1 / 64

Chapter 10 Broadband ATM networks

Chapter 10 Broadband ATM networks. Outline. 10.1 introduction. Introduction. Cells: source information associated with each call being first converted into small fixed-sized packets Cell switching:the cell streams relating to different calls are multiplexed together

felix
Télécharger la présentation

Chapter 10 Broadband ATM networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Chapter 10 Broadband ATM networks

  2. Outline • 10.1 introduction

  3. Introduction • Cells: source information associated with each call being first converted into small fixed-sized packets • Cell switching:the cell streams relating to different calls are multiplexed together • Asynchronous transfer mode (ATM) : the cells relating to different calls have varying time intervals between them • Virtual circuit identifier(VCI): identify the packets relating to the different calls on each line rather than the different calls in the total network

  4. Introduction • Small cells have advantages for constant bit rate traffic but disadvantages that the overheads associated with each cell are higher • A cell size is 53 bytes/octets • A cell comprises a 48-byte payload field with a 5-byte header for the VCI and other fields

  5. 10.2 cell format and switching principles • Protocol connection identifier (PCI) :the virtual circuit identifier used on each link • The PCI is made up of virtual path identifier (VPI) and virtual channel identifier (VCI) • Figure 10.1 • Cell loss priority(CLP):in ATM cell format, it enable the user to specify a preference as to which cells should be discarded

  6. 10.3 switch architectures • Input controller (IC):involves a simple look-up and mapping operation of the VPI/VCI in the header of the incoming cell into the corresponding output VPI/VCI • Buffering is provided in the IC or output controllers to hold simultaneously arriving cells • Control processor is to download routing information into the routing tables in each IC

  7. 10.3 switch architectures • Switch fabrics can be classified as either time division or space division • In a time-division switch, a time-division backplane bus is capable of transferring N cells in a single cell arrival time • In a space-division switch, the switch fabric comprises a matrix of interconnected switching elements that collectively provide a number of alternative paths through the switch

  8. 10.3 switch architectures • Delta switch matrix: a switching fabric comprises multiple switching stages • Figure 10.4 • Routing tag: if the tag bit is a binary 0, a cell arriving on either input is routed to the upper output, else routed to the lover output • Blocking: in figure 10.4, those addressed to ports 2 and 3 arrive at the second switching element simultaneously

  9. Figure 10.4 Delta switch matrix example.

  10. 10.3 switch architectures • One approach to overcoming blocking is for the switching element to discard one of the two cells • Discarding cells can lead to an unacceptably high cell loss rate in large switches • A second approach is to perform the switching operation several times faster than cell arrival rate • A third approach is to introduce buffering into each switching element but has the disadvantage of introducing additional delay to the switching operation

  11. 10.3 switch architectures • Batcher-Banyan switch: no two cells entering the switching matrix require the same output port and no common interconnecting links within the paths through the switch • If two cells arrive simultaneously at different input ports that require the same output port, just one cell is selected for transfer • Sorting matrix and shuffle exchange: all arriving cells are ordered such that they follow a unique path through the switching matrix • Figure 10.5

  12. Figure 10.5 Batcher–Banyan switch matrix.

  13. 10.4 protocol architecture • C-plane are concerned with signaling • U-plane communicate on a user-to-user basis with a similar protocol set in the destination station • M-plane are concerned with the management of the station • Figure 10.6

  14. Figure 10.6 ATM protocol architecture.

  15. 10.4 protocol architecture10.4.1 ATM adaptation layer • ATM adaptation layer(AAL) : it performs an adaptation function between the service provided to the user layer and the underlying ATM layer • Service classes : the AAL layer provides a range of alternative service types • Three criteria • The existence of a time relationship between the source and destination users(voice) • The bit rate associated with the transfer(constant or variable) • Connection mode(connectionless or connection-oriented)

  16. 10.4.1 ATM adaptation layer • Both AAL1 and AAL 2 are connection-oriented • AAL1 provides constant bit rate service • AAL2 provides variable bit rate service • AAL1 is known as circuit(switched) emulationThe convergence sublayer (CS) performs a convergence function between the service offered at the layer interface and the underlying layer

