CN1521662A - Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base - Google Patents
Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base Download PDFInfo
- Publication number
- CN1521662A CN1521662A CNA031185649A CN03118564A CN1521662A CN 1521662 A CN1521662 A CN 1521662A CN A031185649 A CNA031185649 A CN A031185649A CN 03118564 A CN03118564 A CN 03118564A CN 1521662 A CN1521662 A CN 1521662A
- Authority
- CN
- China
- Prior art keywords
- llid
- address
- data base
- ethernet
- filtered data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention provides the process and apparatus for realizing Ethernet passive optical network system dynamic filtration data base, wherein a logic link identification filtration data base is utilized in each EPON interface of the local side device for marking the Ethernet packet that passes the passive optical Ethernet network, through which the far-end ONU device can select the Ethernet packet and authorization information sent to the ONU, when the LLID fitting for itself is received, the Ethernet data packet is transferred, when the LLID not fitting for itself is received, and not a single copy broadcast, then the data pack is filtered.
Description
Technical field
The present invention relates to a kind of implementation method and device of Ethernet passive optical network system dynamics filtered data base, this method and apparatus relates to the realization based on the LLID of the EPON of the Ethernet of gigabit (EPON) (LLID) filtered data base, perhaps can be used as the reference of the realization of Ethernet switch (L2switch) filtered data base, perhaps can be used as the reference of realization of filtered data base of the Ethernet interface of multiservice transport platform equipment (MSTP) bridge joint.
Background technology
Described filtered data base is in the Ethernet transmission system, and Ethernet switch or ethernet passive optical network (EPON) equipment filters the employed address table data of Ethernet bag storehouse device according to the medium layer MAC address.By using filtered data base, Ethernet switch can realize that clean culture transmits other port; Ethernet passive optical network system (EPON) local side apparatus can realize that clean culture forwarding Ethernet bag is to remote equipment.Ethernet switch all uses filtered data base at present, and the port numbers and the VLAN characteristic of port transmitted in record, plays an important role for the network efficiency that improves port, the unnecessary broadcasting of minimizing.Ethernet passive optical network system (EPON) then is a kind of popular new generation broadband connecting system that is about to, this system utilizes LLID (LLID) filtered data base of EPON (EPON) can realize high efficiency forwarding strategy, reduces the inner broadcasting of unnecessary EPON (EPON).
In traditional Ethernet switch, this filtered data base adopts central processing unit system, memory address table Database Systems, and learning functionality and aging function are realized by CPU.But along with the raising of port speed, the learning functionality that software is formed and the speed of aging function are too low, the requirement that does not reach system.Some chip manufacturers competitively develop the hardware filtering Database Systems with learning functionality and aging function.
The implementation method of current dynamic filtration database and device need content adressable memory (CAM) to finish, and port speed is high more, and needed pace of learning and inquiry velocity are fast more.Content adressable memory (CAM). writing the maintenance of clauses and subclauses (entry), is the operation of filtered data base more complicated.The dynamic filtration database is exactly to be that aging characteristics are dynamically wiped aging clauses and subclauses according to the operating position that writes clauses and subclauses, thereby safeguards dynamic, an effective filtered data base.Whether clauses and subclauses (entry) wear out mainly is whether these clauses and subclauses are used (being read out) to come to determine by frequent, and as if not being read out, that just thinks is aging in aging interval.
At present the chip manufacturing of some external manufacturers the dynamic filtration database function, but the needed hardware of aging characteristics is comparatively complicated, does not also see corresponding patented technology.The more important thing is the realization of LLID (LLID) filtered data base that can be used for EPON EPON (EPON), belonging to recent ethernet technology and move towards the main flow means---vital technology in the EPON system, this technology is not seen any report and patented claim as yet.
In EPON (EPON) system based on Ethernet, the network topology structure of system is that point is to multiple spot.These are slightly different with conventional ethernet communication mode.Referring to paper " emulation technology of point-to-point LAN and shared LAN among the EPON " (optical communication 2002 6 phases of research, author willow, modest, the He Yan of hair.)
The signal that each ONU sends can only be received by OLT, rather than is received by other ONUs.EPON can be regarded as a kind of novel LAN, and it is a point-to-point on up (ONU-OLT) direction, is that point arrives multiple spot on descending (OLT-ONU) direction, therefore can be referred to as " putting multiple spot LAN ".
Yet up to the present, two kinds of LAN have only been defined in IEEE 802.3 standards: when having only two websites among the LAN, be called " point-to-point LAN "; When three or three above websites link to each other, just can be called " shared LAN ".The denominator of these two kinds of LAN is that the signal that any one website among the LAN sends will be received by other all station among the same LAN.This obviously be with EPON in the characteristic that can not send mutually between the ONUs with received signal be different.Bridge based on the IEEE802.1D standard, only know " point-to-point LAN " and " sharing LAN ", they are not also known a little to multiple spot LAN, go back the neither one standard code now at a point to the medium of multiple spot, how a router/bridge links to each other with terminal station, bridge or other router.In addition, the EPON of point to multipoint mode and existing IEEE 802.1D Spanning-Tree Protocol (STP) and incompatible.If the IEEE 802.1D bridge of an operation Spanning-Tree Protocol is connected on the EPON (putting multiple spot LAN), will cause the multicopy of Frame to be transmitted or certain site can't receiving data frames.The EPON of Ethernet (EPON) system must solve the problem compatible mutually with traditional Ethernet system, realizes the LAN of a kind of point to multiple spot.
