Computer network

architectures and protocols

theory

Imre Varga
University of Debrecen, Faculty of Informatics

For internal use only!

04 September 2017
 General information
Subject, course:
Computer network architectures and protocols
INHK721 (Computer Science Engineering BSc)
Computer networks (Architectures and protocols)
INJK711-K5 (Business Information BSc)
Wednesday 10:00-12:00, TEOKJ II/106B room (Lecture, I.V.)
Tuesday 12:00-14:00, IF03 room (Practice, INJK711L, I.V.)
Tuesday 16:00-18:00, IF03 room (Practice, INHK721L, A.K.)

subject credit: 5 (INJK711-K5 ), 4 (INHK721)

2
 General information
Teachers:
Dr. Varga, Imre (Tuesday 10, Tuesday 12)
Department of Informatics Systems and Networks
email: [email protected]
www: irh.inf.unideb.hu/user/vargai
room: IF13

Karsai, Andrea (Tuesday 16)

Network and Telecommunication Supplier Unit
email: [email protected]
room: Chemistry Building, C stairway 4/3
3
 General information
Requirements, conditions for practice (INJK711L):
maximum number of absences: 3
late arrival (more than 20 minutes) means absent from
class
2 midterm tests (+1 retake)
to pass a test: reach at least 50%
if a test failed: retake is necessary with extra conditions
retake test: covers the whole semester
result overwrites the worse test result

4
 General information
Requirements, conditions for lecture (INJK711-K5):
written exam
to pass: reach at least 50%
signature + passed theoretical test: suggested grade
theoretical and practical results together determine
the final grade (50%-50%)

Readings:
Andrew S. Tanenbaum: Computer Networks, Prentice-Hall, 2003.

5
 Topics
• Concepts of network
• Layered network architecture
• Protocols and services
• Transmission mediums
• Ethernet
• IP addressing
• Routing
• Network configuration
• Applications (DNS, web, e-mail, ftp, …)
• Many more things…

6
Basics of computer networks

7
 Computer Networks
Definition:
Two or more computers linked together with some
software and hardware tools for an information
transmission related purpose.
Purposes:
• Human communication.
• Sharing resources.
• Increasing reliability.
• Increasing speed.
• etc.
8
 Computer Network Nodes
Node:
Device with own network address. It can communicate
independently (e.g. computer, printer, router).
In a communication a node can act either as a
transmitter (source) or as a receiver (sink).
Categories of network devices and tools:
• End user node: computer, printer, scanner, and any
other devices that provide services directly to the user
• Network linking/connecting tools: devices that enable
communication between end user nodes by connecting
them to each other

9
 Classification of Computer Networks
Based on their sizes:
• Personal Area Network (PAN)
• Local Area Networks (LAN)
• Metropolitan Area Networks (MAN)
• Wide Area Networks (WAN)
Based on switching technology:
• Packet switching
• Circuit switching
• Message switching
10
 Classification of Computer Networks
Personal Area Network Local Area Network
• For one person • For a building
• Size: few meters • Size: max few 100 m
• E.g.: USB, Bluetooth • Ethernet, Wi-Fi

Metropolitan Area Network Wide Area Network

• Covers a city • Covers countries and
• Size: few 10 km more
• Connect LANs • Size: more 100 km

11
 Classification of Computer Networks
Circuit switching
• Establish a dedicated communications channel
(circuit) before the nodes may communicate, they
remains connected during communication
Message switching
• The whole information (message) travels from node
to node (store-and-forward technique)
Packet switching
• Message is cut into smaller units (packet) which are
transmitted independently
12
 Packet switching
Advantages
• Don’t need large memory/disk in routers (cheaper)
• No continually busy lines for long time (interactive)
• While 2nd packet is arriving 1st can be sent (faster)
• Fault tolerant (re-routing, partial retransmission)
• Efficient (not occupied line, if no transmission)
• Charging/fees are based on the amount of sent
information (not the time of connection)
Internet is (mostly) packet switched.

13
 Transmission Speed
Transmission speed
(network speed, bandwidth, bit rate):
Amount of information transmitted during a time
unit. Measure of unit: bit/sec, b/s, bps.
The throughput measured in applications is always
lower than the physical bandwidth.

Larger units:
• 1 kbps = 1000 bps
• 1 Mbps = 1000 Kbps
• 1 Gbps = 1000 Mbps
14
Directions of Information Transmission
One way (simplex) connection:
The transmission of information allowed only one
way is called a one way (simplex) connection
(eg. radio broadcasting).
Alternate way (half duplex) connection:
The transmission allowed both directions, but only
one direction at a time is called a half duplex
connection (eg. CB radio).
Two way (full duplex) connection:
The traffic allowed in both directions simultaneously
is called a full duplex connection (eg. telephone).
15
 Connections of Data Transmission
Point-to-point connection:
The propagation of information performed between
two points (a transmitter and a receiver) is called a
peer-to-peer connection.
Multiple nodes connection, broadcasting:
A transmitter provided information to multiple
receivers is called a multiple nodes connection.
Broadcasting is a multiple nodes connection, where
all receivers get the information inside a given range
(e.g. radio broadcasting).
16
 Basics of Addressing
Unique address (Unicast):
An identifier, assigned to a Imre Varga
Debrecen
Kassai 26
network interface of a node. 4028

Everyone address (Broadcast):

An address, identifying all
nodes (and interfaces of nodes)
in a so called broadcast domain.
Not a list of unique addresses.

17
 Computer Network Protocol
Protocol:
The formal description of all rules and conventions
which determines the communication of network
devices (nodes) (set of communication rules).
Syntax, semantics, timing, etc.

Examples:
HTTP, FTP, IP, DHCP, TCP, UDP, SMTP, POP3, IMAP,
ARP, RARP, ICMP, RIP, EIGRP, OSPF, IPSEC, …

18
 Server-Client Architecture
Server:
A network node (and software) which provides
services for other nodes. The service of a server is
ensured by a server-software (e.g. a web-server).
Client:
A network node (and software) which has some kind
of network service demand. For recourse to the
service the client uses a client-software (e.g. web
browser).
The communication between the server and the client
is described by a high level protocol.
client2
request
server
client1
reply 19
 Peer-to-peer architecture
• No fixed client/serves roles
– Equivalent hosts
• Anyone in the group can communicate with anyone
else directly
– No client server client host1
• Example: Napster, BitTorrent, Skype
host2

host3

host5
host4

20
Transmission Media, Channel, Collision
Transmission media:
Device or material on which the transmission of
information (signal) is performed. (Eg. twisted-pair cable,
coaxial cable, fiber-optic cable, or air).
Transmission channel:
Data path, frequency band for transmitting signals.
Usually, in a transmission media multiple channels (data
path) are formed.
Collision:
A collision occurs when two (or more) nodes transmit
information at the same time on a common transmission
channel.

