<f>WAN standards typically describe both physical layer and data-link layer requirements.
|
WAN physical layer protocols describe how to provide electrical, mechanical, operational, and functional connections for wide area networking services.
|
These services are most often obtained from WAN service providers like Regional Bell Operating Companies (RBOCs), alternate carriers, and Public Telephone and Telegraph (PTT).
|
The WAN physical layer describes the interface between the data terminal equipment (DTE) and the data circuit-terminating equipment (DCE).
|
The DCE is the end of the service provider's side of the communication, and the DTE is the user's device.
|
Typically, the services offered to the DTE are made available through a modem or channel service unit/data service unit (CSU/DSU).
@@
WAN data-link protocols describe how frames are carried between systems on a single data-link.
|
They include protocols designed to operate over dedicated point-to-point facilities, multipoint facilities based on dedicated facilities, and multiaccess switched services such as Frame Relay.
|
The common data-link encapsulations associated with synchronous serial lines are
 
 Synchronous Data Link 
  Control (SDLC)
|
 Point-to-Point (PPP)
|
 High-Level Data Link 
  Control (HDLC)
|
 Link Access Protocol, 
  Balanced (LAPB)
|
SDLC is a bit-oriented protocol developed by IBM for its SNA networks.
|
It defines a multipoint WAN environment that allows several stations to connect to a dedicated facility.
|
SDLC defines a primary station and one or more secondary stations.
|
Communications is always between the primary station and one of its secondary stations.
|
Secondary stations cannot communicate with each other directly.
|
SDLC frames contain an address field that always contains the address of the secondary station involved in the current communication.
|
Because the primary station is always either the sender of the receiver, there is no need to include its address in the frame.
@@
PPP was developed by the Internet Engineering Task Force (IETF).
|
It is designed as a protocol to provide data encapsulation and other network management features on serial links between LAN hosts.
|
PPP is described by RFC 1548.
|
PPP, like other serial protocols, was derived from SDLC.
|
Because of this it has inherited some features of SDLC that have no direct application in PPP. 
|
For example, SDLC frames include an address byte that PPP does not need because it does not assign individual station addresses to the end stations. So, PPP simply fills this byte with 1s. 
|
PPP extends the basic SDLC frame by including a protocol field.
|
This is used to identify the network layer protocol encapsulated in the information field of the frame.
|
PPP uses the Link Control Protocol (LCP) to establish, configure, maintain, and terminate the point-to-point connection.
|
LCP goes through several phases of operation:
 
 link establishment and 
  configuration negotiation
|
 link quality determination
|
 network layer protocol 
  configuration negotiation
|
 link termination
|
LCP serves much the same function as the 802.2 LLC in the LAN protocols.
@@
HDLC (High Level Data Link Control) is an ISO standard.
|
It is the default for Cisco routers.
|
However it might not be compatible between vendors because of the way in which individual vendors have chosen to implement it.
|
The Cisco HDLC frame uses a proprietary type field that acts as a protocol field.
|
This makes it possible for multiple network layer protocols to share the same serial link.
|
HDLC supports both point-to-point and multipoint configurations.
|
LAPB (Link Access Procedure, Balanced) is primarily used with X.25 but can also be used as a simple data link transport.
|
Its frame format is the same as SDLC.
|
It includes capabilities for 
 
 detecting out-of-sequence frames
|
 detecting missing frames
|
 exchanging frames
|
 retransmitting frames
|
 acknowledging frames
@@
The X.25 standard addresses the physical, data-link, and network layers in the OSI model.
|
At the physical layer, X.25 provides a synchronous, bit-serial full-duplex point-to-point circuit for data transmission between the DTE and the DCE.
|
At the data-link layer it deals with the detection of transmission errors and their subsequent correction.
|
At the network layer, X.25 defines three basic types of packet service:
 