The solution of the problems referred to above can adopt " emulation technology of point-to-point LAN and shared LAN " that proposed of this paper.Though these methods implement comparatively simple, it can't utilize the downlink broadcast ability of PON.For point-to-point lan emulation technology, when the downlink broadcast frame, OLT need repeatedly send to each and receive ONU, and this has obviously reduced bandwidth utilization.For putting multiple spot lan emulation technology, this mode can't provide the user to cut apart, and each website can be seen the frame of all transmissions, and waste bandwidth.
Summary of the invention
The present invention is just in order to solve the deficiency of above-mentioned solution.The objective of the invention is to maximally utilise the downlink broadcast ability of PON, and the connection of point-to-point, i.e. clean culture ability still are provided.The EPON of Ethernet (EPON) system employing LLID (LLID) is used for identifying the Ethernet bag by EPON, in order to improve the validity of network.Can select information such as the Ethernet bag that mails to this ONU and mandate by LLID (LLID) far-end ONU equipment.LLID (LLID) filtered data base is just in order to make the Ethernet data bag can carry LLID (LLID) label, the needed parallel database of searching, the EPON interface logic of finishing local side apparatus OLT connects line analysis and assembly method, has improved the validity of network.
Technical scheme of the present invention is: the implementation method of the passive optical network dynamic filtration database of this Ethernet, mainly adopt the LLID filtered data base at each EPON interface of local side apparatus, be used for identifying Ethernet bag by EPON, be that LLID (LLID) filtered data base makes the Ethernet data bag can carry LLID (LLID) label, pass through LLID, far-end ONU equipment can be selected Ethernet bag and the authorization message that mails to this ONU, meet when oneself taking LLID when receiving, just this Ethernet data bag is transmitted, do not meet oneself and take LLID when receiving, when being not single copy broadcast again, just this packet is filtered out, filtered data base comprises mac learning, the aging operation of MAC inquiry and address entry.
The implementation method of the passive optical network dynamic filtration database of described Ethernet, its LLID is the registration process by the EPON system, system has distributed LLID (LLID) for the connection of each ONU, set up link channel, it is single copy broadcast that descending Ethernet data bag utilizes the distinctive broadcast characteristic of EPON, be dealt into far end system ONU, in case terminal system is answered a packet, then its MAC Address and during by the EPON network employed LLID will be filtered database at once and learn, the packet that mails to this destination address later on just all adopts the mode of point-to-point, promptly adopts clean culture LLID value to identify this Ethernet data bag.
The implementation method of the passive optical network dynamic filtration database of described Ethernet, described MAC address learning, be in the forwarding of Ethernet bag, whether the clauses and subclauses of consulting concurrently in the filtered data base have source address, think during coupling that filtered data base has contained these clauses and subclauses, if when not mating, source address is registered in the filtered data base, so that learn new forwarding characteristic.
The implementation method of the passive optical network dynamic filtration database of described Ethernet, its MAC inquiry is to inquire about according to the source MAC of downlink data packet, draw and transmit needed LLID, after receiving packet from exchange side, buffer structure will be initiated once based on destination address " LLID address lookup " process, it is inquired about the destination address of storage by " cam " mechanism, if found this address, just the LLID value is sent layer to RS, if also do not find, to obtain a generic broadcasted LLID value of establishing in advance so is the address, and this LLID value is sent reconciliation sublayer RS layer to EPON.
The implementation method of the passive optical network dynamic filtration database of described Ethernet, the aging operation of its address entry is the address entry of filtered data base, wear out according to the given time interval accurately, MAC Address that OLT receives from ONU and LLID are after being stored by learning process, if in certain digestion time, do not receive the MAC Address that matches, just these clauses and subclauses are wiped, whether whether clauses and subclauses wear out mainly is to be come to determine by frequent the use according to these clauses and subclauses, aging at interval in as if not being read out, just think aging, and the clauses and subclauses that do not meet aging condition continue to be retained in the filtered data base, realize the clauses and subclauses of dynamic refresh database.
A kind of device of realizing the passive optical network dynamic filtration database of above-mentioned arbitrary Ethernet, comprise and provide the study and the service of search capacity to arbitrate, can wipe CAM, priority encoder and LLID storer, and provide aging operation aging storer, learn erasure controller, warn controller, wipe shifting memory, available items shifting memory, ageing state machine and caution shifting memory.
The device of described passive optical network dynamic filtration database, its service arbitration comprises latch, selector switch and frequency divider, to the MAC storer, the study request of MAC Address and the query requests of MAC Address provide Winserve through latch to the target MAC (Media Access Control) address that is used for the source MAC of learning data frame and the Frame that is used to inquire about through selector switch: during the expression MAC address learning carries out or Rinserve: middle signal is carried out in the inquiry of expression MAC Address.
The device of described passive optical network dynamic filtration database, hardware circuit that its aging operation is adopted is the aging storer of record ageing information and shift memory of temporary aging operation information, and these shift memories comprise that wiping shifting memory is used for temporary entry positions information, the available items shifting memory of preparing to wipe and is used to indicate entry positions that new study will deposit to clauses and subclauses and caution shifting memory to be used to the entry positions of indicating those to deposit filtered data base in and entered aging operation limit of consideration when scanning.
The device of described passive optical network dynamic filtration database, it is characterized in that providing the content adressable memory CAM of study and search capacity, adopted and to have wiped CAM, can wipe CAM and have the MAC Address storer, write down the value of the MAC Address of being stored, when clashing, needs can 1 of this row storage of CAM be clashed according to the position of concrete clauses and subclauses existence.
The device of described passive optical network dynamic filtration database, its circuit that can wipe CAM part comprise and are used for scrubber bar target MAC (Media Access Control) address storer and content adressable memory RAM that content adressable memory RAM adopts true double port memory to constitute.
