CHAPTER 11: WAN Topologies

1
CHAPTER 11 WAN Topologies Transmission Services

• MIS 3523 Business Data Communications
• Fall 2001
• Dr. Segall

2
WAN Topologies

• Star
• Hierarchical
• Interconnected
• Ring
• Bus
• Hybrid (Combinations of above).

3
STAR NETWORK

• The central or hub node serves as a message
  switch by accepting a message from the
  originating node and forwarding it to the
  destination node.
• See Figure 11-1 on page 302 for Star
  Configuration.

4
STAR NETWORK

• Advantages
• 1. Provides a short path between any two nodes
  with a maximum of 2 links to transverse.
• 2. Short time to message to pass from central
  node to a peripheral node.
• 3. Provides user with a high degree of network
  control.
• 4. Expandability of network is simple.

5
STAR NETWORK

• Disadvantages
• 1. Congestion at central site of network.
• 2. Low Reliability
• 3. Higher Circuit Costs.

6
HIERARCHICAL NETWORK

• Network topology in which the nodes are arranged
  hierarchically.
• Also know as atree structure.
• See Figure 11-2 on page 304.
• Advantages
• 1. Lower media costs more likely than for star
  topology.

7
HIERARCHICAL NETWORK

• Disadvantages
• 1. Expansion and reconfiguration costs time and
  money, and may require new network routing
  tables.
• 2. Reliability problem causes failure of entire
  network upon failure of ANY node other than those
  at the extremities.
• 3. Congestion at the root and higher level nodes
  is a potential problem.

8
INTERCONNECTED NETWORK

• Network topology in which any node can be
  connected directly to any other node.
• Also called plex or mesh network.
• See Figure 11-3 on page 305.

9
INTERCONNECTED NETWORK

• Types of Interconnected Networks
• 1. Fully Interconnected Network each node is
  connected to every other node with one link. See
  Figure 11-3(a) on page 305.
• 2. Interconnected Network All of the nodes of
  the network are connected, but it is not possible
  to reach some links with a single link from
  another node. See Figure 11-3(b) on page 305.

10
INTERCONNECTED NETWORK

• Advantages
• 1. Good performance with ability to transmit
  large amounts of data.
• 2. Ability to control costs because of shortest
  path topology between nodes.
• 3. Insertion of new node in Interconnected
  network is simple.
• Disadvantages
• 1. Control is distributed rather than centralized.

11
HYBRID NETWORKS

• Combination of above topologies as integrated
  into one network.
• Example Backbone Network consisting of a ring
  with spurs attached.
• Advantages
• 1. Reliability is high.
• 2. Low cost.
• 3. Helps reduce the number of hops, length of
  links, and congestion problems.

12
NETWORK TOPOLOGY CHARACTERISTICS

• See Summary Table 11-2 on page 306.

13
WAN DATA LINK PROTOCOLS

• Asynchronous(async) most often used to connect
  hosts with terminals.
• Synchronous(sync) used between computers and
  between computers and terminals.

14
ASYNCHRONOUS TRANSMISSION

• Oldest and most common data link protocol.
• Each CHARACTER is transmitted individually with
  its own error detection scheme, usually as a
  parity bit.
• Also called the Start-stop protocol with 0 as
  the Start-bit and 1 as the Stop-Bit. (See Figure
  11-4 on page 307.)

15
ASYNCHRONOUS TRANSMISSION

• Compatibility of sending receiving messages
• 1. Stations must agree on the number of bits per
  character before establishing the communication
  link.
• 2. Stations must agree on transmission speed
  because this determines the interval at which the
  line is sampled.
• 3. Must be agreement as to what will terminate
  the message.
• 4. Must agree on maximum speed of link.

16
ASYNCHRONOUS TRANSMISSION

• Transmitting a Character
• Start Bit
• 7 data bits
• Parity Bit
• Stop Bit
• Interrupt Characters
• a set of characters that terminate a message or
  cause an interruption in transmission to perform
  a special action, such as a backspace.