21
Layered Network Architecture

22
 Layered Network Architecture
Why we use layered network architecture?
• To describe a huge protocol is complex and difficult.
• A hierarchical protocol system can be easier
implemented.
• The change tracking is easier.
• Layers can cooperate also in case of different
producers.

23
Philosopher-translator-secretary architecture
The life Das Leben
is… Philosophers ist …

Lang: hu Lang: hu

Az élet… Translators Az élet…

Fax: +36… Fax: +36…

Lang: hu Lang: hu
Secretaries
Az élet… Az élet…

24
Layers (Levels), Protocols, Interfaces
Machine 1 Machine 2
Layer 5 Layer 5 protocol Layer 5
Layer 4/5 interface
Layer 4 Layer 4 protocol Layer 4
Layer 3/4 interface
Layer 3 Layer 3 protocol Layer 3
Layer 2/3 interface
Layer 2 Layer 2 protocol Layer 2
Layer 1/2 interface
Layer 1 Layer 1 protocol Layer 1

Transmission medium
25
 Concepts of Layered Architecture
Layer N protocol:
A protocol which describes the specifications of layer N.
Peers:
Entities which located on the same level of the two
communication endpoints (nodes). In some logical way
the peers communicate each other by the help of the
corresponding layer protocol.
Layer N/N+1 interface:
Connection of boundary surface of layers N and N+1.
Service of Layer N:
Set of actions (service) which are provided to layer N+1
by layer N (through the interface).

26
 Encapsulation
Encapsulation:
Packaging the information arrived from a higher level
with a header of a specific protocol (it is similar when
a traditional mail letter is put in an envelope and the
envelope is addressed).
H2 H3 H4 L5 DATA T2

Protocol Data Unit (PDU):

Entity (contains header and data) handled by the
considered protocol. (It is frequently mentioned as
packet.)
27
 Scheme of Network Communication
Layer
Data L5 protocol Data
5

L4 protocol
4 H4 Data H4 Data

H3 H4 D… H3 …ata L3 protocol H3 H4 D… H3 …ata

3

2 L2 prot.
H2 H3 H4 D… T2 H2 H3 …ata T2 H2 H3 H4 D… T2 H2 H3 …ata T2

28
 OSI Reference Model
Layer Name of PDU

7 Application Layer APDU

6 Presentation Layer PPDU

5 Session Layer SPDU

4 Transport Layer TPDU, Segment

3 Network Layer Packet

2 Data Link Layer Frame

1 Physical Layer Bit

29
 Layers of OSI model
Physical Layer (L1):
Specification and properties of different transmission
mediums in order to implement signal transmission.
• Cables, connectors, modulation, signal coding, etc.

Data Link Layer (L2):

Reliable transmission between two directly
connected devices. Two sublayers: LLC, MAC.
• Physical addressing, media access, logical topology,
acknowledging, etc.

30
 Layers of OSI model
Network Layer (L3):
Connection between any two network nodes (not
just dirrectly connected).
• Routing, traffic control, network addressing, etc.
Transport Layer (L4):
Reliable connection between softwares on two
nodes. Protocols may connectionless or connection-
oriented.
• Data stream, error detection/correction, order
guarantee, etc.

31
 Layers of OSI model
Session Layer (L5):
Relationship-treating between applications during
the dialog, establishing sessions between hosts.
Presentation Layer (L6):
Provides same interpretation of information
(different nodes can use different data structures,
data representation). Encryption, compression, etc.
Application Layer (L7):
Interface between applications and users.
• DNS, web, e-mail, ftp, bittorrent, etc.
32
 Mapping of TCP/IP - OSI Model
Layer OSI model TCP/IP model
7 Application Layer

6 Presentation Layer Application Layer

5 Session Layer

4 Transport Layer Transport Layer

3 Network Layer Network Layer

2 Data Link Layer Network Interface

(host-to-network)
1 Physical Layer Layer

33
Hybrid Reference Model
Layer

5 Application Layer

TCP/IP like
4 Transport Layer

3 Network Layer

2 Data Link Layer

OSI like
1 Physical Layer

34
Network interconnection

35
 Network Interconnection - Basics
Collision domain; Bandwidth domain:
Part of a network, where collisions can be detected
(a common communication channel that is shared by
multiple nodes).
In a collision domain only one information
transmission can be performed at a time.
Broadcast domain:
Part of a network, where information transmitted
with a broadcast address can be detected.

36
 Interconnected networks
Problems with interconnected networks
• Too large distances between nodes
• Too large collision domain: low efficiency, frequent
collisions
• Too large broadcast domain: congestion, too much
packets
• Connected networks can have different
– cabling – packet size
– signals – address space
– speed – protocols
37
 Network Interconnection Devices
Repeater:
Amplifies and repeats the signals sent on transmission
media.
Does not separate the connected subnetworks.
Repeaters with multiple ports is called a HUB.
Bridge:
Working in Data Link Layer it performs selective
connection („Only those packets goes through the
bridge, who tends to other side”).
The interconnected subnets form separate collision
domains.
Usually transmits the broadcasting towards all
interconnected subnets.
38
 Network Interconnection Devices
Switch:
A multiple port device with bridge functionality between
any two ports.
Router:
Working in Network Layer it performs selective
connection, routing, and traffic control.
The interconnected subnets form separate collision
domains and separate broadcast domains.
It is a node with own IP address.
It is also called a gateway in Network Layer (default
gateway).

39
 Network Interconnection Devices
• Which node-pairs don’t disturb each other?
• Who is available form where by broadcast?