Beneficial effect of the present invention:
A) proposed to utilize the LLID filtered data base to realize the method for ethernet passive optical network system first.The EPON interface logic that this method utilizes filtered data base to finish local side apparatus OLT connects line analysis and assembly method, has improved the validity of network.With LLID (LLID) filtered data base of EPON EPON (EPON), solve and to belong to vital, exigence technical solution in the recent ethernet technology EPON system.
In traditional passive optical network, the method that the ATM-PON system of state-of-the-art system is adopted can not directly be applied mechanically.In this system, VP or VC (being equivalent to the LLID among the EPON) filter and finish dealing with by means of the ATM layer, the VP or the VC establishment of connection that belong to certain remote equipment ONU are to set up by the mode or the signaling method of assignment, though ATM layer process chip used content adressable memory (CAM), but the foundation of VP or VC is not dynamic transient change, thereby do not need learning functionality and aging function, there is not the notion of dynamic filtration database.
Because the shortcoming of ATM-PON system is the dismounting cost height of cell, configuration and maintenance are complicated, can not plug and play.The ATM-PON system does not have extensive popularization and application.The speed of development of EPON EPON (EPON) will substantially exceed the growth momentum of ATM-PON.In addition, ATM layer treatment technology be study and aging function not, can't directly apply in the Ethernet passive optical network system.The present invention has proposed to utilize the LLID filtered data base to realize the method for ethernet passive optical network system first, being filtration unit in the ATM-PON system---the revolutionary character that the ATM layer is handled is improved, make the use LLID (LLID) that EPON EPON (EPON) system can be appropriate, filter the Ethernet data bag, realize the clean culture characteristic, improve the efficient of ethernet passive optical network, overcome the shortcoming of point-to-point lan emulation and shared lan emulation method.
B) use hardware to realize the aging operation of dynamic link libraries
The present invention has adopted the technology of the filtered data base of Ethernet switch, but on the filtered data base using method, a great difference is arranged, the technology of the filtered data base of Ethernet switch is applied to the port of a switch and transmits, only relate in the network element, and being the EPON interface, LLID filtered data base of the present invention is used to assemble LLID, this is identified in the whole ethernet passive optical network and uses, and relates to a plurality of network elements.
The present invention also is improved on the formation of filtered data base, in current Ethernet switch, this filtered data base adopts content adressable memory (CAM) to finish, but determines in aging function and the learning functionality that the entry operation that takies need be helped through by central processing unit (CPU) execution command.Along with the raising of port speed, the learning functionality that software is formed and the speed of aging function are too low.
The present invention has realized the method with hardware first, has realized the aging operation of dynamic filtration database.Whole contents addressable memory (CAM) writes the maintenance of clauses and subclauses (entry), all is to use hardware circuit to realize.Though safeguarding filtered data base is the operation of more complicated, but the hardware circuit device of method of the present invention and use is very succinct, employed hardware resource is also few, and institute realizes that the algorithm that wears out is clear, clear and definite, meets the requirement of IEEE802.1D standard to the dynamic filtration database fully.
Description of drawings
Fig. 1 is the structure of dynamic filtration database.Line among the figure in the square frame has a little oblique line, and there is a numeral of having drawn together on the next door.Line between the expression square frame is many lines, and the radical of line is the represented number of this numeral.
Fig. 2 is the aging sequential chart of dynamic filtration database.Compare_en is for comparing enable signal, and erase_en is for wiping enable signal, and learn_en is the study enable signal.Introduction in conjunction with 4 steps of aging 4 stages and study.
Fig. 3 is aging operational flowchart.4 stages that this figure will wear out have been done detailed description.Design though this process flow diagram is the scan pattern according to K=1024, the present invention is not limited to the value of K.
Fig. 4 is the learning process process flow diagram.This figure has done detailed description with 4 steps of an address entry study.
Fig. 5 is the circuit diagram of dynamic filtration database.
Fig. 6 is the circuit diagram that can wipe CAM.
Fig. 7 is the electrical schematic diagram of service arbitration.
Fig. 8 is the transition diagram of ageing state machine.
Embodiment
The implementation method of EPON LLID (LLID) filtered data base:
In EPON (EPON) system based on gigabit Ethernet, each EPON interface of local side apparatus all has the function of exchange of data link layer.Adopt LLID (LLID) to improve the validity of network in the system.Can select information such as the Ethernet bag that mails to this ONU and mandate by LLID (LLID) far-end ONU equipment, meet when oneself taking LLID when receiving, just this Ethernet data bag is transmitted, do not meet oneself and take LLID when receiving, when being not single copy broadcast again, just this packet is filtered out, thereby improved efficient.Therefore each EPON interface of local side apparatus all needs to have the filtered data base of LLID (LLID), can learn and inquire about the pairing LLID LLID of MAC Address, needed LLID (LLID) when determining to send.
Its LLID is the registration process by the EPON system, system has distributed LLID (LLID) for the connection of each ONU, set up link channel, it is single copy broadcast that descending Ethernet data bag utilizes the distinctive broadcast characteristic of EPON, be dealt into far end system ONU, in case terminal system is answered a packet, then its MAC Address and during by the EPON network employed LLID will be filtered database at once and learn, the packet that mails to this destination address later on just all adopts the mode of point-to-point, promptly adopts clean culture LLID value to identify this Ethernet data bag.
Simultaneously, how these Database Systems utilize aging grade for measure, use Limited resources, yet just can reach MAC Address are implemented unconfined effectively dynamic memory.The realization of LLID (LLID) filtered data base is based on the problem that must solve in EPON (EPON) system of gigabit Ethernet.The present invention comprises the implementation method and the filtered data base implement device of such filtered data base, comprises the implement device of mac learning, MAC inquiry and aging operation.
