WO2012032606A1 - フレーム連結装置 - Google Patents
フレーム連結装置 Download PDFInfo
- Publication number
- WO2012032606A1 WO2012032606A1 PCT/JP2010/065314 JP2010065314W WO2012032606A1 WO 2012032606 A1 WO2012032606 A1 WO 2012032606A1 JP 2010065314 W JP2010065314 W JP 2010065314W WO 2012032606 A1 WO2012032606 A1 WO 2012032606A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- frame
- frames
- unit
- concatenated
- mac
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L69/00—Network arrangements, protocols or services independent of the application payload and not provided for in the other groups of this subclass
- H04L69/22—Parsing or analysis of headers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
- G06F13/38—Information transfer, e.g. on bus
- G06F13/382—Information transfer, e.g. on bus using universal interface adapter
- G06F13/385—Information transfer, e.g. on bus using universal interface adapter for adaptation of a particular data processing system to different peripheral devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2213/00—Indexing scheme relating to interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
- G06F2213/38—Universal adapter
- G06F2213/3808—Network interface controller
Definitions
- the present invention relates to a frame connecting device.
- data center that manages multiple servers and storage, and provides connection lines to the Internet and maintenance / operation services.
- L2 Layer 2
- top-of-rack switch In the data center, multiple servers and multiple storages are installed in multiple racks.
- a Layer 2 (L2) switch called a “top-of-rack switch” is installed at the entrance on the communication line of each rack.
- Each top-of-rack switch is connected to a plurality of servers or a plurality of storages accommodated in a rack via downlink lines.
- each top-of-switch is connected to a router connected to the Internet via an uplink line.
- a plurality of virtual machines (VMs) are arranged in one or more physical servers accommodated in a rack, and the operation efficiency of the data center can be increased in proportion to the operating rate of the physical servers and VMs. For this reason, it is desirable not to create an empty port in the top-of-rack switch.
- the downlink side is 1 Gbps x 44 to 48 ports, and the uplink side is 10 Gbps x 4 ports.
- the traffic volume on the downlink side exceeds the traffic volume on the uplink side and non-blocking communication is not performed.
- the above problem is a problem that may occur when the maximum traffic volume on the upstream side exceeds the maximum traffic volume on the downstream side in the relay device as well as the L2 switch used as the above-described top-of-rack switch.
- it is generally desired to apply a communication method with high data transmission efficiency.
- One embodiment of the present invention has been made in view of the above problems, and an object thereof is to provide a technique capable of efficiently transmitting a frame.
- One aspect of the present invention is a storage unit that stores a plurality of frames to be transmitted; A plurality of frames that are read from the storage unit are connected in series, and a generation unit that generates a connected frame to which frame length information of each frame is attached; A frame concatenation apparatus comprising: a concatenation unit that appends information for establishing synchronization with the reception side of the concatenated frame to the concatenated frame.
- frames can be transmitted efficiently.
- FIG. 2 shows a configuration example of a network system including a data center.
- the Ethernet (registered trademark) Media Access Control (MAC) frame format is shown.
- An example of a MAC address table is shown.
- the structural example of L2 switch is shown.
- the structural example of the MAC processing part by the side of an egress (Egress) is shown.
- the format of a concatenated frame generated by the “different DA concatenation” method (fourth processing) is shown.
- the format of the Length field included in the concatenated frame shown in FIG. 6 is shown. It is a table
- the processing flowchart of the MAC processing part by the side of Egress is shown.
- the structural example of the MAC processing part by the side of an entrance (Ingress) is shown.
- the processing flowchart on the Ingress side is shown.
- the format of a concatenated frame generated by the “DA concatenation” method (first processing) is shown.
- the format of a concatenated frame generated by the “VID concatenation” method (second processing) is shown.
- the format of a concatenated frame generated by the “VID and DA concatenation” method is shown.
- the band usage rate of a physical line when connecting MAC frames is shown.
- the horizontal axis indicates the frame size, and the vertical axis indicates the band usage rate.
- the number of frames to be connected is 10.
- the change in the bandwidth usage rate of the physical line according to the number of connected MAC frames is shown.
- the horizontal axis indicates the number of connections, and the vertical axis indicates the band usage rate.
- the frame size is 64 bytes.
- the structural example of the exit side MAC process part in 2nd Embodiment is shown.
- the data structure example of the Length field in 2nd Embodiment is shown. It is a flowchart which shows the operation example of the exit side MAC process part in 2nd Embodiment.
- the structural example of the entrance side MAC process part in 2nd Embodiment is shown. It is a flowchart which shows the operation example of the entrance side MAC process part in 2nd Embodiment.
- FIG. 1 shows a configuration example of a network system including a data center in the embodiment.
- the data center 10 has an area called a rack that accommodates a plurality of servers and storage.
- a server rack 11 in which a plurality of servers (server machines) 1 are accommodated and a storage rack 12 in which computers (server machines) that function as the plurality of storages 2 are accommodated are illustrated.
- a plurality of virtual machines (VMs) can be deployed in a server machine that functions as the server 1 or the storage 2, that is, a physical server. That is, one virtual machine can operate as a computer (server machine) that functions as the server (server machine) 1 or the storage 2.
- the server rack 11 is provided with an L2 switch 13 that functions as a top-of-rack switch connected to each of the plurality of server machines 1 via a communication line (downlink).
- the storage rack 12 is provided with an L2 switch 14 that functions as a top-of-rack switch connected to each of the plurality of storages 2 via a communication line (downlink).
- Each L2 switch 13 and 14 is connected to an L2 switch 15 that functions as an inter-rack switch via a communication line (uplink).
- the L2 switch 15 is connected to the Internet IN via a dedicated communication device (security appliance 16) that integrates a security function such as a firewall, virus gateway, and spam filter and a router function called a "security appliance”. Has been.
- An end user (terminal device) 17 connected to the Internet IN can access the server 1 and the storage 2 in the data center 10 via the Internet IN, and can receive provision of desired network services and data. .
- the L2 switches 13, 14, and 15 relay the MAC frame.
- the L2 switches 13 and 14 shown in FIG. 1 have 1 Gbps ⁇ 44 ports on the downlink side and 10 Gbps ⁇ 4 ports on the uplink side.
- the L2 switch 15 has 10 Gbps ⁇ 8 ports on the downlink side and 10 Gbps ⁇ 4 ports on the uplink side.
- FIG. 2 shows a standard MAC frame format.
- Inter Frame Gap IGF
- a signal called preamble / start frame delimiter (SFD) is stored in an area (field) of 8 bytes from the head of the MAC frame.
- the preamble / SFD is used to detect the head of a MAC frame when synchronization is established in MAC transmission.
- the preamble is a signal in which the signal of the bit string “10101010” is continued for 7 bytes.
- SFD is a 1-byte signal of the bit string “10101011”.
- DA destination address
- SA source address
- a Type / Length field (2 bytes) is provided.
- the value of the type / length field represents the user data length (Length).
- the value of the type / length field indicates the protocol type of the upper protocol.
- the main value (protocol type) represented by the type (Type) includes 0x0800 (IP), 0x0806 (ARP), and 0x8100 (802.1Q Tagged VLAN).
- the user data of the upper layer of the MAC is stored in the data section (data field) next to the type / length field.
- the length of the data part is 46 to 1500 bytes. If the length of the data part is less than 46 bytes, the padding (PAD) is adjusted to 46 bytes.
- FEC Frame Check Sequence
- CRC Cyclic Redundancy Check
- the MAC device that controls the MAC frame monitors the preamble and SFD in the received signal, and when detecting the preamble and SFD, detects the MAC frame including the preamble and SFD.
- the MAC device performs FCS check (CRC check) on the detected MAC frame.
- the MAC device detects the port number and the source MAC address in which the MAC frame has entered, learns that the source MAC address exists at the port, and registers it in the MAC address table (MAC address) Learning).
- the MAC device finds the destination port of the MAC frame from the MAC address table that has already been constructed by learning. When the destination port is found, the MAC device sends out the MAC frame from the destination port (MAC forwarding). On the other hand, when the destination port is not found, the MAC device transmits the MAC frame to the network (all ports) where the MAC frame exists (flooding).
- the MAC address table is updated when the device is replaced.
- aging information on the MAC table after a certain period has been deleted.
- FIG. 3 shows an example of the MAC address table.
- the MAC address is simplified and expressed by one character.
- a device with MAC address A exists for port number 1.
- the frame of destination A is transmitted to port 1.
- a newly learned combination of a MAC address and a port number is added to the MAC address table.
- the MAC frame is transmitted to all ports on the same network as described above.
- each L2 switch 13 and 14 is in a state of transmitting 44 Gbps data (MAC frame) flowing in from the downlink side to the uplink side at 40 Gbps. That is, data relay is not performed in a non-blocking manner.
- an interframe gap (IFG) of 12 bytes, and a preamble and SFD of 8 bytes are inserted for each transmitted MAC frame.
- IFG interframe gap
- preamble and SFD consume a band to be used for transmitting an actual MAC frame (DA, SA, Type / Length,, and data portion).
- the transmission efficiency for the physical bandwidth of the MAC frame is 64 / (64 + 12 + 8) ⁇ 0.76.
- the conventional Ethernet (registered trademark) system MAC frame format
- FIG. 4 shows a configuration example of the L2 switch. 4, the L2 switch 20 is applicable as the L2 switches 13, 14, and 15 shown in FIG.
- the L2 switch 20 includes a Central Processing Unit (CPU) 21, an SDRAM 22, and a flash memory (Flash MEM) 23.
- CPU Central Processing Unit
- SDRAM Secure Digital RAM
- Flash MEM flash memory
- the L2 switch 20 includes a CPU 21, an SDRAM 22, a flash memory 23 and a MAC switch 24 connected via the bus B, and a plurality of PHY chips (PHY devices) 25 and a MAC chip (MAC) connected to the MAC switch 24.
- Device) 26 sets.
- a set of a plurality of PHY chips 25 and a MAC chip 26 is divided into a downlink side and an uplink side, and each can accommodate one or more ports. Further, each MAC chip 26 is connected to the CPU 21 via the bus B.
- the CPU 21 controls the entire operation of the L2 switch 20 by executing various programs such as an operating system (OS) stored in the flash memory 23 and a program for controlling the MAC chip 26.
- OS operating system
- the CPU 21 controls the operation of the MAC chip (MAC concatenation processing (described later)) particularly during MAC frame relay.