  17. 10.4.1 ATM adaptation layer • The segmentation and reassembly (SAR) sublayer performs the necessary segmentation of the source information ready for transfer • Service access point(SAP): is used to direct all information submitted for transfer to the appropriate CS protocol • Figure 10.7

  18. 10.4.1 ATM adaptation layer • Cell losses are overcome in an agreed way by inserting dummy bits into the delivered stream • Cell transfer delay variations are compensated for by buffering segments at the destination • In Figure10.8, SN and SNP field is used to protect the sequence number against single bit errors • Figure 10.8

  19. Figure 10.8 SAR protocol data unit types: (a) AAL 1; (b) AAL 2.

  20. 10.4.1 ATM adaptation layer • AAL3/4 provides a connectionless data transfer service for the transfer of variable length frames up to 65535 bytes in length • The resulting CS PDU is segmented by the SAR protocol into multiple 48-octet SAR-PDUs • MID: in figure10.9 for receiving SAR protocol to relate each incoming segment to the correct PDU • Figure 10.9

  21. Figure 10.9 CS and SAR PDU formats: (a) AAL3/4; (b) AAL 5.

  22. 10.4.1 ATM adaptation layer • Simple and efficient adaptation layer(SEAL):AAL5 provides a similar service to AAL3/4 but with a reduced number of fields in the CS and SAR PDUs • There is no head or tail associated with the SAR PDU • Signaling AAL or SAAL: AAL5 is also used as the AAL layer in the C-plane for the segmentation and reassembly of messages • Figure 10.9

  23. 10.4.2 ATM layer • Its main function is to assign a header to the segment streams generated by the AAL relating to a particular call • Constant bit rate(CBR):supports isochronous traffic • Variable bit rate/real time(VBR/RT): supports variable bit rate traffic with real-time requirements • VBR/NRT • Available bit rate (ABR): bursty traffic with a known minimum, mean, peak rate • Unspecified bit rate(UBR)

  24. 10.4.2 ATM layer • Traffic descriptor includes parameters: • Peak cell rate(PCR) • Sustained cell rate(SCR) • Minimum cell rate (MCR) • Cell delay variation tolerance(CDVT) • Cell loss ratio(CLR) • Cell transfer delay(CTD) • Cell delay variation (CDV) • Generic cell rate algorithm(GCRA): determines whether the parameters for this call/VC conform to the agreed contract(Figure 10.10)

  25. Figure 10.10 Principle of operation of generic cell rate algorithm.

  26. 10.5 ATM LANs • Remote concentrator unit(RCT) is to minimize the number of ports • The conversion of all information into streams of fixed-sized cells has the advantage that the cells relating to the different types of call can be switched in a uniform way • Broadcast server enables a user at a worksation to set up on demand a videoconferencing session between many workstations • Figure 10.11

  27. Figure 10.11 ATM LAN schematic.

  28. 10.5 ATM LANs • Legacy LANs • With on-demand connections, the user device sends a request for a switched virtual connection (SVC) to be set up between itself and destination • Signaling control point(SCP) : a central control • Signaling virtual channel connections (SVCCs): all the signaling messages are transferred across the network to and from the SCP over a set of VCs

  29. 10.5 ATM LANs • Permanent VCs (PVCs): between each workstations and servers • Connectionless server (CLS): a central data forwarding point • Figure 10.12

  30. Figure 10.12 ATM LAN routing example: (a) network segment;(b) example routing table entries.

  31. 10.5.1 call processingconnection-oriented calls • Services such as telephony require a separate VC to be set up for the duration of the call • Signaling virtual channel connection (SVCC): All messages relating to the setting up and clearing of calls are carried over a separate channel • When an SVCC is set up by network management, an entry is also made in the routing table of SCP • To initiate the setting up of a call, the user of the station follows a dialog which involves specifying the address of the required recipient of the call and the call type