Described MAC Address inquiry is before the Ethernet bag is transmitted, to consult the clauses and subclauses in the filtered data base, and the port numbers of during coupling desire being transmitted is read.MAC Address inquiry must be carried out fast, finishes in the transmitting time of an Ethernet bag at least, otherwise will restrict the transmission speed of the transmission of Ethernet bag.Generally need the parallel query system.
Described MAC address learning, in the forwarding of Ethernet bag, whether the parallel clauses and subclauses of consulting in the filtered data base have source address, think during coupling that filtered data base has contained these clauses and subclauses, if when not mating, source address is registered in the filtered data base, so that learn new forwarding characteristic.
The described filtered data base that dynamically is meant has the aging method of address entry, make the clauses and subclauses of address filtering database to wear out according to the given time interval accurately, and the clauses and subclauses that do not meet aging condition continue to be retained in the filtered data base, realize the clauses and subclauses of dynamic refresh database.
Further specify the embodiment of filtered data base below in conjunction with accompanying drawing:
1.EPON the method for the realization of LLID (LLID) filtered data base:
At first by the registration process of EPON system, system sets up link channel for the connection of each ONU has distributed LLID (LLID).Descending Ethernet data bag at the beginning and do not know how to select suitable LLID, therefore just utilizing the distinctive broadcast characteristic of EPON is single copy broadcast, is dealt into far end system ONU.In case terminal system is answered a packet, then its MAC Address and during by the EPON network employed LLID will be filtered database at once and learn.The packet that mails to this destination address (this DA) later on just all adopts the mode of point-to-point.
In general the identification method of LLID is such, and LLID is 16 bit architecture:
The 0-14 bit is clean culture LLID territory
15 bits are mode bit (0=P2P, 1=broadcasting)
The position of LLID is: originally the lead code in the packet header of ethernet frame and SFD are made up of 8 following bytes:
10101010?10101010?10101010?10101010?10101010?1010101010101010?10101011。
The order of serial transmission is from left to right.The most left BIT of each byte is LSB, and the rightest BIT is MSB.The sign of described LLID be exactly lead code and SFD when passing through GMII, replaced by the structure that defines below:
The 0-1 byte is as SPD 10,101,010 10101011
2-4 is as reserve bytes: 10101010
5-6 byte conduct: LLID 11,111,111 11111111 (default value)
7 bytes use x8+x2+x+1 to calculate the CRC of 8 bits as CRC-8.The implementation method of the passive optical network dynamic filtration database of described Ethernet, its MAC address learning, be from the Ethernet that EPON ODN side is received wrap in be forwarded to the exchange kernel in, whether the clauses and subclauses of consulting concurrently in the filtered data base have source address, think during coupling that filtered data base has contained these clauses and subclauses, if when not mating, source address is registered in the filtered data base, so that learn new forwarding characteristic.
Concrete EPON LLID (LLID) filtered data base relevant with database has following operation.
A) learn according to the source MAC of upstream data bag, write down the pairing LLID of this MAC Address:
OLT receives that buffer structure will be initiated once based on source address " address learning " process after the packet that ONU sends over.It is learnt the source address of storage by " address lookup " mechanism, if found this address, current learning process just finishes.If also do not find, will obtain an address space from address table so, and deposit this address and corresponding LLID value in.
B) inquire about according to the source MAC of downlink data packet, draw and transmit needed LLID:
OLT receives from exchange side after the packet that buffer structure will be initiated once based on destination address " LLID (address) inquiry " process.It is inquired about the destination address of storage by " cam " mechanism, if found this address, just the LLID value is sent the layer to RS.The EPON interface is stamped the LLID sign at the RS layer with the Ethernet bag.If also do not find, will obtain a generic broadcasted LLID value (address) of establishing in advance so, and this LLID value will be sent reconciliation sublayer RS layer to EPON.
C) the MAC Address clauses and subclauses are aging
MAC Address that OLT receives from ONU and LLID are after being stored by learning process, if in certain digestion time, do not receive the MAC Address that matches, just these clauses and subclauses should be discharged, use for other MAC Address, therefore need carry out address aging is purpose scanning.During scanning, the clauses and subclauses that meet aging condition are wiped.Whether clauses and subclauses (entry) wear out mainly is whether these clauses and subclauses are used (being read out) to come to determine by frequent, and as if not being read out, that just thinks is aging in aging interval.In order to realize aging this function, require aging bit (Aging Bit) storer is set, and it is searched, operations such as removing are set.
2. the present invention has gone out to provide outside the filtered data base method, also comprises circuit structure:
Main hardware as shown in Figure 1, according to summary of the invention noted earlier and purpose, described filtered data base hardware implementation mode has comprised provides study and the service of search capacity arbitration, can wipe CAM, priority encoder and LLID storer, and the aging aging storer of operating be provided, learn erasure controller, warn controller, wipe shifting memory, available items shifting memory, ageing state machine, comparer and caution shifting memory.
The service arbitration provides the arbitration of address lookup and address maintenance (comprising study and scanning) service, makes the use that can wipe CAM priority encoder and LLID storer be used alternatingly according to time division multiplexing mode.
The described CAM that wipes provides the function of content adressable memory (CAM).An important component part can wiping CAM is content adressable memory (CAM), and width is 48 bits, and the degree of depth is K (K=2^L, for example K=1024), can be that 1 dual-ported memory constitutes by address width 48+L data width.Content adressable memory (CAM) has read operation and write operation.Read operation is the address entry (K=2^L, for example K=1024) all according to 48 bit mac address fast query, sees whether contain these clauses and subclauses, and matcher is exportable high level.Another constituent that can wipe content adressable memory is the pairing 48 bit mac addressed memories of recording storage clauses and subclauses.