N1 N2 N3 N10 N11 N12 N13 N14 N15

Bridge Repeater

N4 N5 N6
N16

N17
Hub Router Switch
N18
N7 N8 N9
N19
40
 Network Interconnection Devices
• Which node-pairs don’t disturb each other?
• Collision domains:

N1 N2 N3 N10 N11 N12 N13 N14 N15

Bridge Repeater

N4 N5 N6
N16

N17
Hub Router Switch
N18
N7 N8 N9
N19
41
 Network Interconnection Devices
• Who is available form where by broadcast?
• Broadcast domains:

N1 N2 N3 N10 N11 N12 N13 N14 N15

Bridge Repeater

N4 N5 N6
N16

N17
Hub Router Switch
N18
N7 N8 N9
N19
42
 Network Interconnection Devices
Subnetworks – based on the functionality of network
interconnection devices – can be connected in different
OSI Layers.
OSI layer Connector item

Transport Layer gateway

and above

Network Layer router

Data Link Layer bridge, switch

Physical Layer repeater, hub

43
 Repeater, switch, router
Machine 1 Machine 2
Layer 5 Layer 5

Layer 4 Layer 4

Layer 3 Router Layer 3

Layer 2 Switch Layer 2

Layer 1 Repeater Layer 1

44
Physical layer

45
 Physical layer
First layer of hybrid model (L1)
Specification and properties of different transmission
mediums in order to implement signal transmission.
Topics
• Cables and connectors
• Topology
• Modulation and signal coding,
• etc.

46
Theoretical basis of communication
Fourier analysis
1
= + sin 2 + b cos 2
2

Frequency band limits the bitrate

• Larger signal frequency range
• Larger bandwidth
• More detailed signal
• More coded information
• Faster transmission speed
• (More noise in the channel decrease the bitrate)
47
 Attenuation
The amplitude of a signal is decreasing during its way in
a transmission media.
The length of a transmission media is determined such
a way that the signal should be interpreted securely
by the receiver.
If large distance has to be covered, the signal has to be
restored by the help of amplifiers (repeaters).
The attenuation depends on frequency (bandwidth
important), thus the amplifiers have to compensate
this with frequency dependent amplification.
48
 Physical transmission and cables
Wired
• Coaxial cable (electric signal)
– Thin, thick
• Twisted pair (electric signal)
– UTP, FTP, STP
• Optical fibre (light)
– Multimode, single mode
Wireless
• Air (electromagnetic waves)
– Radio wave, microwave, infrared
49
 Coaxial cable

5-25mm

Outer cover Outer conductor Insulation Inner conductor

50
 Coaxial cable
Transmission characteristics:
• Due to the concentric structure of conductors, it is
not sensitive for interference and crosstalk
• In case of digital transmission amplifiers are
required in every km
• In case of analog transmission, amplifiers are
required in every several km
Applications:
• Transmission of television broadcasting
• Large distant telephone transmission
• Connection of computers
51
 Unshielded Twisted Pair
schematic:

5.1mm

Outer cover Twisted pairs Colored insulation Copper wire

52
 Unshielded Twisted Pair
Characteristics:
• It is the cheapest media
• Data transmission speed (100Mbps) and the distance
(100m) to be covered are highly limited
• Two isolated copper conductors are twisted and four
such pairs are grouped without shield (UTP)
• Foiled Twisted Pair (FTP): pairs has a common shield
cover
• Shielded Twisted Pair (STP): pairs are shielded
separately
53
 Common twisted-pair cables
Typical
Name Bandwidth Application
construction
CAT. 1 UTP 0.4 MHz phone
CAT. 3 UTP 16 MHz 10Base-T
CAT. 5 UTP 100 MHz 100Base-T
CAT. 5e UTP 100 MHz 1GBase-T
CAT. 6 UTP 250 MHz 10GBase-T
CAT. 7 FTP / STP 600 MHz 10GBase-T
Terminated in 8P8C (RJ45) connector
Maximum length: 100 m (90+4+6 or 90+10)
54
 Optical fiber
Protecting cover
Reflecting light ray

Core 2-125µm

Cladding

a) Refraction b) Total reflection

α n1 α α

β n2

n2< n1
55
 Optical fiber
Characteristics:
• Core and cladding: glass or plastic with different
reflective index
• Works in 1014 - 1015 Hz (infrared) domain
• 3 versions are used: multi mode, single mode, multi
mode graded index
• Light sources: LED, laser diode
• Connectors: ST, SC, FC, MT-RJ, LC, MU, MDI, …

56
 Types of optical fibers
Multi-mode: incoming outgoing
signal signal

50µm

Graded index:

Single-mode:

9µm

57
 Optical fiber
Advantages:
• Larger capacity: High transmission speed can be
achieved (2 Gbps in 10x km).
• Smaller size and weight
• Smaller attenuation: The attenuation is smaller, and
it is constant at a wide frequency range.
• Electromagnetic isolation: Not sensitive for outer
electromagnetic effects, there is no crosstalk.
• Larger repeating distance: Smaller the number of
repeaters is, smaller the possibilities of errors and
the costs are.
58
 Signal, Signal Coding, Modulation
Signal: Physical quantities, depending on place and time,
and carrying information. Information carrier on the
communication channel, it could be analog or digital.
Signal Coding: Mapping the (digital) information onto the
digital carrier signal (e.g. voltage levels, changing of
voltage levels). It is also called line coding.
Modulation: Mapping onto analog carrier signal. The
process of creating the (modulated) signal to be
transmitted through the channel from the modulating
signal coming from the source and the analog carrier
signal. Inverse process is the demodulation. A modem
performs modulation and demodulation, as well.
59
 NRZ signal coding
Non Return to Zero
• ‚0’ bit represented by one signal level (-1)
• ‚1’ bit represented by an other signal level (+1)
• Easy implementation
• No synchronization in case of (several) same bits

bits 1 0 1 1 0 0 0 1 0 1

NRZ

60
 RZ signal coding
Return to Zero
• ‚0’ bit represented by one signal level (-1)
• ‚1’ bit represented by half bit-time (+1) and
half bit-time (+1)
• Double frequency needed
• No synchronization in case of several ‚0’ bits
bits 1 0 1 1 0 0 0 1 0 1

RZ

61
 NRZI signal coding
Non Return to Zero Inverted
• ‚0’ bit represented by keeping the previous signal level
• ‚1’ bit represented by changing the previous signal
level
• No synchronization in case of several ‚0’ bits

bits 1 0 1 1 0 0 0 1 0 1

NRZI

62
 Manchester signal coding
Also called Phase Encoding (PE)
• ‚0’ bit represented by high-low level change at the
middle of bit-time
• ‚1’ bit represented by low-high signal sequence
• Double frequency needed
• Synchronized
bits 1 0 1 1 0 0 0 1 0 1

PE

63
 DM signal coding
Differential Manchester
• ‚0’ bit represented by the same level change at the
middle of bit-time as previously
• ‚1’ bit represented by opposite level change at the
middle of bit-time as previously
• Double frequency needed, synchronized
bits 1 0 1 1 0 0 0 1 0 1
Differential
Manchester