  32. 10.5.1 call processingconnection-oriented calls • Recommendation Q.2931:to support the exchange of signaling messages • If the response is positive, this is communicated back to the SCP by adding entries into each switch routing table that links the port VPI/VCI of the calling station • To set up a conferencing session, the initiator of the session communicates with the SCP • The SCP communicates with the user of each of these stations to ascertain their availability and willingness to participate

  33. 10.5.1 call processingconnectionless calls • One for use with a bridged LAN and the other for use with a router-based LAN • LAN Forum: LAN emulation(LE), LE configuration server (LECS), broadcast and unknown-address server (BUS) • The LECS is used by all stations that are connected to the total ATM LAN to determine the ATM addresses of the LES and BUS • The LES provides an address resolution service to convert from 48-bit MAC addresses into 20-byte ATM addresses

  34. 10.5.1 call processingconnectionless calls • The BUS is responsible for supporting multicasting /broadcasting and for relaying frames to stations whose MAC address is unknown by the LES • Figure 10.13 • All stations have an LE client (LEC) • At the end of the initialization phase, both the LES and BUS have the set of addresses of all active stations that are connected to the ATM LAN • In the first case the bridge operates in the nonproxy mode and in the second the proxy mode

  35. Figure 10.13 LAN emulation: (a) terminology and networking components; (b) unicast protocol architecture; (c) multicast protocol architecture.

  36. 10.5.1 call processingconnectionless calls • Control direct VCC:VCC that connects an LEC to the LES • LE address resolution protocol(LE ARP) forms an address resolution request message and sends this to the LE ARP in the LES • If the LE ARP does not have the ATM address of the destination LEC, the LES sends a copy of the LE ARP request message to all registered LECs using the set of control distribute VCCs

  37. 10.5.1 call processingconnectionless calls • To implement multicasting, a copy of the created MAC frame must be forwarded to all stations that belong to the same multicast group • Multicast send VCC and multicast forward VCC • On receipt of a service primitive with a multicast destination address, the LEP creates a frame and sends this to the LEP over multicast send VCC • On receipt of frame, the LEP broadcasts a copy of the frame to all stations over multicast forward VCC

  38. 10.5.1 call processingconnectionless calls • Echo suppression:The LEP in each station first determines from the LECID at the head of the frame whether it originated the frame. If it did, the LEP discards the frames • If the LECID does not match that of the station, the receiving LEP determines from the multicast address at the head of the frame whether this station is part of the multicast group

  39. 10.5.1 call processing • The ARP in the CLS builds up a routing table containing the address pair of all stations and router ports that are connected to it • Whenever a station IP has a datagram to send, it sends the datagram to the IP layer in the CLS • On receipt of the datagram, the IP in the CLS first reads the destination IP address from the head of the data gram to obtain the ATM address from its routing table • Streaming and pipelining • Figure 10.14

  40. Figure 10.14 Protocol architecture to support classical IP over an ATM LAN.

  41. 10.6 ATM MANs • Distributed queue dual but(DQDB): using the asynchronous transfer mode to implement a MAN • Seamless interconnection • Each DQDB subnetwork is connected to a MAN switching system (MSS) • Figure 10.15 Figure 10.16 • At the head of each bus is a slot generator • In a looped bus architecture, both slot generators are colocated

  42. 10.6 ATM MANs • Besides generating slots, the head of bus is responsible for passing management information to the access nodes • Reconfiguration can be supported in the event of a link or a node failure by replicating the head of bus functions

  43. 10.6.2 protocol architecture • Isochronous: the multiplexed voice samples within the duplex link must be transferred across the bus at the same rate as the access link • The MAC convergence function fragments the data frames submitted into multiple segments and reassembles them back on receipt • Queued arbitrated(QA) and prearbitrated(PA) • The prearbitrated function in each node providing the service is responsible for identifying the reserved slots relating to it

  44. 10.6.2 protocol architecture • Physical layer conversion function establishes a contiguous sequence of slots over the selected physical transmission medium • This has a duration of 125us which ensures that the frame can be used to support various CBR • Figure 10.17

More Related