- the SDRAM 22 is used as a work area for the CPU 21.
- the flash memory 23 functions as a boot memory used for L2 switch activation processing, and stores various programs such as an OS and data used when the programs are executed.
- the PHY chip 25 manages processing related to layer 1 (physical layer) in the OSI reference model.
- the MAC chip 26 manages processing related to the MAC layer.
- the MAC switch 24 performs processing (switching) to transfer the MAC frame input from each MAC chip 26 to the MAC chip 26 corresponding to the destination. For this switching, the MAC address table described above can be used.
- the CPU 21 can issue an instruction to the MAC chip 26 such as a connection method to be used and a frame size to be connected.
- the MAC chip 26 includes an egress-side MAC processing unit 30 (FIG. 5) and an egress-side (ingress) -side MAC processing unit 40 (FIG. 10).
- the MAC processing unit 30 on the egress side performs processing related to the MAC frame that flows in the MAC frame transmission direction (MAC ⁇ PHY).
- the MAC processing unit 40 on the ingress side performs processing related to the MAC frame flowing in the MAC frame reception direction (PHY ⁇ MAC).
- FIG. 5 shows a configuration example of the MAC processing unit on the exit side.
- the egress-side MAC processing unit 30 concatenates a plurality of MAC frames and performs MAC frame concatenation.
- the MAC processing unit 30 on the egress side includes an FCS check unit 31, a frame length check unit 32, a frame length insertion unit 33, a connection type selection unit 34 (hereinafter referred to as selection unit 34), and a connection type insertion unit 35 (hereinafter referred to as “connection type insertion unit”).
- FCS check unit 31 a frame length check unit 32, a frame length insertion unit 33, a connection type selection unit 34 (hereinafter referred to as selection unit 34), and a connection type insertion unit 35 (hereinafter referred to as “connection type insertion unit”).
- Type insertion unit 35 Type insertion unit 35
- deletion unit 36 duplicate byte deletion unit 36
- concatenation frame generation unit 37 hereinafter referred to as generation unit 37
- FCS insertion unit 38 FCS insertion unit
- the type insertion part 35 includes a DA connection type insertion part 35A, a VID connection type insertion part 35B, a VID and DA connection type insertion part 35C, and a different DA connection type insertion part 35D. Further, management tables T1 to T4 are provided between the type insertion units 35A to 35D and the deletion unit 36. The management tables T1 to T4 are examples of storage units.
- the MAC frame transmitted from the MAC switch 24 is input to the FCS check unit 31.
- the FCS check unit 31 detects the MAC frame by detecting the preamble and SFD of the MAC frame, and checks the FCS of the detected MAC frame.
- the frame length check unit 32 checks (calculates) the frame length (DA, SA, Type / Length, data portion, and FCS length) of the MAC frame.
- the frame length insertion unit 33 inserts the calculated frame length (referred to as “frame length bit string”) into the MAC frame.
- the selection unit 34 selects a MAC frame concatenation method (concatenation type) based on an instruction from the CPU 21 (concatenation method designation), and sends the MAC frame to any of the type insertion units 35A to 35D according to the concatenation method. send.
- connection types there are four types of connection methods (connection types): “DA connection”, “V-LAN (ID (VID) connection”, “VID and DA connection”, and “different DA connection” (each type). Will be described later).
- Each type insertion unit 35A to 35D inserts data (referred to as “connection type bit string”) indicating the selected connection method (connection type) into the MAC frame.
- Each type insertion unit 35A to 35D has a management table (buffer) for temporarily storing MAC frames.
- the MAC frame in which the concatenation type byte is inserted is temporarily stored in the corresponding management tables T1 to T4.
- the deletion unit 36 In response to a request from the generation unit 37, the deletion unit 36 reads a plurality of MAC frames to be concatenated from one of the management tables (buffers) T1 to T4. The deletion unit 36 deletes bytes (data, that is, DA and / or VID) that are duplicated among the plurality of read MAC frames, and sends them to the generation unit 37.
- the generation unit 37 generates a concatenation frame (concatenated frame) with a frame size according to an instruction from the CPU 21 based on the control from the CPU 21.
- the generation unit 37 monitors the total size (buffer storage amount) of the MAC frames stored in each of the management tables T1 to T4, and the buffer storage amount is the concatenated frame.
- the deletion unit 36 is instructed to read out the MAC frame from the corresponding management table (concatenation type: other than different DA), or the generation unit directly from the management table T4
- the MAC frame may be read (concatenation type: different DA).
- the deletion unit 36 or the generation unit 37 may periodically read out the MAC frames in the management tables T1 to T4 at a predetermined cycle.
- the FCS insertion unit 38 performs CRC calculation on the concatenated frame generated by the generation unit 37, inserts the calculation result as FCS, and then transmits the concatenated frame toward the PHY chip 25.
- Fig. 6 shows the format of a concatenated frame.
- the concatenated frame format shown in FIG. 6 indicates the format of the concatenated frame when the concatenation type is “different DA”. Specifically, this concatenated frame has the following data structure.
- the preamble and SFD related to the remaining MAC frames are omitted except for the preamble and SFD of the first MAC frame among the plurality of MAC frames to be connected.
- a “Length” field (2 bytes), which is a new field, is inserted after the Type / Length field of the portion corresponding to each of the concatenated MAC frames (in FIG. 6, an asterisk).
- the field marked with (*) is the “Length” field).
- the “Length” field stores the above-described frame length bit string and concatenated type bit string.
- FIG. 7 shows an example of the data structure of the “Length” field.
- a bit string (referred to as a concatenation code) indicating a concatenation type (concatenation method) is expressed by 2 bits and stored in 2 bits (bits b0 and b1) from the head of the “Length” field.
- the frame length bit string is expressed by the remaining 14 bits (bits b2 to b15). Therefore, a frame length of 16384 bytes can be expressed.
- the concatenation type bit string (concatenation code: bits b0, b1) represents a concatenation method (concatenation method).
- FIG. 8 is a correspondence table between a concatenated code expressed by bits b0 and b1 and a concatenation method corresponding to the concatenated code.
- connection methods described in the embodiment (1) “Different DA concatenation” method of concatenating MAC frames having different destination addresses (DA). (2) “DA concatenation” method of concatenating MAC frames having the same DA. (3) “VID concatenation” method of concatenating MAC frames having the same VID. (4) “VID and DA concatenation” method for concatenating MAC frames having the same VID and DA.
- the concatenated code “00” (different DA concatenation method) is inserted into the MAC frame by the different DA concatenation type insertion unit 35D
- the concatenation code “01” (DA concatenation method) is inserted into the MAC frame by the DA concatenation type insertion unit 35A. Inserted into.
- the concatenation code “10” (VID concatenation method) is inserted into the MAC frame by the VID concatenation type insertion unit 35B
- the concatenation code “11” (VID and DA concatenation method) is inserted by the VID and DA concatenation type insertion unit 35C. It is inserted into the MAC frame.
- the concatenated frame has the same format as the standard MAC frame. Therefore, it is possible to manage with the same MAC address table as the MAC address table applied to the standard MAC frame. This means that the concatenated frame can be relayed to a general-purpose L2 switch and used end-to-end.
- the number of bytes in the “Length” field, the number of bits of the concatenated type bit string (code), and the number of bits of the frame length described above are examples, and the number of types of concatenation method and the number of representations are appropriately determined according to the desired field length. It can be changed.
- FIG. 9 is a flowchart illustrating an operation example of the MAC processing unit 30 on the exit side. The process illustrated in FIG. 9 is started when a MAC frame is input to the MAC processing unit 30.
- step S1 the FCS check unit 31 detects a MAC frame and executes an FCS check (CRC check). At this time, if the result of the FCS check is an error, the frame is discarded (step S2). On the other hand, if the result of the FCS check is “OK”, the MAC frame is passed to the frame length check unit 32.
- FCS check CRC check
- the frame length check unit 32 checks (measures) the frame length of the MAC frame (step S3), and passes the frame length bit string indicating the measured frame length and the MAC frame to the frame length insertion unit 33.
- the frame length insertion unit 33 inserts a frame length bit string at a predetermined position (between the Type / Length field and the data portion) of the MAC frame (step S4), and passes the MAC frame to the selection unit 34.
- the selection unit 34 passes the MAC frame to any one of the connection type insertion units 35A to 35D according to the instruction from the CPU 21. That is, the connection type insertion unit 34 determines whether or not the instructed connection type is “DA connection” (step S5).
- connection type is “DA connection”
- selection unit 34 advances the process to step S10.
- the selection unit 34 advances the process to step S6.
- the selection unit 34 determines whether or not the connection type is “VID connection”. At this time, if the connection type is “VID connection”, the selection unit 34 advances the process to step S13. On the other hand, if the connection type is not “VID connection”, the selection unit 34 advances the process to step S7.
- the selection unit 34 determines whether or not the connection type is “VID and DA connection”. At this time, if the connection type is “VID and DA connection”, the selection unit 34 advances the process to step S16. On the other hand, if the connection type is not “VID and DA connection”, the selection unit 34 advances the process to step S8.
- step S8 a concatenated frame is generated in accordance with the “different DA concatenation” method. That is, in step S8, the MAC frame is transferred from the selection unit 34 to the different DA connection type insertion unit 35D.
- the different DA concatenation type insertion unit 35D inserts the code “00” (FIG. 8) indicating the different DA concatenation method into a predetermined position (concatenated code insertion position) in the MAC frame and stores it in the management table (buffer) T4. .
- a plurality of connection target MAC frames stored in the management table T 4 are read out and passed to the generation unit 37.
- the generation unit 37 generates a concatenated frame (FIG. 6) according to the above-described different DA concatenation method from the received plurality of MAC frames, and passes it to the FCS insertion unit 38. Then, the process proceeds to step S9.
- step S9 the FCS insertion unit 38 performs CRC calculation for the concatenated frame and stores the calculation result in the FCS field. Thereafter, the connection frame is output to the outside line and sent to the corresponding PHY chip 25 (FIG. 4). The processing after step S10 will be described later.
- the generation unit 37 attaches a preamble and SFD as synchronization establishment information to the head of the concatenated frame.
- an adding unit for adding the preamble and SFD may be provided in the subsequent stage of the generating unit 37.
- the SFD assignment and the preamble assignment for the connection frame may be performed by different entities (components).
- FIG. 10 shows a configuration example of the MAC processing unit 40 on the entrance side.