 a switched virtual circuit
|
 a permanent virtual circuit
|
 a connectionless, datagram 
  service
|
You can think of a virtual circuit as a logical link between two nodes.
|
A switched virtual circuit, means that a logical connection exists between the two nodes only for as long as it takes to transfer the data. Afterwards it is released.
|
A permanent virtual circuit simply means that a permanent logical connection exists between the two nodes.
|
A permanent virtual circut (PVC) eliminates the need for call set-ups as the DTE is always in data transfer mode.
|
The other transmission service supported by X.25 is a connectionless service.
|
Each data unit, called a datagram, is a separate entity which contains data and control information.
|
The network doesn't identify any relationship between two datagrams and therefore each will be routed and processed independently.
@@
Recall that X.25 addresses the bottom three layers of the OSI model.
|
At the network layer data is broken into units called packets and the X.25 protocol header is added.
|
Once the network layer has assembled the data into packets it passes them to the data-link layer.
|
X.25 uses LAP at the data-link layer which defines the movement of data between the DTE and the DCE.
|
When the data-link layer receives the data, it encapsulates it into frames by adding headers and trailers.
|
Finally the data-link layer passes the frames on to the physical layer where they are transmitted on to the transmission medium.
@@
X.25 has some limitations. 
|
In order for X.25 to operate, the packets must be assembled and disassembled in a consistent manner.
|
For a terminal to communicate across an X.25 network, a device called a PAD is used to assemble and disassemble the packets.
|
A PAD sits between the terminal and the network, and all data to be transmitted on to the network will pass through it.
@@
X.25 packet switching is inappropriate for broadband digital voice and video transmissions and bursty data traffic.
|
Frame relay is a service designed to exploit the high quality and high bandwidths offered by optical fiber links.
|
Frame Relay can be implemented without fast packets. It is typically a narrow band service 56Kbps - 2.078Mbps.
|
Because optical fiber has low error rates, sophisticated error checking is unnecessary.
|
Frame Relay frames use a cyclic redundancy check (CRC) to detect errors.
|
Any corrupted frames are simply discarded.
|
The receiving station requests retransmission of any missing frames.
|
Since Frame Relay is simple and involves no complex processing, it can handle streams of bursty traffic at much higher speeds than is possible in most X.25 networks.
|
Fast X.25 @2.078 Mbps does exist in France today.
|
Under Frame Relay, when a network shows signs of becoming congested, a notification is sent to transmitting stations to ease off transmission.
@@
The Transmission Control Protocol/Internet Protocol (TCP/IP) was developed for the US Department of Defense.
|
It is widely used for large internetworks such as the Internet.
|
TCP/IP is a suite of protocols incorporating protocols at the network, transport, and application layers of the OSI model.
|
At the network layer, the TCP/IP protocols include
 
 Internet Protocol (IP)
|
 Internet Control 
  Message Protocol (ICMP)
|
 Address Resolution 
  Protocol (ARP)
|
 Reverse Address 
  Resolution Protocol 
  (RARP)
|
The TCP/IP transport layers are

 Transmission Control Protocol (TCP)
|
 User Datagram Protocol (UDP)
|
The TCP/IP application layer protocols include

 File Transfer Protocol (FTP)
|
 Trivial File Transfer Protocol 
  (TFTP)
|
 Network File System Protocol 
  (NFS)
|
 Simple Mail Transfer Protocol 
  (SMTP)
|
 Telnet
|
 rlogin
|
 Simple Network Management 
  Protocol (SNMP)
|
Although the TCP/IP suite incorporates many more protocols than TCP and IP themselves, these are the two most important.
|
IP provides connectionless, best-effort delivery routing of datagrams.
|
IP datagrams are not guaranteed to arrive at their destination in the sequence in which they were sent, indeed they are not guaranteed to arrive at all.
|
TCP guarantees packet delivery and data integrity.
|
It is a connection-oriented transport layer protocol that provides reliable full-duplex data transmission.
|
The TCP protocol
 
 initiates handshaking
|
 controls packet sequencing
|
 performs flow control
|
 handles errors
@@
NetWare is a popular network operating system developed by Novell.
|
It is based on XNS from Xerox and sits above lower-layer standards such as Ethernet.
|
The protocol that allows it to link into any of several data- link layer protocols is ODI (open data-link interface).
|
ODI is also known as link support layer (LSL).
|
ODI specifications allow NetWare to be independent of the physical layer networking hardware.
|
IPX, the major NetWare protocol at the network layer, provides a very fast connectionless service.
|
IPX also provides addressing, routing, and other network layer services.
|
SPX, the major protocol at the transport layer, is a connection-oriented protocol, providing error checking and flow control.
|
SPX does not correspond precisely to the OSI transport layer.
|
SPX ensures reliable delivery by retransmitting information that has not been correctly received.
@@
The NetWare upper layers correspond roughly to the OSI application, session, and transport layers.
|
At the upper layers, NetWare provides