Described priority encoder will be exported the matcher of high level and weave into the L bit code according to priority encoder.
The pairing LLID of MAC Address that described LLID storer will be learnt carries out memory storing.
Described aging storer is K 1 bit memory, has the independently true dual-ported memory of 2 cover read-writes.The pairing LLID of the MAC Address of learning carries out the aging bit memory of memory storing, if 1 is fresh address.Being 0 is to be about to wear out.
Every aging bit that has just deposited the address entry correspondence in is made as 1.It is 1 that every address entry that just had been read out once also refreshes.
Described study erasure controller is used to control study and the erasing move that can wipe CAM.In the stage 2 of scan pattern: wear out and wipe (aging out), according to the content of wiping shifting memory, pairing content of wiping content memorizer is wiped in control.Step 2 in mode of learning: be used for learning according to the value of control in the write step with the content of shift register C.
Described caution controller is temporarily stored the store items that needs caution.Be used for the removing of aging bit on the one hand, i.e. the warning that pairing clauses and subclauses are worn out.This content also is used to reaffirm whether reach aging condition in scanning next time.
The described shifting memory of wiping is stored the address entry that empirical tests can wear out well.
Described available items shifting memory is stored the good operable address entry of empirical tests, and the value of this content is that 1 pairing clauses and subclauses are the clauses and subclauses that will write.
Comparer 12 is NAND gate circuit among the figure.
Described ageing state machine carries out the device that each unit component work was counted and controlled to digestion time at interval, contains state machine.
3. the course of work is described
Entire circuit work has two patterns.Mode of learning and scanning (SCAN) pattern.In most cases, circuit is all under mode of learning.Under the mode of learning, constantly new address is learnt.Approximately several ten thousand/ time is scan pattern.Scan pattern temporarily stops learning manipulation, the operation of only wearing out.Setting by pattern, in fact reached such effect, promptly the digestion time interval counter for all address entrys all adopts common one, when rolling counters forward arrives scanning constantly, entire circuit enters scan pattern, the operation of wearing out immediately, the aging operation of all K clauses and subclauses concentrates under the scan pattern to be carried out.
Aging decision condition is: those address entrys of the aging bit of having disposed are not read once again every interior at aging compartment.
Implement under the concrete scan pattern agingly to form (referring to Fig. 3) by 4 stages.
This four-stage is: comparison phase, wear out and wipe, write down and the caution of wearing out (removing aging bit).
Stage 1 is a comparison phase
Whether checking reaches aging condition.The content of output that will caution shifting memory A and current RAM is (NOT-AND operation) relatively, if be that 0 value then meets aging condition all, deposit 1 in and is worth and wipes shifting memory 12, shows that the pairing clauses and subclauses in this address are will be in operation subsequently deleted to fall.Here exist the definition of the aging bit of aging bit memory 4 to be, if this address was just read, just be set to 1, explanation is fresh address.
This stage has also been stored the content of available shifting memory simultaneously, and the above-mentioned logic NOT value of wiping storer of this content is for so that under the future mode of learning, provide the positional information that can be used for putting down in writing clauses and subclauses.
Stage 2 is aging wipe (aging out):
After relatively, wipe shifting memory and had those positional informations through the clauses and subclauses confirming to be wiped free of.The aging stage of wiping is just removed CAM and LLID one by one according to the content of wiping shifting memory 8 RAM has finished clear operation up to k clauses and subclauses, aging end of operation.
Stage 3 is record again:
For next step caution operation of wearing out, simultaneously for the compare operation of scanning process next time, the address entry that keeps a record again and have at present.Concrete is exactly to read the content (style) of aging storer 4 to caution shifting memory 11.
Stage 4 is aging caution:
Aging caution is done the warning that a desire will be deleted with existing address entry exactly, when scanning, if this warning message is not reacted (set), just is not read out once next time, and these clauses and subclauses are with deleted.Aging caution operation is exactly a bit of removing aging storer 4 correspondences according to the content of caution shifting memory.Simultaneously, the content of reading is also read in by caution shifting memory input port, when K bar content all by after a single reading, the content when the caution shifting memory returns to aging caution operation beginning.
Under the situation of scan pattern, except above-mentioned aging operation, locating function (read operation) is also carrying out normally simultaneously.
Generally, scanning can all be worn out to aging address in aging (as 5 seconds) at interval accurately.Also can set in advance at interval aging by register.Be the aging sequential chart of dynamic filtration database as shown in Figure 2.
Between twice scanning, circuit all is to be in mode of learning.
Under the mode of learning situation, CAM, address table carry out normal learning functionality (adopting wMAC) and locating function (RMAC is adopted in read operation).
In the mode of learning, as long as free address entry just can be learnt many new MAC Address.Wherein a rule purpose learning process comprises 4 steps (consulting Fig. 4)
Whether new step 1 be to detect address;
At first consult and deposit CAM whether in, need to use the address of reading of CAM.By the service arbitration, learning manipulation can obtain query time (the service arbitration is the mechanism that realizes two address multiplexes), determines whether content memorizer contains current address.
Step 2 is to write:
When the determined address entry of content in the available shifting memory 9 being used to learn MAC Address and LLID thereof storage.Concrete, MAC Address is write the MAC storer that can wipe shifting memory, the appropriate address clauses and subclauses are write among the CAM, the LLID value is deposited in the LLID storer 3.