- the MAC processing unit 40 releases the MAC frame concatenation, and restores each MAC frame before connection.
- the MAC processing unit 40 includes an FCS check unit 41, a concatenation type determination unit 42, a concatenated frame division unit 43, a duplicate byte restoration unit 44 (hereinafter, restoration unit), and a frame length deletion unit 45 ( Hereinafter, a deletion unit 45), an SFD insertion unit 46, and an FCS insertion unit 47 are provided.
- the dividing unit 43 includes a DA dividing unit 43A corresponding to the “DA concatenation” method, a VID dividing unit 43B corresponding to the “VID concatenation” method, a DA and VID dividing unit 43C corresponding to the “VID and DA concatenation” method, , And a different DA division unit 43D corresponding to the “different DA connection” method.
- the FCS check unit 41 performs an FCS check (CRC check) on the concatenated frame received from the PHY chip 25 (FIG. 3).
- the connection type determination unit 42 determines the connection type of the connection frame. For example, the concatenation type determination unit 42 has an offset position of the first “Length” field from the head of the concatenated frame, and detects a concatenation code (bits b0 and b1) indicating the concatenation type based on the offset position.
- a concatenation code bits b0 and b1
- connection type determination unit 42 passes the connection frame to the different DA division unit 43D.
- connection code is “01 (DA connection)”
- connection type determination unit 42 passes the connection frame to the DA division unit 43A.
- connection code is “10 (VID connection)”
- the connection type determination unit 42 passes the connection frame to the DA division unit 43B.
- the connection code is “11 (VID and DA connection)”
- the connection type determination unit 42 passes the connection frame to the VID and DA division unit 43A.
- a connection frame dividing process corresponding to the connection type is performed.
- the concatenation type is “different DA”
- the different DA division unit 43D has a plurality of unit blocks in which the concatenated frame is a unit from the DA field to the field immediately before the next DA field (FCS field).
- the start position of the DA field of the MAC frame located at the head of the concatenated frame is the same as the start position of the DA field of the standard MAC frame, it can be obtained from the offset position of the DA field of the standard MAC frame.
- the start position (offset position on the concatenated frame) of the DA field in the second and subsequent MAC frame parts from the beginning is the frame length represented by the frame length bit string in the “Length” field included in the immediately preceding MAC frame equivalent part It is calculated from.
- the frame length of the “Length” field included in each MAC frame portion in the concatenated frame indicates the frame length of the MAC frame portion (length from DA to FCS).
- the start position of the DA field of the second MAC frame portion is obtained, the start position of the DA field of the head MAC frame portion is set as a starting point, and the portion corresponding to the frame length of the head MAC frame therefrom Is included in the head MAC frame portion. Then, the start position (offset position) of the DA field of the next (second) MAC frame is immediately after the FCS field included in the first MAC frame portion.
- the third The offset position of the DA field in the MAC frame portion can be obtained.
- the DA field of each MAC frame included in the concatenated frame is detected, and the concatenated frame can be divided at the start position of each DA field.
- the DA dividing unit 43A, the VID dividing unit 43B, the VID, and the DA dividing unit 43C divide the connection frame into a plurality of unit blocks according to the connection type (details will be described later).
- the duplicate byte restoring unit 44 restores a field (DA and / or VID field) omitted during the concatenation process for each of the plurality of unit blocks received from the division units 43A to 43C according to the concatenation type.
- the frame length deletion unit 45 deletes the “Length” field from each unit block received from the duplicate byte restoration unit 44 or the different DA division unit 43D.
- the SFD insertion unit 46 inserts the SFD into each unit block received from the frame length deletion unit 45.
- the FCS insertion unit 47 performs CRC32 calculation for each unit block received from the SFD insertion unit 46, and inserts the calculation result into the FCS field. In this way, the original MAC frame is restored.
- each unit block is sent as a MAC frame following the preamble (7 bytes) at every IFG interval.
- Each sent MAC frame is sent to the MAC switch 24.
- FIG. 11 is a flowchart illustrating an operation example of the MAC processing unit 40 on the entrance side. The process illustrated in FIG. 11 is started when a concatenated frame is input to the MAC processing unit 40.
- step S001 the FCS check unit 41 detects a concatenated frame and executes an FCS check (CRC check). At this time, if the result of the FCS check is an error, the frame is discarded (step S002). On the other hand, when the result of the FCS check is “OK”, the concatenated frame is passed to the concatenation type determination unit 42.
- FCS check CRC check
- connection type determination unit 42 performs a concatenation type check (step S003). That is, the connection type determination unit 42 detects a code indicating the connection type from the “Length” field based on, for example, the offset position of the first “Length” field in the connection frame held in advance, and the connection indicated by the code Recognize type.
- connection type determination unit 42 passes the connection frame to any of the division units 43A to 43D corresponding to the recognized connection type. That is, the connection type determination unit 42 determines whether or not the connection type is “DA connection” (step S004).
- connection type is “DA connection”
- selection unit 34 advances the process to step S013.
- the selection unit 34 advances the process to step S005.
- the selection unit 34 determines whether or not the connection type is “VID connection”. At this time, if the connection type is “VID connection”, the selection unit 34 advances the process to step S019. On the other hand, if the connection type is not “VID connection”, the selection unit 34 advances the process to step S006.
- the selection unit 34 determines whether or not the connection type is “VID and DA connection”. At this time, if the connection type is “VID and DA connection”, the selection unit 34 advances the process to step S025. On the other hand, if the connection type is not “VID and DA connection”, the selection unit 34 advances the process to step S007.
- step S007 a connection frame dividing operation according to the “different DA connection” method is performed. That is, the connection frame is transferred from the connection type determination unit 42 to the different DA division unit 43D.
- the different DA division unit 43D checks the field length of the “Length” field of the first MAC frame portion in the concatenated frame.
- the different DA division unit 43D separates (divides) the part from the head MAC frame part to the FCS field as a unit block according to the field length (step S008).
- the different DA division unit 43D determines whether or not the remaining part of the concatenated frame includes the next MAC frame part (step S009). At this time, if the next MAC frame portion remains, the process returns to step S007, the head “Length” field of the remaining portion is checked, the division position is determined, and the unit block is separated (step S008). . Such a loop process is performed until the next MAC frame portion disappears (NO determination in step S009).
- a plurality of unit blocks created by dividing the concatenated frame are passed to the frame length deleting unit 45.
- the frame length deletion unit 45 deletes the “Length” field included in each unit block (step S010).
- the frame length deletion unit 45 passes each unit block to the SFD insertion unit 46.
- the SFD insertion unit 46 inserts an SFD into each unit block (step S011).
- the SFD insertion unit 46 passes each unit block to the FCS insertion unit 47, and the FCS insertion unit 47 performs FCS (CRC32) calculation for each unit block and stores the calculation result in the FCS field (step S012).
- FCS FCS
- step S013 the SFD insertion unit 46 sends each MAC frame toward the MAC switch 24 following the preamble at an appropriate timing (timing according to IFG). The processing after step S013 will be described later.
- the MAC frame concatenation as described above is performed by uplink communication (L2 switch 13, 14 ⁇ L2 switch 15) and down link connecting each L2 switch 13, 14 and the upper L2 switch 15.
- the present invention can be applied to link communication (L2 switch 15 ⁇ L2 switches 13, 14).
- concatenated frame generation (MAC frame concatenation) process (exit side (transmission side) process
- frame length information of each MAC frame part included in the concatenated frame is inserted, and a concatenated code is inserted for each concatenation method.
- Duplicate byte deletion that can be omitted is performed, and concatenated frames are generated while suppressing the frame size within the maximum frame length.
- the inserted frame is divided into each frame (unit block) before concatenation based on the concatenation type of the received concatenated frame.
- the original frame is restored by deleting long bytes.
- SFD is not assigned to each MAC frame as in the prior art, but only one SFD is assigned to a concatenated frame in which a plurality of MAC frames are concatenated.
- IFG, preamble, and SFD (20 bytes) are omitted. That is, it is possible to avoid insertion of 8 bytes of SFD and preamble, which are information for establishing synchronization, for each MAC frame.
- MAC frame concatenation based on DA concatenation Next, MAC frame concatenation based on the “DA concatenation” method will be described. In the “DA concatenation” method, MAC frames having the same DA are collected and MAC frame concatenation is executed.
- connection type selection unit 34 follows the execution instruction of the “DA connection” method from the CPU 21.
- the MAC frame received from the frame length insertion unit 33 is sent to the DA concatenation type insertion unit 35A.
- DA concatenation type insertion unit 35A inserts a concatenation code “01” indicating “DA concatenation” at a predetermined position of the MAC frame and stores it in the management table T1.
- the management table (buffer) T1 one or more management tables (buffers) T1 for each DA, or different temporary storage areas for each DA are prepared.
- the deletion unit 36 reads a plurality of MAC frames (having the same DA) corresponding to the concatenated frame size (not exceeding the maximum size of the concatenated frame) from a certain management table T1 based on an instruction from the generation unit 37. Subsequently, the deletion unit 36 leaves the DA of the MAC frame scheduled to be located at the head of the concatenated frame among the plurality of read MAC frames, and sets the DA field included in each of the MAC frames scheduled to be located after the second. delete.
- the generation unit 37 generates a concatenated frame according to the “DA concatenation” method.
- FIG. 12 shows a format of a concatenated frame generated by the “DA concatenation” method.
- the difference from the “different DA concatenation” format is that the DA field of the MAC frame portion located after the second is deleted. Therefore, in the “DA concatenation” method, it is possible to improve the transmission efficiency in accordance with the size of the DA field to be deleted, compared to the “different DA concatenation” method.
- step S5 in FIG. 9 If it is determined in step S5 in FIG. 9 that the concatenation type is “DA concatenation”, the concatenation code “01” is inserted into the MAC frame by the DA concatenation type insertion unit 35A in step S10.
- the MAC frame is sorted for each DA (step S11) and temporarily stored in the management table (buffer) T1 corresponding to the DA.
- the deletion unit 36 that has received a read instruction from the generation unit 37 reads the number of MAC frames (having the same DA) according to the concatenated frame size from the management table T1.
- the deletion unit 36 deletes the DA field from each MAC frame scheduled to be placed second and later in the concatenated frame (step S12).
- the DA field of the MAC frame scheduled to be arranged at the head of the concatenated frame is not deleted.
- the deletion unit 36 gives a plurality of MAC frames to the generation unit 37.