 NetBIOS emulation
|
 file system support (using NCP)
|
 the NetWare shell
|
 NetWare Remote Procedure Calls (RPC)
|
 NetWare Streams
|
 Message Handling Systems (MHS)
|
 Btrieve
|
NetBIOS was developed by IBM for microcomputer communications.
|
NetWare uses a NetBIOS emulator rather than NetBIOS itself.
|
This emulator allows programs written for NetBIOS to run in the NetWare system.
|
NetWare Core Protocols (NCP) provide the following server functions to users:

 file service
|
 print service
|
 name management
|
 file locking
|
 synchronization
|
 accounting
|
The NetWare shell works in conjunction with NCP to provide transparent access to file and printer services.
|
It intercepts client application requests and processes them.
|
If the request requires network services, it is passed on to the lower layers for transmission across the network.
|
If it does not require network services, the shell sends the request to the local operating system for processing.
|
NetWare RPCs (remote procedure calls) are general-purpose tools providing remote transparent access to other programs.
|
NetWare Streams is a special linking protocol that connects upper-layer processes with the transport layer.
|
It works together with TLI (transport link interface) and LSL (or ODI) to allow NetWare to run over a variety of transport technologies and network drivers.
|
NetWare Streams accepts calls from upper-layer processes and relays them to a driver.
|
MHS (message handling systems) provides the technology for transferring electronic mail across networks, and a mailbox in which to store messages.
|
MHS doesn't provide a user interface.
|
Btrieve provides rapid access to database files and records.
|
Finally, NetWare offers a series of value-added processes, also known as NetWare loadable modules (NLMs) to provide functions such as
 
 file sharing
|
 printer sharing
|
 communication services
|
 database services
@@
Apple Computers Inc. developed the AppleTalk suite of networking protocols for the Apple Macintosh.
|
AppleTalk provides connectivity not only for the Macintosh but also for 

 DOS-based PCs
|
 UNIX-based computers
|
 IBM mainframes
|
 DEC VAX computers
|
At the lower layers, AppleTalk works in conjunction with Apple's own lower-layer protocol, LocalTalk.
|
AppleTalk protocols can also run over standard lower-layer protocols such as Ethernet and Token Ring.
|
The AppleTalk Address Resolution Protocol (AARP) sits above the data-link layer and maps data-link addresses to protocol addresses.
|
This allows AppleTalk protocols to run over any data-link layer protocols.
|
At the network layer, the Datagram Delivery Protocol (DDP) offers a connectionless service.
|
The following protocols work with DDP:

 ZIP (zone information 
  protocol) maps AppleTalk 
  network numbers to zones
|
 RTMP (routing table 
  maintenance protocol) 
  maintains AppleTalk routing 
  tables
|
 NBP (name binding protocol) 
  maps AppleTalk names to 
  addresses
@@
There are two major protocols at the transport layer:
 