64
 BMC signal coding
Biphase Mark Coding
• ‚0’ bit represented by changing level for full bit-time
• ‚1’ bit represented by changing level for half bit-time,
then changing level again for half bit-time
• Double frequency needed
• Synchronized
bits 1 0 1 1 0 0 0 1 0 1

BMC

65
 4B5B coding
4bit-5bit coding
• Maps group of 4 bits onto group of 5 bits
• Uses a conversion table
• Max 3 „0” bits are next to each other
• There are special and unused 5-bit groups
• Further coded by e.g. NRZI
bits 0000 0110 1100 0001
4B5B 11110 01110 11010 01001

66
 Modulation
Digital signal

Carrier signal

Amplitude
modulation (AM)
Frequency
modulation (FM)
Phase
modulation (PM)
67
 Bases of Wireless Communication
• Wave: A kind of changing, which results point-to-
point (cyclic) energy transfer.
• Amplitude: Distance between the zero and the
maximum signal height.
• Frequency (f): Number of cycles in one second.
• Time of period (T): The time of one cycle T = 1/f.
• Wave length (λ): Distance between two identical
signal height values.
• Speed of light (c): velocity of an electromagnetic ray
C=fλ
68
 Bases of Wireless Communication
x(t) = A sin (2πf t + ϕ) λ

Example (Wi-Fi):
f = 2.4 GHz = 2.4*109 Hz
λ = 125 mm = 0.125*10-3 m
T = 41.7 ns = 4,17*10-10 s
c = 300 000 km/s = 1.08*109 km/h
69
 Wireless transmission
Propagation and detection of electromagnetic signals
are performed by antennas.
The two ways of transmitting:
• Directed: focused electromagnetic ray. Antennas
should be positioned very precisely.
• Omnidirectional (not directed) : radiation can be
received with multiple antennas
Three frequency ranges for wireless transmission:
• 2 - 40 GHz (microwave transmission) (directed)
• 30 MHz - 1 GHz (radio frequency) (omnidirectional)
• 3 1011 - 2 1014 Hz (infrared)
70
 Communication satellites
Relaying/forwarding either in space or on the ground
• Geostationary satellites (Arthur C. Clarke)
– Altitude 35800km
• Medium-Earth orbit satellites
– Altitude 5000-20000km
– Global Positioning System (GPS)
• Low-Earth orbit satellites
– Altitude 150-2000km
– Iridium, Globalstar, Teledesic

71
 Topologies
Physical topology:
Investigates the placement of nodes and their
connection possibilities. (Cable topologies).
Logical topology:
Investigates the logical sequence and order of nodes.
Topologies:
• Bus
• Ring
• Mesh
• Star
72
 Bus topology

node node node node

drop drop drop drop

line line line line

cable tap tap tap tap cable

end end
backbone

73
 Bus topology
One long cable acts as a backbone to link all the devices
in the network. Nodes are connected to the common
main cable by drop lines and taps.

Advantage:
• Easy installation
• Simple and cheap
Disadvantage:
• Difficult fault isolation
• Bandwidth is shared on all links

74
 Ring topology
node

node node
peer-to-peer lines

node

75
 Ring topology
Each device has a dedicated point-to-point line
connected only to the two devices on both sides.

Advantage:
• Easy installation
• Fault isolation is simplified
Disadvantage:
• Changing a devices can affect the network
• Bandwidth is shared on all links

76
 Mesh topology
node

node
node

node node
peer-to-peer lines

77
 Mesh topology
Every device has a dedicated point-to-point link to
(almost) every other device.

Advantage:
• Mesh topology is robust
• Lines are not shared (in most of cases)
Disadvantage:
• A fully connected mesh network therefore has
N(N-1)/2 physical channels to link N devices

78
 Star topology
node

node
node

central
controller

peer-to-peer lines
node node

79
 Star topology
Each device has a dedicated point-to-point link only to
a central controller (usually a switch).

Advantage:
• if a link fails, only that link is affected
• Lines are not shared
Disadvantage:
• Failure of the central hub renders the network
unserviceable

80
 Extended star (tree) topology
node node
node

central
central
controller
controller

central
node controller
node

node node

peer-to-peer lines
81
Data link layer

82
 Data link layer
Second layer of hybrid model (L2)
Reliable transmission between two directly connected
devices. Two sublayers: LLC, MAC.
Topics
• Physical addressing (identification)
• Media access
• Logical topology
• etc.

83
 Data link Layer
Network layer Higher levels

Logical Link Control

IEEE 802.2
sublayer
Data link layer
IEEE IEEE IEEE IEEE Media Access Control
802.3 802.4 802.5 802.11 sublayer

Physical layer Transmission medium

IEEE 802.2 = Logical Link Control (LLC) protocol

IEEE 802.3 = CSMA/CD
IEEE 802.4 = Token bus Media Access Control (MAC)
IEEE 802.5 = Token ring protocols
IEEE 802.11 = Wi-Fi
84
 Framing
Breaking the bit stream into frames
• Character count
5 5 1 3 6 4 0 2 3 7 4 3 5 9 6 4 4 6 9 1 without error
5 5 1 3 6 6 0 2 3 7 4 3 5 9 6 4 4 6 9 1 with error
• Byte/character stuffing (starting/end flag byte)
F t o t a l f r a m e c o n t e n t F
F T o t a l @ F r a m e C o n t e n t F
F B i g . @ F r a n k @ @ g m a i l . c o m F
• Bit stuffing
Similar to byte stuffing just special bit pattern is stuffed.
85
 Error detection and correction
• Checksum: A special value calculated based on the
frame content and attached to the end of the frame
before sending
• After receiving the frame it is calculated again and
compared by the original/attached value
• If the two values are different, an error occurred
during transmission.
• Examples
– Parity bit
– Cyclic Redundancy Check (CRC)
86
 Data-link technologies
WAN LAN (MAN)
• SLIP • Aloha
• PPP • Ethernet
• ATM • Wi-Fi
• X.25 • Token-ring
• Frame-relay • Token-bus
• ISDN • FDDI
• ADSL • DQDB