- the generating unit 37 generates a concatenated frame having the format shown in FIG. 12 according to the “DA concatenation” method (step S8).
- the subsequent processing after step S9 (FCS insertion) is the same as in the “different DA concatenation” method.
- connection type determination unit 42 determines that the connection type is “DA connection” in the MAC processing unit 40 on the entrance side illustrated in FIG. 10
- the connection type determination unit 42 inputs the connection frame to the DA division unit 43A.
- the dividing process by the DA dividing unit 43A is different from the process of the different DA dividing unit 43D in the “different DA concatenation” method in the following points.
- the second and subsequent MAC frame portions in the concatenated frame do not include a DA field.
- the frame length stored in the “Length” field of the second and subsequent MAC frames is a value that is larger by the deleted DA field (6 bytes).
- the DA dividing unit 43A obtains the start position of the third and subsequent MAC frame parts, that is, the start position of the SA field, a value obtained by subtracting 6 bytes from the field length of the “Length” field of the immediately preceding MAC frame part. Is the frame length of the immediately preceding MAC frame part, and the start position of the SA field of the next MAC frame part is obtained.
- step S004 a connection frame is sent from the connection type determination unit 42 to the DA division unit 43A (step S013).
- the DA dividing unit 43A detects the frame length in the “Length” field of the head MAC frame portion in the concatenated frame (S014). Subsequently, the DA dividing unit 43A specifies the end position of the FCS field in the first MAC frame based on the frame length, and separates the MAC frame portion (unit block) there (step S015).
- the DA dividing unit 43A passes the unit block to the restoration unit 44.
- the restoration unit 44 determines whether or not to restore the DA based on whether or not the delivered unit block is the head MAC frame part (step S016). At this time, if the unit block is the head MAC frame portion, the process proceeds to step S018, and it is determined whether or not there is a next MAC frame portion (remaining concatenated frame).
- step S014 If there is a next MAC frame portion, the process returns to step S014, and the frame length of the “Length” field located at the headmost side is detected for the remaining portion of the concatenated frame, and separation is performed based on this frame length. (Separation position) is determined, and unit block separation processing is performed (step S015).
- the value obtained by subtracting the DA field length from the frame length is the start position of the SA field in the next MAC frame part (next Used to determine the unit block separation position.
- the restoration unit 44 copies the value of the DA field of the first (first) MAC frame part (unit block). It is given to the second and subsequent unit blocks (step S017). In this way, the DA field is restored in the unit block.
- a plurality of unit blocks in which the DA field is restored as necessary are generated by the loop processing of steps S014 to S018.
- the frame length deleting unit 45 deletes the “Length” field for each of the plurality of unit blocks (step S010). Thereafter, processing similar to the “different DA connection” method is performed (steps S011, S012).
- a concatenated frame that can be handled as a standard MAC frame can be generated.
- the number of MAC frames that can be included in the concatenated frame according to the size of the DA field to be deleted can be increased compared to the “different DA concatenation” method. It is possible to increase transmission efficiency.
- MAC frame concatenation based on VID concatenation method Next, MAC frame concatenation based on the “VID concatenation” method will be described.
- MAC frame concatenation is performed by collecting MAC frames having the same VID (VLAN ID).
- connection type selection unit 34 sends the MAC frame received from the frame length insertion unit 33 to the VID connection type insertion unit 35B in accordance with the execution instruction of the “VID connection” method from the CPU 21.
- the VID concatenation type insertion unit 35B inserts a concatenation code “10” indicating “VID concatenation” at a predetermined position of the MAC frame and stores it in the management table T2. At this time, as the management table (buffer) T2, one or more management tables (buffers) T2 for each VID, or different temporary storage areas for each VID are prepared.
- the deletion unit 36 based on an instruction from the generation unit 37, a plurality of MAC frames (having the same VID) corresponding to the concatenated frame size (not exceeding the maximum size of the concatenated frame) from a certain management table (buffer) T2. Is read. Subsequently, the deletion unit 36 leaves the VLAN tag field of the MAC frame scheduled to be located at the head (first) of the concatenated frame among the plurality of read MAC frames, and each MAC frame scheduled to be located after the second The VLAN tag field included in is deleted.
- the generation unit 37 generates a concatenated frame according to the “VID concatenation” method.
- FIG. 13 shows a format of a concatenated frame generated by the “VID concatenation” method.
- VLAN tag field (4 bytes) is provided as a VID storage field between the SA field and the Type / Length field.
- the difference between the concatenated frame format shown in FIG. 13 and the “different DA concatenation” format (FIG. 6) is that the VLAN tag field of each MAC frame portion located after the second is deleted. Therefore, the “VID concatenation” method can improve the transmission efficiency according to the size of the VLAN tag field to be deleted, compared with the “different DA concatenation” method.
- step S6 of FIG. 9 If it is determined in step S6 of FIG. 9 that the concatenation type is “VID concatenation”, the concatenation code “10” is inserted into the MAC frame by the DA concatenation type insertion unit 35B in step S13.
- the MAC frame is sorted for each VID (step S14) and temporarily stored in the management table (buffer) T2 corresponding to the VID.
- the deletion unit 36 that has received a read instruction from the generation unit 37 reads the number of MAC frames (having the same VID) according to the concatenated frame size from the management table (buffer) T2, and the second and subsequent in the concatenated frame.
- the VLAN tag field is deleted from each MAC frame scheduled to be placed in (Step S15). However, the VLAN tag field of the MAC frame scheduled to be arranged at the head (first) in the concatenated frame is not deleted.
- the plurality of MAC frames are given to the generation unit 37.
- the generating unit 37 generates a concatenated frame having the format shown in FIG. 13 according to the “VID concatenation” method (step S8).
- the subsequent processing after step S9 (FCS insertion) is the same as in the “different DA concatenation” method.
- connection type determination unit 42 determines that the connection type is “VID connection” in the MAC processing unit 40 on the ingress side illustrated in FIG. 10, the connection type is input to the VID division unit 43B.
- the processing by the VID dividing unit 43B is different from the processing of the different DA dividing unit 43D in the “different DA concatenation” method in the following points.
- the second and subsequent MAC frame portions in the concatenated frame do not include a VLAN tag field.
- the frame length stored in the “Length” field of the second and subsequent MAC frames has a value that is larger by the deleted VLAN tag field (4 bytes).
- the VID dividing unit 43B obtains the value obtained by subtracting 4 bytes from the field length of the “Length” field of the immediately preceding MAC frame part when obtaining the start position of the third and subsequent MAC frame parts, that is, the start position of the DA field. Is the frame length of the previous MAC frame part, and the start position of the DA field of the next MAC frame part is obtained.
- step S005 a connection frame is sent from the connection type determination unit 42 to the VID division unit 43B (step S019).
- the VID dividing unit 43B detects the frame length in the “Length” field of the first (first) MAC frame portion in the concatenated frame (step S020). Subsequently, the VID dividing unit 43B specifies the end position of the FCS field in the first MAC frame in the concatenated frame, and separates the MAC frame portion (unit block) there (step S021).
- the VID dividing unit 43B passes the unit block to the restoration unit 44.
- the restoration unit 44 determines whether or not to restore the VID based on whether or not the delivered unit block is the head MAC frame part (step S022). At this time, if the unit block is the head MAC frame portion, the process proceeds to step S024, and it is determined whether or not there is a next MAC frame portion (remaining concatenated frame).
- step S024 is performed by the VID dividing unit 43B. Therefore, the processes of steps S020, S021, and S024 are performed by the VID dividing unit 43B, and the processes of steps S022 and S023 are performed by the restoring unit 44. For this reason, the order of both processes does not have to follow the flow. For example, the process of step S022 and the process of step S024 may be executed in parallel.
- step S020 If there is a next MAC frame portion (remaining portion of the concatenated frame), the process returns to step S020, and the frame length of the “Length” field located at the headmost side in the remaining portion of the concatenated frame is detected. Based on the length, separation (separation position) is determined, and unit block separation processing is performed (step S021).
- the value obtained by subtracting the VLAN tag field length (4 bytes) from the frame length is the DA field in the next MAC frame part. This is used to determine the start position of (the next unit block separation position).
- the restoration unit 44 determines the value of the VLAN tag field of the first (first) MAC frame part (unit block), That is, the VID is duplicated and given to the second and subsequent unit blocks (step S023). In this way, the VLAN tag field is restored in the unit block.
- a plurality of unit blocks in which the VLAN tag field is restored as necessary are generated by the loop processing of steps S020 to S024.
- the frame length deleting unit 45 deletes the “Length” field for each of the plurality of unit blocks (step S010). Thereafter, processing similar to the “different DA connection” method is performed (steps S011, S012).
- a concatenated frame that can be handled as a standard MAC frame can be generated.
- the number of MAC frames that can be included in the concatenated frame according to the size of the VLAN tag field to be deleted can be increased compared to the “different DA concatenation” method. It is possible to increase transmission efficiency.
- connection type selection unit 34 sends the MAC frame received from the frame length insertion unit 33 to the VID and DA connection type insertion unit 35C in accordance with the execution instruction of the “VID and DA connection” method from the CPU 21.
- the VID and DA concatenation type insertion unit 35C inserts a concatenation code “11” indicating “VID and DA concatenation” at a predetermined position of the MAC frame and stores it in the management table T3.
- a concatenation code “11” indicating “VID and DA concatenation” at a predetermined position of the MAC frame and stores it in the management table T3.
- the management table (buffer) T3 one or more management tables (buffers) T3 for each VID and DA, or different temporary storage areas for each VID and DA are prepared.
- the deletion unit 36 Based on an instruction from the generation unit 37, the deletion unit 36 has a plurality of MAC frames (the same VID and the same) that correspond to the concatenated frame size (not exceeding the maximum size of the concatenated frame) from a certain management table (buffer) T3. Read DA). Subsequently, the deletion unit 36 leaves the DA field and the VLAN tag field of the MAC frame scheduled to be located at the head (first) of the concatenated frame among the plurality of read MAC frames, and is scheduled to be located after the second. The DA field and VLAN tag field included in each MAC frame are deleted.
- the generation unit 37 generates a concatenated frame according to the “VID and DA concatenation” method.
- FIG. 14 shows a format of a concatenated frame generated by the “VID and DA concatenation” method.