Step 3: upgrade aging bit:
When reading, in passing after finding MAC Address, refresh aging bit, make it to equal 1 at once.Enable (similar choosing) signal this moment, only the aging storer 4 of gating carries out write operation.Can not change the content of the LLID storer 3 of storage LLID.Step 4 is to determine next bar writing position:
Read available shifting memory 9, up to finding an available items.Counter is counted when reading, and when stopping to read available shifting memory, the value of this counter is writing position (address).
Fig. 5 is the circuit diagram of the case study on implementation of dynamic filtration database.
By using Fig. 1 to Fig. 4 and above-mentioned introduction, available field-programmable integrated circuit (FPGA) is achieved the dynamic filtration database.Point out method in conjunction with the present invention, the LLID dynamic filtration database that can be used for EPON is realized.
Wherein the LLID storer adopts 16 bit wides, 1024 (=K) individual address space storeies.Shifting memory B is for wiping shifting memory, 1024; Shifting memory A is caution shifting memory, 1024; Shifting memory C is available shifting memory, 1024.CNT is a counter among the figure, and dff is a d type flip flop.
Fig. 8 is the transition diagram of ageing state machine. provide the transfer process of 5 signals in the ageing state machine output map 2, be 5 seconds digestion period.System clock is 62.5M.
Fig. 6 is the circuit theory diagrams that can wipe the CAM part.
Be used for scrubber bar target MAC (Media Access Control) address storer among the figure, adopt 1024 address spaces of 48 bit wides.Content adressable memory adopts true double port memory to constitute.The A port that is used to write is a 2^58 address space, and data width is 1, and the B port that is used to read is a 2^48 address space, and data width is 1024.
Fig. 7 is the electrical schematic diagram of service arbitration, comprises latch, selector switch and frequency divider.
WMAC[47:0]: the source MAC that is used for the learning data frame;
RMAC[47:0]: the target MAC (Media Access Control) address of the Frame that is used to inquire about; The study request of Wreq:MAC address;
The query requests of Rreq:MAC address;
Winserve: in the service, during the expression MAC address learning carries out;
Rinserve: in the service, during the inquiry of expression MAC Address is carried out;
Scan mode: in the scanning, the expression control plane just the scan address storage unit wherein frequency divider to tell the cycle according to system clock be 8*32=256 delicate (times of 32 bytes), dutycycle is 50%.As long as the service decision logic has query requests Rreq at high level, with regard to corresponding Rinserve.As long as the service decision logic has study request Wreq in low level, with regard to corresponding Winserve, at Scan_mode: not corresponding study request of valid period in the scanning.The speed that generally can wipe CAM all=256 delicate in.
Core of the present invention is to adopt the LLID filtered data base at each EPON interface of local side apparatus, can select information such as the Ethernet bag that mails to this ONU and mandate by LLID (LLID) far-end ONU equipment, the EPON interface logic of finishing local side apparatus OLT connects line analysis and assembly method, has improved the validity of network.Therefore, every employing LLID filtered data base is realized this purpose, all belongs to protection scope of the present invention.
Claims (10)
1. the implementation method of an Ethernet passive optical network system dynamics filtered data base, mainly adopt the LLID filtered data base at each EPON interface of local side apparatus, be used for identifying Ethernet bag by EPON, be that LLID (LLID) filtered data base makes the Ethernet data bag can carry LLID (LLID) label, pass through LLID, far-end ONU equipment can be selected Ethernet bag and the authorization message that mails to this ONU, meet when oneself taking LLID when receiving, just this Ethernet data bag is transmitted, do not meet oneself and take LLID when receiving, when being not single copy broadcast again, just this packet is filtered out, filtered data base comprises mac learning, the aging operation of MAC inquiry and address entry.
2. the implementation method of Ethernet passive optical network system dynamics filtered data base according to claim 1, it is characterized in that LLID is the registration process by the EPON system, system has distributed LLID LLID for the connection of each ONU, set up link channel, it is single copy broadcast that descending Ethernet data bag utilizes the distinctive broadcast characteristic of EPON, be dealt into far end system ONU, in case terminal system is answered a packet, then its MAC Address and during by the EPON network employed LLID will be filtered database at once and learn, the packet that mails to this destination address later on just all adopts the mode of point-to-point, promptly adopts clean culture LLID value to identify this Ethernet data bag.
3. the implementation method of Ethernet passive optical network system dynamics filtered data base according to claim 1 and 2, it is characterized in that described MAC address learning, be in the forwarding of Ethernet bag, whether the clauses and subclauses of consulting concurrently in the filtered data base have source address, think during coupling that filtered data base has contained these clauses and subclauses, if when not mating, source address is registered in the filtered data base, so that learn new forwarding characteristic.
4. the implementation method of Ethernet passive optical network system dynamics filtered data base according to claim 1 and 2, it is characterized in that the MAC inquiry is to inquire about according to the source MAC of downlink data packet, draw and transmit needed LLID, after receiving packet from exchange side, buffer structure will be initiated once based on destination address " LLID address lookup " process, it is inquired about the destination address of storage by " cam " mechanism, if found this address; Just the LLID value is sent the layer to RS, if also do not find, will obtain a generic broadcasted LLID value of establishing in advance so is the address, and this LLID value is sent reconciliation sublayer RS layer to EPON.
5. the implementation method of Ethernet passive optical network system dynamics filtered data base according to claim 1 and 2, it is characterized in that the aging operation of address entry is the address entry of filtered data base, wear out according to the given time interval accurately, MAC Address that OLT receives from ONU and LLID are after being stored by learning process, if in certain digestion time, do not receive the MAC Address that matches, just these clauses and subclauses are wiped, whether whether clauses and subclauses wear out mainly is to be come to determine by frequent the use according to these clauses and subclauses, aging at interval in as if not being read out, just think aging, and the clauses and subclauses that do not meet aging condition continue to be retained in the filtered data base, realize the clauses and subclauses of dynamic refresh database.
