission data that does not require immediacy can be transmitted by normal aggregation, and data that requires immediacy can be transmitted immediately. As a result, it is possible to suppress a decrease in response to the application, which is a drawback of aggregation, while maintaining the advantages of aggregation.

Embodiments according to the present invention will be described below with reference to the drawings.
[Example 1]
[System overview]

FIG. 2 is an overall view of a system assumed by the present invention. The PC 201 and the wireless access point 202 are network-connected wirelessly, and the access point 202 and the PC 203 are network-connected by wire. The PC 201 can communicate with the PC 203 via the access point 202. Hereinafter, a case where the PC 201 transmits data to the PC 203 will be described as an example.
[Network stack structure]

FIG. 3 shows a network stack structure in the PC 201. From the top, the application layer, TCP / UDP layer (transport layer), IP layer (network layer), MAC layer (data link layer), and physical layer are configured. The application layer is an application that uses a protocol of the network layer, and is, for example, HTTP or SNMP as described in the background art. Since this stack structure is general, detailed description is omitted. The embodiment relates to a case where the PC 201 transmits data to the PC 203, and the case where the application uses TCP or UDP will be described.
[TCP header]

  FIG. 4 shows details of the TCP header. In this embodiment, information included in the TCP header is analyzed in the MAC layer, and it is determined whether or not to perform frame transmission by aggregation. Originally, the information of the TCP header set on the transmission side is used only in the TCP on the reception side, but in the present invention, it is also used in the MAC layer on the transmission side.

As shown in FIG. 4, the TCP header contains various pieces of information in segment units. In this embodiment, PSH in the code bit segment is used. PSH is a push flag, and is a bit indicating whether or not received data should be immediately passed to a higher-level application. Whether or not this bit is turned on is set by the TCP on the transmitter side based on an instruction from the upper application. That is, it is instructed whether or not the TCP, SNMP, etc. are turned on as necessary. On the other hand, when the PSH of the receiver TCP is ON, the received data is immediately passed to the application.
[Transmitter Flow (Main Task)]

  FIG. 5 is an operation flow of the transmitter according to the present embodiment. As shown in the drawing, the operations of both layers of the TCP layer (steps 501 to 504) and the MAC layer (steps 505 to 508) are collectively shown. The transmitter is composed of a main task shown in the figure and an aggregation monitoring task shown in FIG. 6 later. First, the main task will be explained.

  In step 501, TCP data is received from an application.

  In step 502, it is determined whether there is an instruction to turn on PSH for the received TCP data. If there is an instruction, the process proceeds to step 503, and if not, the process proceeds to step 504.

  In step 503, PSH is turned on and TCP data is passed to the MAC layer, while in step 504, it is turned off and passed.

  In step 505, it is determined whether the data passed from the TCP layer is TCP. If TCP, the process proceeds to step 506, and if not, the process proceeds to step 508.

  In step 506, it is determined whether PSH of the TCP header is ON. If ON, the process proceeds to step 507, and if not, the process proceeds to step 508.

  In step 507, an instruction to send the transferred TCP data is issued to an aggregation monitoring task to be described later.

In step 508, the transferred TCP data is aggregated. That is, TCP data is buffered in a predetermined area.
[Transmitter Operation Flow (Aggregation Monitoring Task)]

  The aggregation monitoring task described below is for monitoring the transmission start timing of a transmission frame and transmitting the aggregated frame when the timing comes.

  In step 601, it is determined whether a predetermined time has elapsed. That is, it is determined whether or not the transmission start timing has arrived. As a result of the determination, if the timing has arrived, the process proceeds to step 603, and if not, the process proceeds to step 602.

  In step 602, it is determined whether a transmission instruction from the main task has been received. As a result of the determination, if it is received, the process proceeds to step 603, and if not, the process returns to step 601.

  In step 603, the transmission frame buffered in a predetermined area is transmitted.

As described above, two triggers for transmitting a frame, one based on normal aggregation and one that detects that PSH is ON, are used. Therefore, while maintaining the advantages of aggregation, the trigger is immediately used as needed. Can send frames.
[Another embodiment of the first embodiment]

  In the first embodiment, by analyzing the PSH segment of the TCP header in the MAC layer, the frame can be transmitted immediately even during the aggregation. In addition to PSH, the TCP header includes a segment that can include an intention of immediate transmission from a higher-level application. This is the reserved area shown in FIG.

  The reserved area is a segment whose use is not determined at present, and is normally set to zero indicating that it is invalid. As another form of the first embodiment, this reserved area may be used to notify the MAC layer that transmission data is not aggregated. In that case, it is necessary to determine the meaning of the bits to be set in the reserved area between the upper application and TCP. For example, when bit 0 is 1, transmission is performed.

In the MAC layer, the reserved area in which the bit is set may be analyzed to determine whether to aggregate or transmit the frame immediately.
[Example 2]

  In the first embodiment, by analyzing the PSH segment of the TCP header in the MAC layer, the frame can be transmitted immediately even during the aggregation. The present invention is naturally applicable even if the transport layer is other than TCP, and in this embodiment, an example of using UDP is shown.

  The UDP header does not have a segment corresponding to TCP PSH. For this reason, when the upper application transmits UDP data, a segment included in the header of the IP layer located between the UDP layer and the MAC layer is used.

  FIG. 8 shows details of the IP header. As shown in the figure, a flag segment is provided in the IP header. The flag is a bit for controlling the division of the IP packet, and its meaning is as shown in FIG.