- the difference between the concatenated frame format shown in FIG. 14 and the format of the “VID concatenation” method (FIG. 13) is that not only the VLAN tag field but also the DA field is deleted for each MAC frame portion located after the second. Is a point. Therefore, in the “VID and DA concatenation” method, it is possible to improve the transmission efficiency according to the size of the DA field and VLAN tag field to be deleted, compared to the “different DA concatenation” method and the “VID concatenation” method. Become.
- step S7 of FIG. 9 If it is determined in step S7 of FIG. 9 that the concatenation type is “VID and DA concatenation”, the concatenation code “11” is inserted into the MAC frame by the VID and DA concatenation type insertion unit 35C in step S16.
- the MAC frame is sorted for each VID and DA (step S17) and temporarily stored in the management table (buffer) T3 corresponding to the VID and DA.
- the deletion unit 36 that has received a read instruction from the generation unit 37 reads the number of MAC frames (having the same VID and the same DA) according to the concatenated frame size from the management table (buffer) T3, and in the concatenated frame
- the DA field and the VLAN tag field are deleted from each MAC frame scheduled to be placed after the second (step S18).
- the DA field and VLAN tag field of the MAC frame scheduled to be arranged at the head (first) in the concatenated frame are not deleted.
- the plurality of MAC frames are given to the generation unit 37.
- the generating unit 37 generates a concatenated frame having the format shown in FIG. 14 according to the “VID and DA concatenation” method (step S8).
- the subsequent processing after step S9 (FCS insertion) is the same as in the “different DA concatenation” method.
- connection type determination unit 42 determines that the connection type is “VID and DA connection” in the MAC processing unit 40 on the ingress side illustrated in FIG. 10, the connection type is input to the VID and DA division unit 43C.
- the processing by the VID and DA dividing unit 43C is different from the processing of the different DA dividing unit 43D in the “different DA concatenation” method in the following points.
- the second and subsequent MAC frame portions in the concatenated frame do not include a DA field and a VLAN tag field.
- the frame length stored in the “Length” field of the second and subsequent MAC frame portions is larger by the deleted DA field and VLAN tag field (10 bytes).
- the VID and DA dividing unit 43C subtracts 10 bytes from the field length of the “Length” field of the immediately preceding MAC frame part when determining the start position of the third and subsequent MAC frame parts, that is, the start position of the SA field.
- the start position of the SA field of the next MAC frame portion is obtained using the obtained value as the frame length of the immediately preceding MAC frame portion.
- connection frame is sent from the connection type determination unit 42 to the division unit 43C (step S025).
- the dividing unit 43C detects the frame length in the “Length” field of the first (first) MAC frame portion in the concatenated frame (step S026). Subsequently, the dividing unit 43C specifies the end position of the FCS field in the first MAC frame in the concatenated frame, and separates the MAC frame portion (unit block) there (step S027).
- the division unit 43C passes the unit block to the restoration unit 44.
- the restoration unit 44 determines whether or not to restore the VID and DA based on whether or not the delivered unit block is the head MAC frame part (step S028). At this time, if the unit block is the head MAC frame portion, the process proceeds to step S030, and it is determined whether or not there is a next MAC frame portion (remaining concatenated frame).
- step S030 The determination in step S030 is performed by the dividing unit 43C. Therefore, the processes of steps S026, S027, and S030 are performed by the dividing unit 43C, and the processes of steps S028 and S029 are performed by the restoring unit 44. For this reason, the order of both processes does not have to follow the flow. For example, the process of step S028 and the process of step S030 may be executed in parallel.
- step S026 If there is a next MAC frame portion (remaining portion of the concatenated frame), the process returns to step S026 to detect the frame length of the “Length” field located at the most leading side in the remaining portion of the concatenated frame. Based on the length, separation (separation position) is determined, and unit block separation processing is performed (step S027).
- the value obtained by subtracting the DA field length and the VLAN tag field length (10 bytes) from the frame length is as follows. Is used to determine the start position of the SA field in the MAC frame portion (separation position of the next unit block).
- the restoration unit 44 performs the DA field of the first (first) MAC frame part (unit block).
- the value (destination MAC address) and the value (VID) of the VLAN tag field are duplicated and given to the second and subsequent unit blocks (step S023). In this way, the DA field and VLAN tag field are restored to the unit block.
- a plurality of unit blocks in which the DA field and the VLAN tag field are restored as necessary are generated by the loop processing of steps S026 to S030.
- the frame length deleting unit 45 deletes the “Length” field for each of the plurality of unit blocks (step S010). Thereafter, processing similar to the “different DA connection” method is performed (steps S011, S012).
- a concatenated frame that can be handled as a standard MAC frame can be generated. Also, in the “VID and DA concatenation” method, it is possible to increase the number of MAC frames that can be included in the concatenated frame according to the size of the DA and VLAN tag fields to be deleted compared to the “different DA concatenation” method. It is possible to increase the transmission efficiency.
- the switch 24 can be realized using, for example, a semiconductor memory.
- the PHY chip 25 and the MAC chip 26 are realized by using a dedicated or general-purpose integrated circuit (hardware) such as Application Specific Integrated Circuit (ASIC) or Field Programmable Gate Array (FPGA), or a combination thereof. Can do.
- ASIC Application Specific Integrated Circuit
- FPGA Field Programmable Gate Array
- Each component of the exit side MAC processing unit 30 described with reference to FIG. 5 and the entrance side MAC processing unit 40 described with reference to FIG. 10 can be realized by ASIC, FPGA, or a combination thereof.
- the management table (buffer) is created on a storage area of a storage (for example, a semiconductor memory or a flash memory) included in the MAC chip 26.
- exit side MAC processing unit 30 and the entrance side MAC processing unit 40 are mounted on one MAC chip 26 .
- the exit side MAC processing unit 30 and the entrance side MAC processing unit 40 may be mounted on different chips.
- the configurations of the egress-side MAC processing unit 30 and the ingress-side MAC processing unit 40 are exemplary, and the egress-side MAC processing unit 30 and the ingress-side MAC processing unit 40 can be realized by one or more integrated circuits, respectively.
- the transmission efficiency for the physical bandwidth of the MAC frame is as follows. 64 / (64 + 12 + 8) ⁇ 0.76
- the transmission efficiency is as follows. [(64 + 2) ⁇ 100] / [(64 + 2) ⁇ 100 + 12 + 8] ⁇ 0.999697 Therefore, it is possible to increase the transmission efficiency to 99%.
- the L2 switch 13 (14) has 48 1 Gbps downlink lines and 4 10 Gbps uplink lines.
- the maximum use bandwidth on the uplink side is as follows.
- DA concatenation a plurality of MAC frames having the same DA are bundled (concatenated), so that only the first (first) MAC frame portion in the concatenated frame is included.
- the DA field is left.
- the transmission efficiency at this time is about 77% as follows. 68 / (68 + 12 + 8) ⁇ 0.773
- the transmission efficiency is as follows. (68 + 64 ⁇ 99) / (68 + 64 ⁇ 99 + 12 + 8) ⁇ 0.997
- the transmission efficiency can be increased to 99%.
- FIG. 15 shows the bandwidth usage rate when 10 MAC frames are concatenated. Although the usage rate changes according to the frame size, it can be seen that it is improved by up to about 20% compared to the conventional non-connected method.
- FIG. 16 shows a change in bandwidth usage rate depending on the number of MAC frames connected.
- the frame size of the MAC frame is fixed at 64 bytes. It can be seen that the bandwidth utilization rate is improved to nearly 100% when the connection is performed as compared with the conventional method without connection (the number of connections is one).
- the concatenated frame uses the DA, SA, (VLAN tag), and Type / Length field portions that form the first MAC frame portion of the concatenated frame as the MAC frame header, and the last FCS immediately after the Type / Length field.
- the size of the data portion of the concatenated frame is set as the frame length in the Type / Length field of the head MAC frame portion.
- the standard MAC frame format since the upper limit of the size of the data portion is 1500 bytes, the same handling as the standard MAC frame is possible when the data portion size of the concatenated frame is 1500 bytes or less.
- the upper limit of the data portion size of the concatenated frame is 1500 bytes
- 84 bytes 64 + 12 + 8) are required for each MAC frame, so the number of MAC frames that can be transmitted in the same band is only 17. Accordingly, it is possible to improve transmission efficiency (transmission efficiency) and bandwidth utilization even under the operation where the upper limit of the data portion size of the concatenated frame is 1500 bytes.
- the CPU 21 and the MAC device 26 perform the connection process and the restoration process.
- the MAC device 26 includes the function of the CPU 21, and the connection process as described above using only the MAC device 26.
- the restoration process may be executed.
- the L2 switch in which the egress MAC processing unit 30 functions as a frame connection device and the ingress MAC processing unit 40 functions as a frame restoration device has been described.
- the frame concatenation apparatus and the frame restoration apparatus can also be applied to a communication apparatus that performs MAC frame transfer control, such as a switching HUB.
- the frame connection device and the frame restoration device can be applied (mounted) not only to a relay device such as an L2 switch or a switching HUB, but also to a terminal device that transmits and receives MAC frames.
- the concatenated code may be inserted only in the Length field of an arbitrary MAC frame part (for example, the first MAC frame part) of the concatenated frame.
- the first Length field may be 2 bytes, and the size of the Length field in other MAC frames may be reduced.
- the second embodiment includes a configuration common to the first embodiment. Therefore, in the second embodiment, differences from the first embodiment will be mainly described, and the same reference numerals are given to configurations common to the first embodiment, and description thereof will be omitted.
- the upper limit size of the data part is 1500 bytes
- the upper limit size of the frame length that can be expressed by the Type / Length field is 1500 [bytes]. Therefore, when a concatenated frame size (frame length) exceeding 1500 bytes is written in the Type / Length field, the general-purpose L2 switch according to the standard MAC transfer protocol recognizes the value of this Type / Length field as the frame length. I can't.
- Each of the frame concatenation device and the frame restoration device in the second embodiment can be mounted as an egress MAC processing unit and an ingress MAC processing unit in the MAC chip (MAC device) 26 as in the first embodiment.
- FIG. 17 shows a configuration example of the egress MAC processing unit in the second embodiment.
- the exit-side MAC processing unit 30A in FIG. 17 is different from the exit-side MAC processing unit 30 (FIG. 5) in the first embodiment in the following points.
- the rewrite unit 39 is inserted between the generation unit 37 and the FCS insertion unit 38.
- the rewriting unit 39 checks the value of the Type / Length field included in the head MAC frame portion (hereinafter referred to as the head frame (head MAC frame)) in the concatenated frame generated by the generation unit 37.