6. device of realizing Ethernet passive optical network system dynamics filtered data base, comprise study and the service of search capacity arbitration are provided, can wipe CAM, priority encoder and LLID storer, and provide aging operation aging storer, learn erasure controller, warn controller, wipe shifting memory, available items shifting memory, ageing state machine and warn shifting memory.
7. according to the device of the described passive optical network dynamic filtration of claim 6 database, it is characterized in that the service arbitration comprises latch, selector switch and frequency divider, to the MAC storer that can wipe CAM, the study request of MAC Address and the query requests of MAC Address provide Winserve through latch to the target MAC (Media Access Control) address that is used for the source MAC of learning data frame and the Frame that is used to inquire about through selector switch: during the expression MAC address learning carries out or Rinserve: middle signal is carried out in the inquiry of expression MAC Address.
8. according to the device of claim 6 or 7 described passive optical network dynamic filtration databases, it is characterized in that the hardware circuit that aging operation is adopted is the aging storer of record ageing information and the shift memory of temporary aging operation information, these shift memories comprise that wiping shifting memory is used for temporary entry positions information, the available items shifting memory of preparing to wipe and is used to indicate entry positions that new study will deposit to clauses and subclauses and caution shifting memory to be used to the entry positions of indicating those to deposit filtered data base in and entered aging operation limit of consideration when scanning.
9. require the device of described passive optical network dynamic filtration database according to right 6 or 7, it is characterized in that providing the content adressable memory CAM of study and search capacity, adopted and to have wiped CAM, can wipe CAM and have the MAC Address storer, write down the value of the MAC Address of being stored, when clashing, needs can 1 of this row storage of CAM be clashed according to the position of concrete clauses and subclauses existence.
10. according to the device of claim 6 or 7 described passive optical network dynamic filtration databases, the circuit that it is characterized in that wiping the CAM part comprises and is used for scrubber bar target MAC (Media Access Control) address storer and content adressable memory CAM that content adressable memory CAM adopts true double port memory to constitute.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031185649A CN1291341C (en) | 2003-01-30 | 2003-01-30 | Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB031185649A CN1291341C (en) | 2003-01-30 | 2003-01-30 | Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1521662A true CN1521662A (en) | 2004-08-18 |
| CN1291341C CN1291341C (en) | 2006-12-20 |
Family
ID=34284914
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB031185649A Expired - Fee Related CN1291341C (en) | 2003-01-30 | 2003-01-30 | Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1291341C (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100435527C (en) * | 2005-08-25 | 2008-11-19 | 广东省电信有限公司研究院 | Method for realizing efficient video multicasting in ethernet passive optical entwork system |
| CN100438439C (en) * | 2006-05-19 | 2008-11-26 | 华为技术有限公司 | Method for preventing MAC address cheat |
| CN100459499C (en) * | 2005-10-12 | 2009-02-04 | 杭州华三通信技术有限公司 | Method for transmitting upgrade software to optical network unit in Ethernet passive optical network |
| CN101098291B (en) * | 2006-06-29 | 2010-04-14 | 中兴通讯股份有限公司 | Method for preventing disturbance of medium accessing control address table on access equipment |
| CN101047670B (en) * | 2006-04-14 | 2010-04-21 | 华为技术有限公司 | MAC address table ageing, operation method and process system thereof |
| CN101110611B (en) * | 2007-08-31 | 2010-12-08 | 杭州华三通信技术有限公司 | Method and apparatus for echo wave filtering in EOC system |
| CN102340708A (en) * | 2010-07-22 | 2012-02-01 | 高通创锐讯通讯科技(上海)有限公司 | Method for two-layer multicast in EPON (Ethernet Passive Optical Network) system |
| CN101494595B (en) * | 2008-01-22 | 2012-02-08 | 杭州华三通信技术有限公司 | Method and system for forwarding data message, coaxial line terminal and optical network unit |
| CN101557541B (en) * | 2009-05-26 | 2012-04-18 | 杭州华三通信技术有限公司 | Data package transmission method, system and device thereof |
| CN1976260B (en) * | 2005-11-28 | 2012-06-20 | 阿尔卡特公司 | Burst mode optical receiver |
| CN102932227A (en) * | 2012-10-29 | 2013-02-13 | 烽火通信科技股份有限公司 | Method for realizing virtual local area network (VLAN) network bridge connection technology |
| CN101188461B (en) * | 2005-11-28 | 2013-09-04 | 阿尔卡特公司 | Method and system for optical receiver to receive signal from a plurality of optical network units |
| CN103560916A (en) * | 2013-11-04 | 2014-02-05 | 中国电子科技集团公司第四十一研究所 | High-speed data packet filtering circuit and method |
| CN104378303A (en) * | 2014-11-21 | 2015-02-25 | 瑞斯康达科技发展股份有限公司 | Method and system for forwarding multicast service, ONU and OLT |
| CN104506436A (en) * | 2014-12-22 | 2015-04-08 | 上海斐讯数据通信技术有限公司 | Data message sending method for Ethernet passive optical network |
| CN105282268A (en) * | 2015-10-22 | 2016-01-27 | 上海斐讯数据通信技术有限公司 | Management method and system for passive optical network media access control (PON MAC) address information for optical line terminal (OLT) equipment |
-
2003
- 2003-01-30 CN CNB031185649A patent/CN1291341C/en not_active Expired - Fee Related