87
 Media access
Static (suitable, if node number is small and constant)
• Time-Division Multiplexing (TDM)
• Frequency-Division Multiplexing (FDM)
Dynamic (suitable, if node number is large or changing)
• No carrier sense
• Time-slotted
• Token
• Carrier Sense Multiple Access (CSMA)
• Collision Detection (CD) / Collision Avoidance (CA)
• Code Division Multiple Access (CDMA)
88
 TDM
• Each source can send periodically only in a given
time interval
• Low speed sources, high speed channel

source 1

source 2

source 3

source 4

1 2 3 4 5 6 7 time

89
 FDM
• Each source use a separate (not overlapping)
frequency sub-band to modulate signals
• Example: radio and TV broadcasting
• In case of optical signals it is also referred
Wave-length Division Multiplexing (WDM)
frequency

source 1

source 2

source 3

1 2 3 4 5 6 7 time
90
 Pure ALOHA
• Wireless communication between Hawaiian islands
• Anybody can send a frame anytime
• Many collisions occur (max efficiency 18.4%)

source 1

source 2
transmitted
source 3

source 4 collided

1 2 3 4 5 6 time
91
 Slotted ALOHA
• Time slot is applied
• A source can start sending only at beginning of slot
• Many collisions occur (max efficiency 36.8%)

source 1

source 2
transmitted
source 3

source 4 collided

1 2 3 4 5 6 time
92
 Token-ring
• Logical topology: ring (physical topology: star)
• Special frame (token) always orbits in the network
• Device can transmit only if it has control of the token
• Who has the token send a data-frame, it is forwarded
from node to node
• Destination gets data-frame and forward an
acknowledgement
• Source gets acknowledgement, removes the frame,
passes the token to the next node
• No collision occurs
93
 Carrier sense multiple access
The same channel is used by several nodes
If the channel is busy no one else starts transmission
• 1 persistent CSMA: if the channel become idle/free,
waiting/ready node starts sending immediately
• p-persistent CSMA: if the channel become idle, ready
node starts sending with p probability or waits the
next time slot with 1-p probability
• non-persistent CSMA: if the channel is in use, node
wait (immediately, before channel become idle) a
random time period
94
 Ethernet
The most popular technology for wired LANs based on
Carrier Sense Multiple Access with Collision Detection
(CSMA/CD) media access method.

version standard year speed

‘Classical’ Ethernet IEEE 802.3 1980 10 Mbps
Fast Ethernet IEEE 802.3u 1995 100 Mbps
Gigabit Ethernet IEEE 802.3ab 1999 1.000 Mbps
10Gigabit Ethernet IEEE 802.3ae 2002 10.000 Mbps
100Gigabit Ethernet IEEE 802.3ba 2010 100.000 Mbps
95
 Ethernet cabling
Name Cable Max. segment Speed
10Base5 Thick coax 100 m 10 Mbps

Ethernet
Classical
10Base2 Thin coax 185 m 10 Mbps
10Base-T Twisted pair 100 m 10 Mbps
10Base-F Fiber optics 2000 m 10 Mbps
100Base-T4 Twisted pair 100 m 100 Mbps

Ethernet
Fast
100Base-TX Twisted pair 100 m 100 Mbps
100Base-FX Fiber optics 2000 m 100 Mbps
1000Base-SX Fiber optics 550 m 1000 Mbps

Ethernet
Gigabit
1000Base-LX Fiber optics 5000 m 1000 Mbps
1000Base-CX 2 pairs of STP 25 m 1000 Mbps
1000Base-T 4 pairs of UTP 100 m 1000 Mbps

96
 Ethernet frame format
Preamble 7 bytes (7 x ‘10101010’ for synchronization)
Direction of
transmission Start of Frame 1 byte (‘10101011’)

Destination address 6 bytes

Header

Source address 6 bytes

Length/Type 2 bytes (IEEE/Ethernet)
Data

Data/Payload 0 - 1500 bytes

Pad (if necessary) 0 - 46 bytes (minimum frame length is 64 bytes)

Tail

CRC checksum 4 bytes

97
 Ethernet (MAC) address
6 bytes wide identifier of network cards written in
hexadecimal number system separated per bytes.
Example: 00-26-9E-93-75-AA

UOI: ID of serial
manufacturer number
(3 bytes) (3 bytes)
MAC address is ‘burned’ into the
network interface.
There are no network cards in the
world with same MAC address.
Broadcast address: FF-FF-FF-FF-FF-FF
98
Ethernet frame transmission (CSMA/CD)
Waiting for data to transmit.
Formatting the frame.

Is the channel busy?

Yes No
Waiting the gap time.
Starting transmission.

No Yes
Is there a collision?

Complete transmission. Sending disturbing signals.

Increasing the number of trials.

Indicating unsuccessful Yes Max. number of trials

transmission to higher level. is reached (16)?
No
Computing the delay and
waiting a random time period.
99
Computing the delay and wait
Binary Exponential Backoff algorithm

Number of trials is N.

Yes No
N < 10

Random number (T) Random number (T)

0 <= T <= (2N-1) 0 <= T <= (210-1)

Delay time is = T * 51.2µs

Wait this random time period!

100
Receiving an Ethernet frame

No Is there any incoming signal?

Yes
Bit synchronization, waiting for the
start of frame (delimiter).

Reading the frame.

No
CRC and frame length are OK?
Yes
Destination addr. = own addr. No
or a broadcast addr.?
Yes
Forwarding the data field of frame to a
Discarding the frame.
higher protocol layer for processing.

101
 Ethernet switching
• A collision domain occurs when multiple computers are
connected to the single, shared transmission media (line).
• Devices in second layer (bridge or switch) provide
switching divide the collision domains.
• Each port of a switch forms a separate collision domain.
• These devices control the transmission of frames by MAC-
addresses assigned to the Ethernet devices.
• Switches for each port stores the MAC addresses of the
accessible devices from that port in a switching table.
• Switches upload and maintain their switching tables
(cache) dynamically.
102
 Ethernet switching
Store the source MAC address of the
incoming frame into swiching table!

Is the destination address Yes

a broadcast address?
No

Is the destionation address No Transmit the frame on each port

found in switching table? (excluding the incoming port)!
Yes

Transmit the frame

on the corresponding port!
MAC address Port
00-26-9E-93-75-AA 1
00-1E-64-60-0E-B0 2
08-00-27-00-FC-E1 3
103
 Virtual LANs
• Goal: decoupling the logical topology from the
physical topology
– Migration has not got wiring effects
– Redefinition of broadcast domain
• Based on VLAN-aware switches
• Each MAC addresses or each ports of a switch
belongs to a given VLAN 1 2 3 4

– Tagged by VLAN name sw

• IEEE 802.1Q supports VLAN
on Ethernet 5 6 7 8

104
 Wi-Fi
A set of standards for implementing
Wireless Local Area Network (WLAN)
computer communication.
Located in Physical and Data link layer.
More important standards:
• IEEE 802.11a (1999)
• IEEE 802.11b (1999)
• IEEE 802.11g (2003)
• IEEE 802.11n (2009)
105
 Wi-Fi
IEEE 802.11b:
• 13 overlapping channels (EU)
with 5 MHz bandwidth on 2.4 GHz.
• Maximum 11 Mbps speed.
• Different coding/modulating technologies.