- the rewriting unit 39 does not particularly perform processing when the value of the Type / Length field is a value indicating the type of the upper protocol.
- the rewriting unit 39 analyzes the upper protocol in the head frame. If the type (type) of the upper protocol is found as a result of the analysis, the rewriting unit 39 rewrites the value of the Type / Length field to a value (Type value) indicating the type of the upper protocol.
- the analysis of the upper protocol can be realized, for example, by analyzing the upper protocol header stored in the data part of the first frame in the concatenated frame.
- the type / length field of the concatenated frame The value is interpreted as a type value of the upper protocol by the receiving device of the concatenated frame.
- the concatenated frame is handled as a normal MAC frame by the communication device on the receiving side (for example, a general-purpose L2 switch or switching HUB). Is possible. In other words, the concatenated frame can be transmitted to the destination via a general-purpose L2 switch or switching HUB.
- FIG. 18 shows an example of the data structure of the “Length” field in the second embodiment. Unlike the “Length” field (FIG. 7) in the first embodiment, whether or not the third bit (bit b2) has rewritten the value of the Type / Length field of the first frame (first MAC frame) in the concatenated frame into a Type value. It is treated as rewrite information indicating that.
- bit b2 is “0”, it indicates that rewriting to the Type value has not been performed, and if bit b2 is “1”, it indicates that rewriting to the Type value has been performed. Show.
- bits b3 to b15 indicate the frame length (frame size) of the MAC frame. Since bits b3 to b15 are 13 bits, the frame size can be expressed up to 8192 bytes in the second embodiment.
- bit b2 of the Length field in the head frame is referred to in the process of restoring a plurality of MAC frames from the concatenated frame on the concatenated frame receiving side. For this reason, the bit b2 of the Length field inserted in the second and subsequent MAC frame portions is not substantially used. Therefore, the data structure (FIG. 7) described in the first embodiment can be applied to the Length field in each of the second and subsequent MAC frames excluding the top frame.
- FIG. 19 is a flowchart showing an operation example of the egress MAC processing unit in the second embodiment. The difference from the first embodiment is that a process by the rewrite unit 39 is inserted as step S8A between step S8 and step S9.
- step S8A the rewrite unit 39 receives the connection frame from the generation unit 37.
- the rewriting unit 39 refers to the value of the Type / Length field of the first frame in the concatenated frame, and determines whether the field value indicates a frame length (frame size) or a Type value.
- the rewrite unit 39 passes the concatenated frame to the FCS insertion unit 38 without performing any particular processing.
- the rewrite unit 39 analyzes the upper protocol type and displays the Type / Length field with the Type value indicating the upper protocol type obtained as the analysis result. Rewrite the value. Further, the rewriting unit 39 sets the value of the bit b2 in the Length field in the first frame to “1”. Thereafter, the rewriting unit 39 passes the concatenated frame to the FCS insertion unit 38.
- the configuration and operation of the egress MAC processing unit 30A are the same as the configuration and operation of the egress side MAC processing unit 30 in the first embodiment.
- FIG. 20 shows a configuration example of the ingress MAC processing unit in the second embodiment.
- the ingress MAC processing unit 40A in FIG. 20 differs from the ingress MAC processing unit 40 (FIG. 10) in the first embodiment in the following points.
- the playback unit 48 is inserted between the FCS check unit 41 and the connection type determination unit 42.
- the reproducing unit 48 checks the value of the bit b2 included in the Length field in the concatenated frame received from the FCS check unit 41.
- the playback unit 48 When the bit b2 is “0”, the playback unit 48 performs playback processing on the assumption that the Type / Length field included in the first frame has not been rewritten (the original value is the Type value).
- the connection frame is passed to the connection type determination unit 42 without any change. That is, the playback unit 48 passes through the connection frame.
- the reproducing unit 48 Performs playback processing.
- the playback unit 48 acquires the value of the Length value (frame size) stored in the Length field of the first frame, and rewrites the value of the Type / Length field with the acquired value. As a result, the value of the Type / Length field returns to the original Length value. Thereafter, the playback unit 48 passes the connection frame to the connection type determination unit 42.
- FIG. 21 is a flowchart showing an operation example of the ingress side MAC processing unit in the second embodiment. The difference from the first embodiment is that a b2 check and reproduction process by the reproduction unit 48 is inserted between step S001 and step S003 as step S002A.
- step S002A the reproduction unit 48 checks the value of the bit b2 of the Length field of the first frame in the concatenated frame received from the FCS check unit 41 (type / length field rewriting presence / absence determination information). It is determined whether it is “0” or “1”.
- the reproducing unit 48 passes the concatenated frame to the concatenated type determining unit 42. On the other hand, if the bit b2 is “1”, the reproducing unit 48 rewrites the value of the Type / Length field of the head frame to the value of the frame length (frame size) of the Length field of the head frame. Thereafter, the playback unit 48 passes the connection frame to the connection type determination unit 42.
- the configuration and operation of the ingress MAC processing unit in the second embodiment are the same as those in the first embodiment.
- the rewriting unit 39 indicates whether the value of the Type / Length field of the first frame in the concatenated frame indicates the Type value or the Length value. Determine whether.
- the rewriting unit 39 rewrites the Length value to the Type value.
- the concatenated frame since the format of the concatenated frame conforms to the standard MAC format, the concatenated frame is handled as a normal MAC frame by a general-purpose L2 switch or switching HUB that performs standard MAC frame transfer. That is, the concatenated frame can be sent to the destination via a general-purpose L2 switch or switching HUB.
- the reproducing unit 48 refers to the bit b2 included in the first frame of the concatenated frame. If the value of the Type / Length field has been rewritten, the playback unit 48 rewrites the value of the Type / Length field with the value of the frame length of the Length field of the first frame, thereby rewriting the value of the Type / Length field. Return to the value of. As a result, the value of the Type / Length field of the first frame (unit block) divided by the dividing unit 43 (any of the dividing units 43A to 43D) can be returned to the original value.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Security & Cryptography (AREA)
- Software Systems (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Small-Scale Networks (AREA)
- Communication Control (AREA)