  Bit 0 is a reserved bit and is not normally used. Bit 1 indicates whether or not division is possible. 0 is divisible and 1 is not divisible. Bit 2 indicates whether or not it is the end of the divided packet. 0 indicates the end of the divided packet and 1 indicates the middle of the divided packet. In this embodiment, it is determined whether or not to immediately transmit a frame by analyzing bit 2 in the MAC layer.

  FIG. 10 is an operation flow of the transmitter according to the present embodiment. Similarly to the first embodiment, the flow of this embodiment is composed of a main task and an aggregation monitoring task. Since the aggregation monitoring task is the same as that of the first embodiment, the description thereof is omitted.

  In step 1001, UDP data is received from the application.

  In step 1002, it is determined whether or not the IP is divided and terminated. This is determined based on an instruction from the upper application. If the result of determination is division and termination, processing proceeds to step 1003, and otherwise processing proceeds to step 1004.

  In step 1003, bit 2 of the flag field is set to 0. On the other hand, in step 1004, bit 2 of the flag field is set to 1. Steps 1003 and 1004 are actually performed in the IP layer.

  In step 1005, it is determined whether the data passed from the upper layer is UDP. As a result of the determination, if it is UDP, the process proceeds to step 1006, and if not, the process proceeds to step 1007.

  In step 1006, the IP flag field is analyzed to determine whether bit 2 is 0 or not. If the result of the determination is 0, the process proceeds to step 1007, and if not, the process proceeds to step 1008.

  In step 1007, a transmission instruction is issued to the aggregation monitoring task. That is, it is instructed to transmit immediately without being aggregated.

In step 1008, the frames are aggregated. That is, the UDP data is buffered in a predetermined area.
[Aggregation monitoring task in all examples]

  The aggregation operation of the aggregation monitoring task in the first embodiment (including another form) and the second embodiment described so far transmits a buffered frame triggered by arrival of a predetermined time. However, the frame transmission trigger is not limited to this, and other frames can be used. For example, when a predetermined number of frames are buffered. It is also possible to combine this with the arrival of a predetermined time. [Summary]

In the present invention, the transmission data set by the application is judged at the MAC layer, and if there is a break, it is transmitted immediately. As a result, transmission data that does not require immediacy can be transmitted by normal aggregation, and data that requires immediacy can be transmitted immediately. As a result, it is possible to suppress a decrease in response to the application, which is a drawback of aggregation, while maintaining the advantages of aggregation.

Aggregation operation at transmitter Overall system diagram Network stack structure TCP header Operation flow of transmitter according to embodiment 1 Operation flow of the aggregation monitoring task TCP header IP header Meaning of flags in IP header Operation flow of transmitter according to embodiment 2

503 Communication setting for transmission data 504 Communication setting for transmission data 506 Communication setting determination 507 Transmission instruction 508 Aggregation 601 Determination of elapse of predetermined time 602 Transmission instruction reception determination 603 Transmission

Claims (4)

In a computer network, a transmitter having a processing structure of at least two layers of an upper layer processing unit and a lower layer processing unit and transmitting data based on the processing structure,
(A) The upper layer processing unit
(A-1) In the receiver that is the other party of communication, a setting (communication setting) related to communication that is intended to be used by the upper layer processing unit of the receiver that is at the same level as itself is applied to the transmission data. Communication setting means;
(A-2) transmission data transmission means for passing transmission data set for communication to a lower layer than itself;
With
(B) The lower layer processing unit
(B-1) transmission data receiving means for receiving transmission data passed from a layer higher than itself;
(B-2) transmission data holding means for holding the received transmission data in a predetermined area;
(B-3) transmitting means for transmitting the transmission data held in the predetermined area to the receiver;
(B-4) first determination means for determining whether or not a predetermined condition managed only in the lower layer processing unit that is itself is satisfied;
(B-5) a first transmission instruction means for instructing the transmission means to transmit transmission data when the predetermined condition is satisfied;
(B-6) second determination means for determining whether or not the communication setting performed by the communication setting means is a predetermined communication setting;
(B-7) a second transmission instruction means for instructing the transmission means to transmit transmission data when it is determined that the communication setting is a predetermined communication setting;
Transmitter with. The lower layer processing unit is a processing unit that executes processing in the data link layer,
The predetermined condition is the arrival of a predetermined time,
The transmitter according to claim 1, wherein the predetermined communication setting indicates a delimiter of transmission data in the upper layer processing unit. A transmitter in a computer network,
Upper layer processing means;
A program that functions as a lower layer processing means,
In the upper layer processing means, at least,
At the receiver that is the other party of communication, the transmission data is subjected to settings (communication settings) that are intended to be used by the upper layer processing means of the receiver that is at the same level as the receiver.
A step of passing the transmission data with communication settings to a lower layer than itself is executed,
In the lower layer processing means, at least,
Receives transmission data passed from a layer higher than itself,
Hold the received transmission data in a predetermined area,
Determine whether a predetermined condition managed only by the lower-layer processing means that is itself is satisfied,
When the predetermined condition is satisfied, the transmission data held in the predetermined area is transmitted to the receiver,
Determining whether the communication setting applied to the transmission data is a predetermined communication setting;
A program for executing a step of transmitting transmission data held in the predetermined area to a receiver when the communication setting is determined to be a predetermined communication setting. The lower layer processing means is means for executing processing in the data link layer,
The predetermined condition is the arrival of a predetermined time,
The program according to claim 3, wherein the predetermined communication setting indicates a delimiter of transmission data in the upper layer processing means.