IEEE 802.11a:
• Technology working on 5 GHz (light-like propagation).
• Maximum 54 Mbps data transmission speed.
• Requires a separate radio frequency (RF) unit (5 GHz).
106
 Wi-Fi
IEEE 802.11g:
• New coding and modulating technology on 2.4 GHz
(PBCC, OFDM).
• 54 Mbps maximum data transmission speed.
• Retain frequency (2.4 GHz) provides a backward
compatibility for 802.11b systems.

IEEE 802.11n:
• Technology working on both 2.4 GHz and 5 GHz.
• 600 Mbps maximum data transmission speed.
• More antennas.
107
 Wi-Fi
• Access point (AP): network device allows a Wi-Fi
device to connect the wired network
(bridge between IEEE 802.11 and IEEE802.3)
• Wi-Fi router: complex device
– Router
Wi-Fi antenna
– Switch
Wi-Fi router
access point
– Access point
– Other router switch
• Storage
• web/ftp server WAN port LAN ports

108
 Wi-Fi
Infrastructure mode
• Wi-Fi devices directly connect only to a base station
(access point)
• Multiple access point can be present
(roaming)
Ad-hoc mode
• No base station (AP)
• Wireless devices directly connect to each other
(peer-to-peer)
• Wi-Fi Direct

109
 Wi-Fi
SSID: ‚name’ of the network
Security solutions:
• Open access
– Encryption free
• SSID is hideable
• Wired Equivalent Privacy (WEP)
– Easily breakable, week encryption
• Wi-Fi Protected Access (WPA)
– Temporal Key Integrity Protocol (TKIP; 128 bit)
• Wi-Fi Protected Access 2 (WPA2)
– Advances Encryption Standard (AES; 128/256 bit)
110
 FDDI
• Fiber Distributed Data Interface
• MAN (or LAN) media access
• Dual ring topology
• Multi-mode optical cables
• 100Mbps data rate
• Up to 200km
• Up to 1000 nodes
• Fault tolerance
• 4B5B signal coding

111
 PPP
• Point-to-Point Protocol (RFC 1661)
• WAN data-link layer protocol
• Establish direct connection between two nodes
• Used for costumer dial-up internet access (ISP to home)
• Authentication, compression, encryption, error detection
• Used over serial line, trunk line, cellular phone, optic link
• Directives: PPP over Ethernet, PPP over ATM
• Parts
– LCP: link establish, configure, testing
– NCP: supports L3 protocols: IP, IPX, AppleTalk, etc.
112
 PPP
Working scheme
• Costumer PC calls provider’s router via a modem
• Router’s modem answering, establishing physical
connection
• PC sends LCP packets to configure PPP
• NCP packets configure network layer (e.g. IP address)
• Normal Internet traffic
• NCP frees up IP address, close network layer
• LCP shuts down data-link layer connection
• Modem hang up the phone releasing physical layer
113
 N-ISDN
(Narrowband) Integrated Service Digital Network
Network services over PSTN (Public Switched
Telephone Network).
Standard channel types:
• A: 4KHz analog (telephone)
• B: 64kbps digital (voice and data)
• C: 8/16kbps digital
• D: 16/64kbps digital (signaling)

114
 N-ISDN
Standard channel combinations:
• Basic Rate Interface (BRI)
– 2B + 1D(16) channels
• Primary Rate Interface (PRI)
– 23B + 1D(64) channels (USA)
– 30B + 1D(64) channels (EU)
• Hybrid Rate Interface
– 1A + 1C channels
Its bandwidth is not enough today.

115
 B-ISDN
(Broadband) Integrated Service Digital Network
Network service demands:
• Data-, voice-, video-, multimedia transfer, interactive
communication (different bandwidth needs)
Computers using this services are connected by B-ISDN.
Privte
Data LANs telephone
centers
B-ISDN

Multi- ISDN
media work-
servers stations

116
 ATM
• Asynchronous Transfer Mode (ATM)
• Different media have different needs (low latency,
constant bitrate, nothing special, etc.)
• Protocol over ISDN, PSTN, SONET/SDH network
• Fixed-sized frames (cell: 5+48 bytes)
• In OSI data-link (L2) and physical (L1) layer
• Connection-oriented (VC: virtual circuit)
• Similar to both circuit switching and packet
switching networks
• Uses asynchronous (no clock) TDM
117
 ADSL
• Asymmetrical Digital Subscriber Line
• Most user: large download, but small upload
• Digital communication on twisted pair
max. download max. upload
ADSL 8.0 Mbps 1.0 Mbps
ADSL2 12.0 Mbps 1.0 Mbps
ADSL2+ 24.0 Mbps 1.0 Mbps

4kHz 26kHz 138kHz 1104kHz

112kHz 966kHz

phone upload download frequency

118
 ADSL reference model
Costumer Premises Equipments Provider Premises Equipments

Filter/splitter DSLAM

ADSL Data
ADSL modem network
ATM

Filter/splitter

Telephone
network
Computer PSTN
POTS/IDSN
Telephone
phone
center
119
 Frame Relay
• Standardized WAN technology
• It uses a packet switching methodology
• FR specifies the physical and logical link layers of
digital telecommunications channels
• Data is encapsulated in variable-size units (frames)
• Nodes are connected by virtual circuits (VC)

router FR switch FR switch router

LAN1 VC LAN2
DTE/FRAD DCE DCE DTE/FRAD
Frame Relay network
120
Network layer

121
 Network layer
Third layer of hybrid model (L3)
Connection between any two network nodes (not just
dirrectly connected).
Topics
• Network addressing
• Routing
• Subnetting
• etc.

122
 The IP network protocol
IP (Internet Protocol) (RFC 791)
• The network layer protocol of TCP/IP reference model.
• Widely used, it is the basic element of Internet.
• Most important characteristics:
– Structure of IP header.
– IP addressing, address classes.
– Fragment supporting.
– Datagram services towards Transport Layer.