Abstract
Description
前記記憶部から読み出される前記複数のフレームが直列に連結され、且つ各フレームのフレーム長情報が付与された連結フレームを生成する生成部と、
前記連結フレームに、当該連結フレームの受信側との同期確立用情報を付与する付与部とを含む
フレーム連結装置である。
<ネットワークシステムの構成例>
図1は、実施形態における、データセンタを含むネットワークシステムの構成例を示す。図1において、データセンタ10は、複数のサーバやストレージを収容するラックと呼ばれる領域を有している。図1に示す例では、複数のサーバ(サーバマシン)1が収容されたサーバラック11と、複数のストレージ2として機能するコンピュータ(サーバマシン)が収容されたストレージラック12とが例示されている。サーバ1やストレージ2として機能するサーバマシン、すなわち物理サーバ内には、複数の仮想マシン(VM)を配備することができる。つまり、1つの仮想マシンは、サーバ(サーバマシン)1またはストレージ2として機能するコンピュータ(サーバマシン)として動作することもできる。
図4は、L2スイッチの構成例を示す。図4において、L2スイッチ20は、図1に示したL2スイッチ13,14,15として適用可能なものである。L2スイッチ20は、Central Processing Unit(CPU)21と、SDRAM22と、フラッシュメモリ(Flash MEM)23とを備えている。
次に、図4に示したMACチップ26で実行されるMACコンカチネーション処理について説明する。MACコンカチネーション処理にあたり、CPU21は、MACチップ26に対して、使用する連結方式や連結するフレームサイズなどの指示を出すことができる。
図5は、出口側のMAC処理部の構成例を示す。出口側のMAC処理部30は複数のMACフレームを連結し、MACフレームコンカチネーションを行う。
(1)異なる宛先アドレス(DA)を有するMACフレームを連結する“異DA連結”方式。
(2)同一のDAを有するMACフレームを連結する“DA連結”方式。
(3)同一のVIDを有するMACフレームを連結する“VID連結”方式。
(4)同一のVID及びDAを有するMACフレームを連結する“VID及びDA連結”方式。
図9は、出口側のMAC処理部30の動作例を示すフローチャートである。図9に示す処理は、MAC処理部30にMACフレームが入力されることによって開始される。
次に、MACチップ(MACデバイス)26における入口側のMAC処理部40について説明する。図10は、入口側のMAC処理部40の構成例を示す。MAC処理部40は、MACフレームコンカチネーションを解除し、連結前の各MACフレームを復元する。
図11は、入口側のMAC処理部40の動作例を示すフローチャートである。図11に示す処理は、MAC処理部40に連結フレームが入力されることによって開始される。
次に、“DA連結”方式に基づくMACフレームコンカチネーションについて説明する。“DA連結”方式では、同一のDAを有するMACフレームを集めてMACフレームコンカチネーションが実行される。
次に、“VID連結”方式に基づくMACフレームコンカチネーションについて説明する。“VID連結”方式では、同一のVID(VLAN ID)を有するMACフレームを集めてMACフレームコンカチネーションが実行される。
次に、“VID及びDA連結”方式に基づくMACフレームコンカチネーションについて説明する。“VID及びDA連結”方式では、同一のVID(VLAN ID)及び同一のDAを有するMACフレームを集めてMACフレームコンカチネーションが実行される。
上述した実施形態によれば、以下の効果を得ることが可能である。例えば、64バイトのMACフレームを送信する場合、MACフレームの物理帯域に対する送信効率は、以下となる。
64/(64+12+8)≒0.76
これに対し、上述した“異DA連結”方式に基づき64バイトのMACフレームを100個連結した場合には、その送信効率は、以下となる。
[(64+2)×100]/[(64+2)×100+12+8]≒0.99697
従って、送信効率を99%まで高めることが可能である。
48×0.76=36.48[Gbps]
一方、標準MACフレームのフレームサイズが64バイトであるときにおけるアップリンク側の最大使用帯域は、以下となる。
40×0.76=30.4[Gbps]
これに対し、上述した“異DA連結”方式を適用すれば、アップリンク側の最大使用帯域は、以下のようになる。
40×0.99697=39.8788
従って、MACフレームコンカチネーションを実施しない場合には、アップリンク回線で1Gbps×48本分の帯域を収容できなかったが、“異DA連結”方式による連結フレームを生成することで、ダウンリンク回線の帯域を収容することができる。
(6600-6)/(64+2-6)=109.9
すなわち、MACフレームを109.9個連結したことになり、“異DA連結”方式と比較して9.9%の効率向上となる。
68/(68+12+8)≒0.773
これに対し、“VID連結”方式を用いて同一VIDを有するMACフレームを100個束ねる(連結する)と、送信効率は次のようになる。
(68+64×99)/(68+64×99+12+8)≒0.997
このように、送信効率を99%まで高めることができる。
[68+(64-6)×99]/[68+(64-6)×99+12+8]≒0.997
図15は、10個のMACフレームを連結した場合における帯域使用率を示す。フレームサイズに応じて使用率は変化するが、従来の連結しない方式に比べて最大約20%向上していることがわかる。
上述したように、連結フレームは、連結フレームにおける先頭のMACフレーム部分をなすDA,SA,(VLANタグ),Type/Lengthのフィールド部分をMACフレームヘッダとし、Type/Lengthフィールド直後から最後尾のFCSフィールドの直前までの領域をデータ部と考えることで、標準MACフレームとして扱うことができる。このとき、先頭のMACフレーム部分のType/Lengthフィールドには、連結フレームのデータ部のサイズがフレーム長として設定される。
(64+2)×22+12+8=1472byte
これに対し、MACフレームを連結しない場合には、MACフレーム1つ当たり84バイト(64+12+8)が必要になるので、同一の帯域で送信可能なMACフレーム数は17に止まる。従って、連結フレームのデータ部サイズ上限を1500バイトとする運用下でも、送信効率(伝送効率)及び帯域使用率の向上を図ることができる。
以下、本発明の第2実施形態について説明する。第2実施形態は、第1実施形態と共通の構成を含む。このため、第2実施形態は、主として第1実施形態との相違点について説明し、第1実施形態と共通の構成に関しては、同一の符号を付して説明を省略する。
これによって、分割部43(分割部43A~43Dのいずれか)で分割される先頭フレーム(単位ブロック)のType/Lengthフィールドの値をオリジナルの値に戻すことができる。
21・・・CPU
22・・・SDRAM
23・・・フラッシュメモリ
24・・・MACスイッチ
25・・・PHYチップ(PHYデバイス)
26・・・MACチップ(MACデバイス)
31、41・・・FCSチェック部
32・・・フレーム長チェック部
33・・・フレーム長挿入部
34・・・連結タイプ選択部
35・・・連結タイプ挿入部
36・・・重複バイト削除部
37・・・連結フレーム生成部
38,47・・・FCS挿入部
42・・・連結タイプ判定部
43・・・分割部
44・・重複バイト復元部
45・・・フレーム長削除部
46・・・SFD挿入部
Claims (19)
- 送信対象の複数のフレームを格納する記憶部と、
前記記憶部から読み出される前記複数のフレームが直列に連結され、且つ各フレームのフレーム長情報が付与された連結フレームを生成する生成部と、
前記連結フレームに、当該連結フレームの受信側との同期確立用情報を付与する付与部と
を含むフレーム連結装置。 - 前記複数のフレームの夫々は、同一の宛先情報を含んでおり、
前記生成部は、前記複数のフレームの何れか1つが前記宛先情報を含み、且つ残りの1以上のフレームから前記宛先情報が夫々省略された前記連結フレームを生成する
請求項1に記載のフレーム連結装置。 - 前記複数のフレームの夫々は、同一の仮想網識別情報を含んでおり、
前記生成部は、前記複数のフレームの何れか1つが前記仮想網識別情報を含み、且つ残りの1以上のフレームから前記仮想網識別情報が夫々省略された前記連結フレームを生成する
請求項1に記載のフレーム連結装置。 - 前記複数のフレームの夫々は、同一の宛先情報及び同一の仮想網識別情報を含んでおり、
前記生成部は、前記複数のフレームの何れか1つが前記宛先情報及び前記仮想網識別情報を含み、且つ残りの1以上のフレームから前記宛先情報及び前記仮想網識別情報が夫々省略された前記連結フレームを生成する
請求項1に記載のフレーム連結装置。 - 前記生成部は、
(1)同一の宛先情報を有する複数のフレームの何れか1つが前記宛先情報を含み、且つ残りの1以上のフレームから前記宛先情報が夫々省略された連結フレームを生成する第1の処理と、
(2)同一の仮想網識別情報を有する複数のフレームの何れか1つが前記仮想網識別情報を含み、且つ残りの1以上のフレームから前記仮想網識別情報が夫々省略された連結フレームを生成する第2の処理と、
(3)同一の宛先情報及び同一の仮想網識別情報を有する複数のフレームの何れか1つが前記宛先情報及び前記仮想網識別情報を含み、且つ残りの1以上のフレームから前記宛先情報及び前記仮想網識別情報が夫々省略された連結フレームを生成する第3の処理と、
(4)異なる宛先情報を有する複数のフレームが連結された連結フレームを生成する第4の処理と
の少なくとも2つから1つを選択して実行し、
前記連結フレーム又は前記複数のフレームの夫々に対し、前記生成部で選択される前記第1~第4の処理の何れかを特定する情報を挿入する挿入部をさらに含む
請求項1に記載のフレーム連結装置。 - 前記複数のフレームのフレーム長を夫々検出する検出部と、
前記検出されたフレーム長の夫々を、前記各フレームに挿入するフレーム長挿入部とをさらに含む
請求項1に記載のフレーム連結装置。 - 前記複数のフレームの何れか1つは、前記連結フレーム中の先頭に配置されるフレームである
請求項2から5の何れか1項に記載のフレーム連結装置。 - 前記フレームは、上位プロトコルのタイプを示すタイプ値とフレーム長を示すレングス値との一方が格納されるタイプ/レングスフィールドを含むMACフレームであり、
前記連結フレームにおいて先頭に位置するMACフレームのタイプ/レングスフィールドにレングス値が格納されている場合に、このレングス値をタイプ値に書き換えるとともに、タイプ/レングスフィールドの値が書き換えられたことを示す書換情報を連結フレームに付与する書換部
をさらに含む請求項1から7の何れか1項に記載のフレーム連結装置。 - 複数のフレームが直列に連結され、且つ各フレームのフレーム長情報が付与された連結フレームを受信する受信部と、
前記連結フレームに含まれる各フレームのフレーム長情報に基づいて、前記連結フレームを前記複数のフレームに分割する分割部と、
分割された前記複数のフレームの夫々から前記フレーム長情報を削除する削除部と
を含むフレーム復元装置。 - 前記連結フレームは、前記複数のフレームの何れか1つが宛先情報を含み、且つ残りの1以上のフレームから前記宛先情報が夫々省略された状態を有し、
前記分割部で分割された、前記宛先情報を含まない前記1以上のフレームの夫々に対して、前記複数のフレームの何れか1つに含まれる前記宛先情報を付与する復元部をさらに含む
請求項9に記載のフレーム復元装置。 - 前記連結フレームは、前記複数のフレームの何れか1つが仮想網識別情報を含み、且つ残りの1以上のフレームから前記仮想網識別情報が夫々省略された状態を有し、
前記分割部で分割された、前記仮想網識別情報を含まない前記1以上のフレームの夫々に対して、前記複数のフレームの何れか1つに含まれる前記仮想網識別情報を付与する復元部をさらに含む
請求項9に記載のフレーム復元装置。 - 前記連結フレームは、前記複数のフレームの何れか1つが宛先情報及び仮想網識別情報を含み、且つ残りの1以上のフレームから前記宛先情報及び前記仮想網識別情報が夫々省略された状態を有し、
前記分割部で分割された、前記宛先情報及び前記仮想網識別情報を含まない前記1以上のフレームの夫々に対して、前記複数のフレームの何れか1つに含まれる前記宛先情報及び前記仮想網識別情報を付与する復元部をさらに含む
請求項9に記載のフレーム復元装置。 - 前記分割部は、
(1)連結された複数のフレームの何れか1つが宛先情報を含み、且つ残りの1以上のフレームから前記宛先情報が夫々省略された状態の前記連結フレームを分割する第1の処理と、
(2)連結された複数のフレームの何れか1つが仮想網識別情報を含み、且つ残りの1以上のフレームから前記仮想網識別情報が夫々省略された状態の前記連結フレームを分割する第2の処理と、
(3)連結された複数のフレームの何れか1つが宛先情報及び仮想網識別情報を含み、且つ残りの1以上のフレームから前記宛先情報及び前記仮想網識別情報が夫々省略された状態の前記連結フレームを分割する第3の処理と、
(4)異なる宛先情報を有する複数のフレームが連結された前記連結フレームを分割する第4の処理と
の少なくとも2つから1つを選択して実行し、
前記分割部は、前記連結フレームに含まれる前記第1~第4の処理の何れかを特定する情報に従って、前記第1~第4の処理の何れかを選択する
請求項9に記載のフレーム復元装置。 - 前記複数のフレームの何れか1つは、前記連結フレーム中の先頭に配置されるフレームである
請求項9から12の何れか1項に記載のフレーム復元装置。 - 前記フレームは、上位プロトコルのタイプを示すタイプ値とフレーム長を示すレングス値との一方が格納されるタイプ/レングスフィールドを含むMACフレームであり、
前記連結フレームにおいて先頭に位置する先頭MACフレームのタイプ/レングスフィールドの値が書き換えられていることを示す書換情報が前記連結フレームに含まれている場合に、前記先頭MACフレームに対応するフレーム長情報に基づいて前記先頭MACフレームのタイプ/レングスフィールドの値を書き換える再生部
をさらに含む請求項9から14の何れか1項に記載のフレーム復元装置。 - 送信対象の複数のフレームを格納する記憶部と、
前記記憶部から読み出される前記複数のフレームが直列に連結され、且つ各フレームのフレーム長情報が付与された連結フレームを生成する生成部と、
前記連結フレームに、当該連結フレームの受信側との同期確立用情報を付与する付与部と、
前記同期確立用情報が付与された前記連結フレームを送出する送出部と
を含む通信装置。 - 前記記憶部に格納される前記複数のフレームはアップリンク及びダウリンクの一方から受信され、
前記送出部は、前記連結フレームを前記アップリンク及び前記ダウンリンクの他方へ向けて送出する
請求項16に記載の通信装置。 - 記憶部から読み出される送信対象の複数のフレームが連結され、且つ各フレームのフレーム長情報が付与された連結フレームを生成し、
前記連結フレームに、当該連結フレームの受信側との同期確立用情報を付与する
ことを含むフレーム連結方法。 - 複数のフレームが直列に連結され、且つ各フレームのフレーム長情報が付与された連結フレームを受信し、
前記連結フレームに含まれる各フレームのフレーム長情報に基づいて、前記連結フレームを前記複数のフレームに分割し、
分割された前記複数のフレームの夫々から前記フレーム長情報を削除する
ことを含むフレーム復元方法。
Priority Applications (5)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2010800689917A CN103081417A (zh) | 2010-09-07 | 2010-09-07 | 帧连接装置 |
JP2012532760A JP5573955B2 (ja) | 2010-09-07 | 2010-09-07 | フレーム連結装置 |
EP10856956.7A EP2615787A1 (en) | 2010-09-07 | 2010-09-07 | Frame concatenation device |
PCT/JP2010/065314 WO2012032606A1 (ja) | 2010-09-07 | 2010-09-07 | フレーム連結装置 |
US13/788,297 US9060030B2 (en) | 2010-09-07 | 2013-03-07 | Frame concatenation apparatus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2010/065314 WO2012032606A1 (ja) | 2010-09-07 | 2010-09-07 | フレーム連結装置 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/788,297 Continuation US9060030B2 (en) | 2010-09-07 | 2013-03-07 | Frame concatenation apparatus |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012032606A1 true WO2012032606A1 (ja) | 2012-03-15 |