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100435527C (en) * | 2005-08-25 | 2008-11-19 | 广东省电信有限公司研究院 | Method for realizing efficient video multicasting in ethernet passive optical entwork system |
| CN100459499C (en) * | 2005-10-12 | 2009-02-04 | 杭州华三通信技术有限公司 | Method for transmitting upgrade software to optical network unit in Ethernet passive optical network |
| CN101188461B (en) * | 2005-11-28 | 2013-09-04 | 阿尔卡特公司 | Method and system for optical receiver to receive signal from a plurality of optical network units |
| CN1976260B (en) * | 2005-11-28 | 2012-06-20 | 阿尔卡特公司 | Burst mode optical receiver |
| CN101047670B (en) * | 2006-04-14 | 2010-04-21 | 华为技术有限公司 | MAC address table ageing, operation method and process system thereof |
| CN100438439C (en) * | 2006-05-19 | 2008-11-26 | 华为技术有限公司 | Method for preventing MAC address cheat |
| CN101098291B (en) * | 2006-06-29 | 2010-04-14 | 中兴通讯股份有限公司 | Method for preventing disturbance of medium accessing control address table on access equipment |
| CN101110611B (en) * | 2007-08-31 | 2010-12-08 | 杭州华三通信技术有限公司 | Method and apparatus for echo wave filtering in EOC system |
| CN101494595B (en) * | 2008-01-22 | 2012-02-08 | 杭州华三通信技术有限公司 | Method and system for forwarding data message, coaxial line terminal and optical network unit |
| CN101557541B (en) * | 2009-05-26 | 2012-04-18 | 杭州华三通信技术有限公司 | Data package transmission method, system and device thereof |
| CN102340708A (en) * | 2010-07-22 | 2012-02-01 | 高通创锐讯通讯科技(上海)有限公司 | Method for two-layer multicast in EPON (Ethernet Passive Optical Network) system |
| CN102932227A (en) * | 2012-10-29 | 2013-02-13 | 烽火通信科技股份有限公司 | Method for realizing virtual local area network (VLAN) network bridge connection technology |
| CN102932227B (en) * | 2012-10-29 | 2015-07-01 | 烽火通信科技股份有限公司 | Method for realizing virtual local area network (VLAN) network bridge connection technology |
| CN103560916A (en) * | 2013-11-04 | 2014-02-05 | 中国电子科技集团公司第四十一研究所 | High-speed data packet filtering circuit and method |
| CN103560916B (en) * | 2013-11-04 | 2017-01-11 | 中国电子科技集团公司第四十一研究所 | High-speed data packet filtering circuit and method |
| CN104378303A (en) * | 2014-11-21 | 2015-02-25 | 瑞斯康达科技发展股份有限公司 | Method and system for forwarding multicast service, ONU and OLT |
| CN104378303B (en) * | 2014-11-21 | 2018-03-23 | 瑞斯康达科技发展股份有限公司 | Method and system, ONU and the OLT of forwarding multicasting business |
| CN104506436A (en) * | 2014-12-22 | 2015-04-08 | 上海斐讯数据通信技术有限公司 | Data message sending method for Ethernet passive optical network |
| CN105282268A (en) * | 2015-10-22 | 2016-01-27 | 上海斐讯数据通信技术有限公司 | Management method and system for passive optical network media access control (PON MAC) address information for optical line terminal (OLT) equipment |
| CN105282268B (en) * | 2015-10-22 | 2018-05-01 | 上海斐讯数据通信技术有限公司 | Management method and its system for the PON mac address informations of OLT device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1291341C (en) | 2006-12-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1291341C (en) | Method and apparatus for realizing Ethernet passive optical network system dynamic filtration data base | |
| CN1021395C (en) | synchronous-asynchronous and asynchronous-synchronous converter | |
| KR100585327B1 (en) | Method for adaptively reassigning addresses of nodes according to changes in volumn of wireless network | |
| CN1770720A (en) | Method and apparatus for supporting multiple wireless universal serial bus (USB) hosts in coordinator-based wireless network | |
| CN105847148B (en) | A kind of data frame transmission control method and system | |
| CN1852240A (en) | Bridge-connection transmitting method | |
| CN1761238A (en) | Optimization of routing forwarding database in a network processor | |
| CN1402480A (en) | Method for managing address of medium access control bridge, and said bridge | |
| CN1433190A (en) | Exchange node classifying and marking rules | |
| CN1809032A (en) | Method of dynamically learning address on MAC layer | |
| CN1946054A (en) | Transmission method and device for high speed data flow and data exchange device | |
| CN1691828A (en) | Handover supporting method and apparatus and handover method and apparatus | |
| CN1691629A (en) | Method for implementing layer-2 equipment interconnection in resilient packet ring (RPR) based network | |
| CN1856967A (en) | Automatic switching of setting associated with network | |
| CN1773949A (en) | Switching in method for virtual special network and realizing apparatus | |
| CN101079775A (en) | Method for dividing virtual LAN, data transfer and wireless packet gateway | |
| CN1741499A (en) | Virtual circuit exchanging method based on MAC studying | |
| CN1376348A (en) | Method for managing bandwidth in a communication network comprising a cordless connection | |
| CN1677951A (en) | Data exchange method based on virtual local area network | |
| CN1767495A (en) | Method for assuring two-layer Ethernet exchanger data safety in city area transmission equipment | |
| CN1249956C (en) | Method for virtual Ethernet adapter card | |
| CN1677950A (en) | Data exchange method based on virtual local area network | |
| CN1282342C (en) | Method for implementing transparent bridging of radio local network bridging | |
| CN1719769A (en) | Method for classifying received data pocket in network apparatus | |
| CN1299459C (en) | Channel sharing ring protection supported protection switching protocol implementing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20061220 Termination date: 20160130 |
|
| EXPY | Termination of patent right or utility model |