123
 Structure of IP header
Consists of 32-bit words.
Length: Minimum 5, maximum 15 words.

Version IHL Type of service Whole length

DM
Identifier Fragment offset
F F
Time To Live Transport layer prot. Header checksum

Sender (source) IP address

Receiver (destination) IP address

Optional field(s) [0-10 words]

124
 IP addresses
• Identifies the node in Network Layer.
• 32 bit (4 byte) long.
• Dotted decimal notation
– eg. 157.45.190.57
• Managing identifiers
– InterNIC
– IANA
• For organisations not unique addresses but address
domains (network identifiers) are assigned.

125
 IP addresses
• The first part on an IP address identifies the network,
the second part identifies the node (inside the
network).
IP address

Network identification Host identification

• The IP routing based on the network identifiers.

• How many bits should be in network IDs?
– If too small, the large domains will be unused.
– If too large, only small subnetworks can be handled.
126
 Classes of IP addresses
1st byte 2nd byte 3rd byte 4th byte

Class A 0 Network Host

1 7 24

Class B 1 0 Network Host

2 14 16

Class C 1 1 0 Network Host

3 21 8

127
 Network mask
Network mask (netmask):
• A 32 bit mask, which contains bits with values of 1 in
place of network and subnetwork identifiers, and
bits with values of 0 in place of host identifiers.
Prefix length:
• The number of value 1 in netmask (number of binary
places in netmask).
Network identification Host identification

IP address

Netmask 11111111111111000000000000000000
128
 Law of First Byte

Class Leading bit(s) First byte Netmask Prefix

A 0....... 0-127 255.0.0.0 8

B 10 . . . . . . 128-191 255.255.0.0 16

C 110 . . . . . 192-223 255.255.255.0 24

129
 Special IP addresses
• Not specificated host
00000000000000000000000000000000
• ID of the specific network (network ID)
Network 000000000000000000
• Broadcast on the specific network
Network 111111111111111111
• Loopback address
01111111 Anything
130
 Fragmentation
• Cutting the packet/datagram into pieces at 8-bytes
units
• Nodes do it due to datalink MTU
• Sometimes fragments are also fragmented at internal
nodes (routers)
• Only the destination merges the fragments
• In the IP header „Fragment offset” field tells the
position of the fragment in the original packet
• DF and MF header bits are also used

131
 Fragmentation example
• Sending a packet of 1900 bytes
Source Router Destination
MTU=1024 (+20) MTU=512 (+20)
• „Original” packet:
DF=0, MF=0, offset=00000 00000000
• From source to router
DF=0, MF=1, offset=00000 00000000 (0 =0 /8)
DF=0, MF=0, offset=00000 10000000 (128=1024/8)
• From router to destination
DF=0, MF=1, offset=00000 00000000 (0 =0 /8)
DF=0, MF=1, offset=00000 01000000 (64 =512 /8)
DF=0, MF=1, offset=00000 10000000 (128=1024/8)
DF=0, MF=0, offset=00000 11000000 (192=1536/8)
132
Problems of Dual Address systems
In Network and Data Link Layers two independent
address systems (IP addresses and Ethernet addresses)
are considered.
• For encapsulation of Data Link Layer (forming an
Ethernet frame) the physical address (MAC address)
belonging to the IP address has to be determined.
• In certain cases it could be necessary to determine
the IP address by the help of Ethernet address.

133
 Network Address −> Physical Address
ARP (Address Resolution Protocol):
• Each node records physical addresses belonging to
the network addresses in a table (ARP table).
• How get a new data (pair of addresses) into the
table?
1. ARP question:
Who knows the physical address of the network address X?
2. Each node of subnet receives and processes the frame of
the question by a broadcast message.
3. If a node ‘identifies itself’ with network address X, sends an
answer to the ARP question with own physical address.

134
Physical Address −> Network Address
RARP (Reverse Address Resolution Protocol)
• RARP servers stores network addresses of given
physical addresses
• Servers replays to (broadcast) queries
BOOTP (BOOTstrap Protocol)
• Its operation is similar to RARP
• It works not just in a broadcast domain
– using BOOTP relay agents

135
Physical Address −> Network Address
DHCP (Dynamic Host Configuration Protocol):
• Allows assignment of IP address domain.
• In case of more DHCP servers, the handled address
domains should not overlap (in default).
• Clients get the IP address (and other network setup)
for a renewable time period.
• If client and server are in different network it uses
relay agents.

136
Physical Address −> Network Address
DHCP scheme of functioning :
1. DHCP question: Who can give me an IP address?
2. Each node of subnet receives the frame of the question
by a broadcast message.
3. A DHCP servers process the question: If there is a free IP
address in the handled address domain, then send an
answer to DHCP question with that IP address.
4. The client chooses one from the received DHCP answers,
and sends a feedback of its choice to the corresponding
DHCP server.
5. The DHCP server books the choice of address (the
address became occupied), and confirms client on
booking.
137
Physical Address −> Network Address
DHCP scheme of functioning :
client server

time

138
Problems with classful IP networks
• Class A networks are to large, Class C networks are to
small, Class B networks are full.
Solutions:
• Private IP domains (e.g. 192.168.0.0/16) with
Network Address Translation (NAT)
• Classless IP addressing: the border between network
and host ID is shiftable (e.g. netmask 255.240.0.0)
• IPv6, new version of Internet Protocol

139
 IPv6
• Continuous lunching (from 1994)
– Long coexists with IPv4
– Dual stack (IPv4 vs IPv6)
– Tunneling (IPv6 inside IPv4)
• 128 bits long addresses
– Network prefix (first 64 bits)
– Interface ID (last 64 bits)
• Large address space (approx. 1038 address)

140
 IPv6
• Representation: 8 groups of 4 hexadecimal digits
FE80:0000:0000:0000:32E4:00DF:FE27:8D3F
• Shorter form
FE80::32E4:DF:FE27:8D3F
• Special addresses
– 2000::/3 global unicast
– FD00::/8 local unicast (IPv4 private)
– FE80::/10 link-local unicast (valid only locally)
– FF00::/8 multicast
– ::1 loopback
141
 IPv6
• Header structure (fix 40 bytes)
version traffic class flow label
payload length next header hop limit

source address

destination address

• +Extension headers
142
 Private IP networks
• Network that uses private IP address space
• Commonly used for home and office LANs, when
globally routable addresses are not necessary
• Must use a network address translator (NAT)
• Private domains:
– 10.0.0.0/8
– 172.16.0.0/12
– 192.168.0.0/16

143
 Network Address Translation
Local address realm 1 Global address realm Local address realm 2
57.68.5.2 192.168.0.5
Server2 Client3
192.168.0.5
192.168.0.1 193.8.23.4 202.1.0.19 192.168.0.1
Client2 1
4 NAT box NAT box
2 63.12.111.5
3 Server1 172.16.14.9
192.168.1.7
Client1 72.16.10.5 Client4
Client5

• 1: Source 192.168.0.5 Destination 63.12.111.5

• 2: Source 193.8.23.4 Destination 63.12.111.5
• 3: Source 63.12.111.5 Destination 193.8.23.4
• 4: Source 63.12.111.5 Destination 192.168.0.5
144
 CIDR
Classless Inter-Domain Routing
Main problem:
We want to divide a network to different sized subnets.
(Original subnetting results same subnet size.)
Not the number of subnets is important, but the
number of nodes in a given subnet.
IP classes are not so important.
Network-host border can be shifted.
The result depends on the arriving time of demands.