17
ASYNCHRONOUS TRANSMISSION

• Methods of Message Termination
• 1. Interrupt characters
• 2. Time-out
• If event does not happen in time allowed, the
  time-out expires and process is initiated.
• 3. Character count
• Transmission is complete when a specified number
  of characters have been received.

18
ASYNCHRONOUS TRANSMISSION

• Buffer overflow/overrun
• Buffer is either too small or too full to receive
  the transmitted data. Data is thus lost due to
  lack of storage space.
• Double Buffering
• Used when buffer overflow/overrun occurs to avoid
  losing characters.
• Effectiveness of Asynchronous Transmissions
• See Table 11-2 on page 309.

19
Why is Asynchronous Transmission so popular?

• 1. Oldest protocol so was only available method
  for long time.
• 2. Well developed and well suited to many types
  of applications.
• 3. Allows wide variety of hardware options at low
  prices.
• 4. Data entry at speeds compatible with async
  protocol.
• Primary Drawback
• Inefficient use of the circuit.

20
SYNCHRONOUS TRANSMISSION

• Transmission protocol where sender receiver are
  synchronized.
• Data is transmitted in a block (See Figure
  11-5(a)) at a time. ( Not character (See Figure
  11-5(b)) at a time as in async!!!)
• Types of Synchronous Data Link Protocols
• 1. Character oriented
• 2. Byte count oriented
• 3. Bit oriented.

21
Data Delineation in Synchronous Protocols

• 1. Positional Protocols
• fixed-length fields that indicate message size as
  character size embedded in the message
• 2. Framing Protocols
• uses reserved characters or bit patterns to
  delineate data control fields within the
  message.
• 3. Byte Count Protocols
• includes the number of characters being
  transmitted within the message.

22
Standards for Character Synchronous Protocols

• American National Standards Institute (ANSI)
• IBM Binary Synchronous Communications (BISYNC or
  BSC)
• Introduced by IBM in 1967
• Only 2 common BISYNC data codes
• ASCII
• EBCDIC
• Control Characters include
• SYN (synchronization)
• STX (start of text)
• ETX (end of text)

23
Standards for Character Synchronous Protocols

• Transparency
• ability to send any bit string as data in the
  message
• with BISYNC involves the insertion of extra
  characters in the message
• Limitations of BISYNC
• Essentially a half-duplex protocol
• Effectiveness of BISYNC Protocol
• Summarized in Table 11-3 on page 313.

24
Byte Count Synchronous Protocols

• Difference between BYTE count synch protocols and
  BISYNC is their signaling of the beginning and
  end of messages.
• Header is a fixed length and data field is of
  variable length.
• Advantage of byte count protocols is their
  transparency.

25
Byte Count Synchronous Protocols

• Some implementations of byte count protocols also
  include message sequence numbers.
• Performance
• much the same as BISYN except for
• 1. transparency is present in Byte count synch
  protocol and requires insertion of extra
  characters in BISYN
• 2. efficiency is greater for Byte count synch
  protocols if message sequencing or true
  full-duplex operations are allowed.

26
Bit Synchronous Protocols

• Use bits rather than bytes to delineate data and
  provide message control
• 1st bit-oriented synchronous data link protocol
  was introduced by IBM in 1972.
• Current major bit synchronous protocols are
• SDLC (Synchronous Data Link Control)
• ADCCP (Advanced Data Communications Control
  Procedure)
• HDLC (High-Level Data Link Protocol)
• LAPB (Link Access Procedure, Balanced)

27
Bit Synchronous Protocols

• SDLC (Synchronous Data Link Control)
• Developed by IBM and operates in full-duplex or
  half-duplex mode in both point-to-point and
  multipoint configurations.
• ADCCP (Advanced Data Communications Control
  Procedure)
• An ANSI standard bit-oriented data link control
• HDLC (High-Level Data Link Protocol)
• operates in full-duplex or half-duplex mode in
  both point-to-point and multipoint
  configurations.
• ISO standard
• LAPB (Link Access Procedure, Balanced)
• protocol specified for X.25 networks