 ATP (AppleTalk Transaction 
  Protocol) is a transaction-based 
  protocol that provides reliable 
  delivery of data
|
 ADSP (AppleTalk Data 
  Stream Protocol) provides 
  reliable delivery also, but it is 
  byte-stream-based rather than 
  transaction-based
|
ADSP is a more conventional transport protocol than ATP.
|
PAP (Printer Access Protocol) and ASP (AppleTalk Session Protocol) are both session layer protocols.
|
PAP is designed for connections between workstations and servers of all types.
|
ASP initiates, maintains and terminates sessions.
|
The presentation layer protocol AFP (AppleTalk Filing Protocol) provides access to remote files.
|
AppleShare Print Server, AppleShare File Server and AppleShare PC are all application layer protocols.
|
AppleShare File Server uses the services provided by AFP in order to access files.
|
Together they provide a file access service that is transparent to the user.
|
AppleShare PC allows DOS-based PCs to link into AppleTalk servers.
@@
AppleTalk defines the concept of a network visible entity (NVE).
|
An NVE is a network addressable service, for example a protocol socket.
|
Nodes are not themselves NVEs, but they run network service processes which are NVEs.
|
NVEs have multiple entity names and a series of attributes.
|
An entity name is simply a character string. It might specify a particular socket, for example.
|
An attribute specifies NVE characteristics.
@@
Banyan Virtual Integrated Network Service (VINES) implements a distributed network operating system based on a proprietary protocol family derived from Xerox's XNS.
|
It uses a client/server architecture.
|
The two lower layers of the VINES stack are implemented with a variety of media access mechanisms such as HDLC, Ethernet, and Token Ring.
|
VINES uses VINES Internetwork Protocol (VIP) to perform network layer functions such as addressing and routing.
|
At the network layer VINES also supports
 
 its own Address Resolution 
  Protocol (ARP)
|
 its own version of the Routing 
  Information Protocol (RIP) called 
  Routing Table Protocol (RTP)
|
 the Internet Control Protocol 
  (ICP)
|
VINES network layer addresses are 48-bit entities.
|
This is divided into a network portion, 32-bits, and a subnetwork portion, 16-bits.
|
The network portion is sometimes called the ^Rserver^r portion, and the subnetwork portion is sometimes called the ^Rhost^r portion.
|
VINES provides three transport layer services:
 
 unreliable datagram service
|
 reliable message service
|
 data stream service
|
The unreliable datagram service sends packets that are routed on a best-effort basis.
|
They are not acknowledged at their destination.
|
The reliable message service is a virtual circuit service that provides reliable, sequenced, and acknowledged delivery of messages between network nodes.
|
The data stream service supports the controlled flow of data between two processes.
|
It is an acknowledged virtual circuit service that supports the transmission of messages of unlimited size.
|
VINES uses the Remote Procedure Call (RPC) model for communication between clients and servers.
|
At the application layer, VINES offers file service and print service applications, as well as StreetTalk, which provides a globally consistent name service for an entire internetwork.
@@
DECnet (^RD^rigital ^RE^rquipment ^RC^rorporation ^Rnet^rwork) is described by the DNA (^RD^rigital ^RN^retwork ^RA^rrchitecture).
|
DNA is a seven-layer architecture that is fully OSI-compliant.
|
At the lower layers DNA can run over 

 802.3/Ethernet
|
 FDDI
|
 X.25
|
It also supports the OSI standard protocols HDLC and LAPB, in addition to DEC's own data-link protocol DDCMP.
|
At the network layer, DNA supports a variety of protocols:
 
 OSI connectionless (CLNS, 
  CLNP)
|
 OSI connection-oriented 
  (CONS)
|
 X.25
|
 IP
|
 ES-IS (end-system to 
  intermediate-system) routing 
  protocol
|
 IS-IS (intermediate-system to 
  intermediate-system) routing 
  protocol
|
At the transport layer DNA supports TP0, TP2, TP4, and TCP.
|
DEC's own major protocol at the transport layer is NSP (Network Services Protocol).
|
Like TP4 and TCP, NSP offers a connection-oriented service with flow control.
@@
DNA's session layer, supports both the OSI session layer protocols and DEC's own session control protocol.
|
One of the main functions of DNA's session control protocol is name-to-address mapping.
|
This has to do with enabling upper and lower layers to understand each other.
|
This protocol also controls access to the lower layers and to network resources.
|
DNA applications include presentation features, so DNA does not really have a separate presentation layer.
|
However, it does support the OSI presentation layer.
|
At the application layer, DNA supports a wide variety of applications, including
 
 FTAM
|
 VT
|
 MOTIS/MHS
|
 ACSE
|
 ROSE
|
 CMIP
|
DNA also supports gateways to other protocol stacks, for example SNA.