Family
ID=45810231
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2010/065314 WO2012032606A1 (ja) | 2010-09-07 | 2010-09-07 | フレーム連結装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US9060030B2 (ja) |
EP (1) | EP2615787A1 (ja) |
JP (1) | JP5573955B2 (ja) |
CN (1) | CN103081417A (ja) |
WO (1) | WO2012032606A1 (ja) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2016105563A (ja) * | 2014-12-01 | 2016-06-09 | Necプラットフォームズ株式会社 | 通信装置、通信方法、及びプログラム |
JP2019008768A (ja) * | 2017-06-28 | 2019-01-17 | 株式会社ギウォンテクKiwontech | 詐欺メールの危険度を判断する方法およびそのための装置 |
US10200296B2 (en) | 2014-05-09 | 2019-02-05 | Nec Corporation | Packet transport apparatus, packet transport system and packet transport method |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9471772B2 (en) | 2011-06-01 | 2016-10-18 | Paypal, Inc. | Password check by decomposing password |
CN105701053B (zh) * | 2014-11-24 | 2020-01-17 | 中兴通讯股份有限公司 | 串行数据发送、接收方法及装置 |
JP7147403B2 (ja) * | 2018-09-12 | 2022-10-05 | 株式会社デンソー | 中継装置 |
US11936490B2 (en) * | 2019-08-06 | 2024-03-19 | Nippon Telegraph And Telephone Corporation | L2 switch, communication method, and communication program |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003324445A (ja) * | 2002-05-07 | 2003-11-14 | Alps Electric Co Ltd | 無線伝送方式 |
JP2004282740A (ja) * | 2003-03-13 | 2004-10-07 | Samsung Electronics Co Ltd | 通信システムのパケット伝送装置及び方法 |
JP2006527570A (ja) * | 2003-06-12 | 2006-11-30 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | データ送信のための連結されたフレーム構造 |
JP2007515113A (ja) * | 2003-11-24 | 2007-06-07 | インテロン・コーポレーション | 複数の局が共有媒体を介して通信するネットワークにおける動作方法 |
JP2009290744A (ja) * | 2008-05-30 | 2009-12-10 | Fujitsu Ltd | フレーム中継装置およびフレーム中継方法 |
JP2010119043A (ja) * | 2008-11-14 | 2010-05-27 | Mitsubishi Electric Corp | データ送受信装置 |
Family Cites Families (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6894999B1 (en) * | 2000-11-17 | 2005-05-17 | Advanced Micro Devices, Inc. | Combining VLAN tagging with other network protocols allows a user to transfer data on a network with enhanced security |
JP2002198994A (ja) | 2000-12-26 | 2002-07-12 | Nec Corp | Gfpフレーム転送装置およびgfpフレーム転送方法 |
US20020199021A1 (en) * | 2001-06-26 | 2002-12-26 | Niels Beier | Method and apparatus for using the type/length field in an ethernet mac header for carrying generic tags/labels |
US6801545B2 (en) | 2002-02-04 | 2004-10-05 | Fujitsu Network Communications, Inc. | Container transport for packets in connection oriented protocols |
US7408957B2 (en) * | 2002-06-13 | 2008-08-05 | International Business Machines Corporation | Selective header field dispatch in a network processing system |
KR101018204B1 (ko) * | 2003-05-16 | 2011-02-28 | 파나소닉 주식회사 | 통신 네트워크 시스템에서의 사용을 위한 통신 단말 장치, 방법, 프로그램, 기록 매체, 및 집적 회로 |
US7317687B2 (en) | 2003-06-12 | 2008-01-08 | Koninklijke Philips Electronics N.V. | Transmitting data frames with less interframe space (ifs) time |
CN1806418A (zh) * | 2003-06-12 | 2006-07-19 | 皇家飞利浦电子股份有限公司 | 用于数据传输的级联帧结构 |
US7684568B2 (en) | 2003-11-24 | 2010-03-23 | Intellon Corporation | Encrypting data in a communication network |
US7633970B2 (en) * | 2004-05-07 | 2009-12-15 | Agere Systems Inc. | MAC header compression for use with frame aggregation |
US20080080524A1 (en) | 2004-06-11 | 2008-04-03 | Tsunehito Tsushima | Router Device, Communication Device, Routing Method, Routing Program, and Computer-Readable Recording Medium Recorded with Routing Program |
EP1624638B1 (en) | 2004-08-05 | 2006-10-25 | Alcatel | Access control method and apparatus |
JP2008109471A (ja) * | 2006-10-26 | 2008-05-08 | Nec Corp | Lanシステム、送信装置、受信装置、lanシステム制御方法、プログラム、及び複数のフレーム |
-
2010
- 2010-09-07 WO PCT/JP2010/065314 patent/WO2012032606A1/ja active Application Filing
- 2010-09-07 CN CN2010800689917A patent/CN103081417A/zh active Pending
- 2010-09-07 EP EP10856956.7A patent/EP2615787A1/en not_active Withdrawn
- 2010-09-07 JP JP2012532760A patent/JP5573955B2/ja not_active Expired - Fee Related
-
2013
- 2013-03-07 US US13/788,297 patent/US9060030B2/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003324445A (ja) * | 2002-05-07 | 2003-11-14 | Alps Electric Co Ltd | 無線伝送方式 |
JP2004282740A (ja) * | 2003-03-13 | 2004-10-07 | Samsung Electronics Co Ltd | 通信システムのパケット伝送装置及び方法 |
JP2006527570A (ja) * | 2003-06-12 | 2006-11-30 | コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ | データ送信のための連結されたフレーム構造 |
JP2007515113A (ja) * | 2003-11-24 | 2007-06-07 | インテロン・コーポレーション | 複数の局が共有媒体を介して通信するネットワークにおける動作方法 |
JP2009290744A (ja) * | 2008-05-30 | 2009-12-10 | Fujitsu Ltd | フレーム中継装置およびフレーム中継方法 |
JP2010119043A (ja) * | 2008-11-14 | 2010-05-27 | Mitsubishi Electric Corp | データ送受信装置 |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10200296B2 (en) | 2014-05-09 | 2019-02-05 | Nec Corporation | Packet transport apparatus, packet transport system and packet transport method |
JP2016105563A (ja) * | 2014-12-01 | 2016-06-09 | Necプラットフォームズ株式会社 | 通信装置、通信方法、及びプログラム |
JP2019008768A (ja) * | 2017-06-28 | 2019-01-17 | 株式会社ギウォンテクKiwontech | 詐欺メールの危険度を判断する方法およびそのための装置 |
Also Published As
Publication number | Publication date |
---|---|
EP2615787A1 (en) | 2013-07-17 |
JP5573955B2 (ja) | 2014-08-20 |
US9060030B2 (en) | 2015-06-16 |
US20130182720A1 (en) | 2013-07-18 |
JPWO2012032606A1 (ja) | 2013-12-12 |
CN103081417A (zh) | 2013-05-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11115339B2 (en) | Network congestion control method, device, and system | |
JP5573955B2 (ja) | フレーム連結装置 | |
JP7035227B2 (ja) | データパケット検出方法、デバイス、及びシステム | |
US8351352B1 (en) | Methods and apparatus for RBridge hop-by-hop compression and frame aggregation | |
US9225656B2 (en) | Quality of service in a heterogeneous network | |
US10404605B2 (en) | Packet processing method, device and computer storage medium | |
US7486674B2 (en) | Data mirroring in a service | |
US20160119236A1 (en) | Source routing with fabric switches in an ethernet fabric network | |
CN109218178A (zh) | 一种报文处理方法及网络设备 | |
WO2014136864A1 (ja) | パケット書換装置、制御装置、通信システム、パケット送信方法及びプログラム | |
US7944924B2 (en) | Handling of received implicit null packets | |
WO2018132699A1 (en) | Managing network traffic in virtual switches based on logical port identifiers | |
CN102136989B (zh) | 报文传输的方法、系统和设备 | |
US20130163419A1 (en) | Flexible and scalable data link layer flow control for network fabrics | |
JP2009065429A (ja) | パケット転送装置 | |
US10374935B2 (en) | Link discovery method, system, and device | |
WO2013059991A1 (zh) | 数据报文处理方法和系统、报文转发设备 | |
TW201021466A (en) | Data path acceleration of a network stack | |
WO2018149338A1 (zh) | 基于sdn的远端流镜像控制方法、实现方法及相关设备 | |
US10009272B2 (en) | Packet transmission apparatus for processing operations, administration, and maintenance (OAM) packet and protection switching message | |
RU2643469C2 (ru) | Способ передачи фреймов ethernet через программно-конфигурируемые сети (sdn) | |
WO2012105677A1 (ja) | パケット処理装置、パケット処理方法およびプログラム | |
CN110224887A (zh) | 一种以太网y.1564测试帧自动识别的方法与装置 | |
JP6222505B2 (ja) | 入力パラメータを生成するための方法および装置 | |
US9137158B2 (en) | Communication apparatus and communication method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201080068991.7 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 10856956 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2012532760 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2010856956 Country of ref document: EP |