28
Synchronous Data Link Control

• Possible to configure stations in a loop (See
  Figure 11-10).
• Data are transmitted in one direction around the
  loop.
• One station is designated as primary station, and
  others as secondary stations.
• SDLC Frame Format (See Figure 11-11)
• Flag (8 bits)
• Address (8 bits)
• Control (8 bits)
• Data (variable length)
• CRC (16 bits) frame check sequence
• Flag (8 bits)

29
Synchronous Data Link Control

• Frame transmission packet in bit-oriented
  protocols
• SDLC is a positional protocol, which means that
  each field except the data field has a specific
  length and location relative to adjacent fields.
  (quote from page 316!!)
• Frame Types
• 1. information
• 2. supervisory
• 3. unnumbered

30
Synchronous Data Link Control

• Ns (Number sent) subfield
• field on the transmission frame on the SENDERs
  system used to represent the frame sequence
  number being transmitted.
• Nr (Number received) subfield
• field on the transmission frame on the
  RECEIVERs system used to represent the frame
  sequence number the RECEIVING station expects to
  receive next.
• See Figure 11-12 on page 317 for control fields
  for Information frames (I-frames)
• P/F is Poll/Final bit in Figure 11-12.

31
Synchronous Data Link Control

• Transparency is implemented in SDLC by bit
  insertion also known as bit stuffing.
• Effectiveness of SDLC Bit Synchronous Protocol is
  summarized in Table 11-4 on page 318.
• Supervisory frame control functions
• Receive Ready (RR)
• Receive Not Ready (RNR)
• Frame Reject (REJ)

32
Comparison of Asynchronous Synchronous Protocols

• Asynchronous
• 1. Character-at-a-time transmission
• 2. Modems are NOT synchronized
• 3. Error detection commonly is PARITY.
• 4. Fixed overhead per CHARACTER.
• 5. LESS efficient use of communication link.
• 6. LOWER cost devices.

• Synchronous
• 1. BLOCK transmission.
• 2. Modems ARE synchronized.
• 3. Error detection is commonly CRC or PARITY plus
  LRC.
• 4. Fixed overhead per BLOCK.
• 5. MORE efficient use of communication link.
• 6. HIGHER cost devices.

33
THE OSI NETWORK LAYER

• Performs 4 major functions
• 1. Routing
• 2. Network Control
• 3. Congestion Control
• 4. Collection of Accounting Data
• Network Control
• Involves sending receiving of node status
  information to other nodes to determine the best
  routing for messages.

34
Message Routing

• Routing in LANs using Routing Tables
• Network Routing Manager
• A designated node that has an overview of network
  functioning, location of any bottlenecks, and
  location of underused facilities.
• Periodically recalculates the optimal paths
  between nodes and constructs and distributes new
  routing tables to all nodes.

35
Disadvantages of Network Routing Manager

• 1. Routing managers ability to receive many
  messages from the other nodes increases the
  probability of congestion.
• 2. Some nodes may receive newly calculated
  routing tables before others leading to
  inconsistencies in how messages are to be routed.
• 3. Transmission of the routing tables may bias
  the statistics being gathered to compute the next
  routing algorithm.
• 4. Reliability of routing manager affects
  selection of alternative routing manager(s).

36
Routing

• Distributed Routing Algorithm
• Each node calculates its own routing table based
  on status information periodically received from
  other nodes.
• Static Routing
• One particular path between 2 nodes is ALWAYS
  used.
• Adaptive Routing
• Evaluates the existing paths and chooses the one
  that will provide the best path for a message.

37
MESSAGE ROUTING ALGORITHMS

• 1. Quickest Link shortest route
• 2. Best Route (e.g. no. of required hops, speed
  of links, type of link, congestion, etc.)
• 3. Broadcast message is broadcast to ALL
  stations, but ONLY the station to which the
  station is addressed accepts it.
• 4. Weighted weights are assigned to paths
  according to perceived use.

38
FUNCTIONS OF OSI TRANSPORT LAYER

• End-to-end reliability
• Packet Sequencing
• Flow Control
• Error Control
• Addressing
• Security
• Message Segmentation
• Connection Management

39
WAN Topologies Transmission Services

• THE END !!!