145
 IP subnets
Why is it necessary to create subnets?
• The logical functionality of the institute can be a
reason.
• On an IP network more than one broadcast domains
(usually with the same size) have to be created.

How can we create a subnet?

• Some of the higher position bits of host ID of an IP
address will be used identifying the subnet.
• The new network-node boundary is denoted with
the network mask (longer prefix is used).
146
 ICMP
Internet Control Message Protocol (RFC 792)
• Implemented together with IP
• When an unexpected event happens it is reported
by ICMP messages
• Encapsulated into IP packet (8-bytes header)
• Several network utilities based on ICMP
• Examples:
– Destination unreachable (routing)
– Time exceeded (TTL)
– Echo request (ping)
– Echo reply (ping)
147
 Routing
• How to find the destination?

source
destination

148
 Routing table
Each node have a „list” about its (direct) connections
and knows who is the „best informed” of them.
This „list” is called routing table.
The „best informed” node in a network called default
gateway.
If a node wants to send a packet to an other, it searches
for the destination in its connection list. If it is in the
list, the sender can know how to reach it, else it
sends the packet to default gateway (as a next hop)
maybe it can forward the packet to the addressee.
149
 Routing table example
Computer with 2 network interfaces
172.16.0.0
255.240.0.0
eth1: 172.16.0.23

192.168.1.254

eth0: 192.168.1.100 Internet

192.168.1.0
255.255.255.0

Destination Gateway Netmask Interface

Network ID (genmask)
192.168.1.0 0.0.0.0 255.255.255.0 eth0

172.16.0.0 0.0.0.0 255.240.0.0 eth1

0.0.0.0 192.168.1.254 0.0.0.0 eth0

150
 Routing process
How the routing works in case of sending to an given IP:
• Step 1: See the first row of routing table.
• Step 2: Make AND operation between the given
destination IP and the netmask in the row.
• Step 3: If the result equal to Network ID in the row,
send the packet on your interface written at the end
of row. (If gateway given send to the gateway, else
directly to destination on the link.) Ready.
• Step 4: Otherwise see the next row (if exists) and go
to Step 2. If no further row stop with error.
151
 Routing example
Sending a packet to 193.6.128.5. Via which interface?
First row 193. 6.128. 5
& 255.255.255. 0
193. 6.128. 0 ≠ 192.168.1.0
Second row 193. 6.128. 5
& 255.240. 0. 0
193. 0. 0. 0 ≠ 172.16.0.0
Third row 193. 6.128. 5
& 0. 0. 0. 0
0. 0. 0. 0 = 0.0.0.0
Send the packet to gateway 192.168.1.254
(this is the next hop) via the interface eth0 (192.168.1.100). 152
 Maintenance of routing tables
Static (nonadaptive) routing
• The routing tables are treated by the system
administrator (root).

Dynamic (adaptive) routing

• Routers automatically change information between
each other to update their routing tables.
– Internal: finding optimal route
– External: finding trusted route

153
 Routing concepts
• Autonomous system: Administrative routing unit
with same routing strategy
• Metrics: Describes the quality of routes (distance,
cost, bandwidth)
• Routed protocols: General protocols controlled by
routers (IP, ICMP, etc.)
• Routing protocols: Controls the routing process
– Distance vector routing: RIP, EIGRP, BGP
– Link-state routing: OSPF, IS-IS, etc.

154
 Distance Vector Routing
Operation:
• Routers store the shortest distance to all nodes and
the next node on the shortest path
• Routers exchange this information between
neighbors periodically and automatically
• Routers check (based on the new information)
whether there is better path than the stored one.
Examples:
• RIP, EIGRP, etc.

155
 Mathematical background
• Direct cost (distance)

, =
cost, if i and j in the same network
∞ , otherwise

• Distance on the shortest path

0, if i = j
D , = min , ! + "(!, ) , otherwise
where k runs over the neighbors of i

156
 Composition of routing table
0, =
• Initial state: D , =
∞, ' ()*+ ,)
All i node knows d(i,k) to all k neighbors.
• Algorithm (Bellman-Ford):
1. All i node get D(k,j) from k neighbors.
2. Node i calculates D(i,j) based on Step 1.
3. If the new D(i,j) smaller then its previous value
store it and the shortest path to j goes through k.
4. Continue at Step 1.
After finite iteration we get the optimal routes.
157
 Problems with DV-routing
• Count to infinity:
The method slowly respond to topological changes.
After any change in the network longer time needed
to find the optimal path.
• Too small initial value:
If optimal path damaged available longer path can’t
overwrite it.
Solution: longer distance arrives from the direction
of optimal path overwrite it.

158
 Examples of problems
Slow convergence (after turn on)
• How far is router ’A’ from a given router?
A B C D E F

? ? ? ? ? Initially
1 ? ? ? ? After 1 exchange
1 2 ? ? ? After 2 exchange
1 2 3 ? ? After 3 exchange
1 2 3 4 ? After 4 exchange
1 2 3 4 5 After 5 exchange

Several exchange needed to discover changes.

159
 Examples of problems
Count to infinity (after break down)
• How far is router ’A’ from a given router?
A B C D E F

1 2 3 4 5 Initially
3 2 3 4 5 After 1 exchange
3 4 3 4 5 After 2 exchange
5 4 5 4 5 After 3 exchange
5 6 5 6 5 After 4 exchange
7 6 7 6 7 After 5 exchange
∞ 7 8 7 8 7 After 6 exchange
…
Never said: ’A’ unreachable
160
 RIP
Routing Information Protocol (RFC 1058)
• Distance vector based internal routing protocol
• Old, but continuously developed
• Maximum 15 router long paths
• Information sending in each 30 seconds
• If topology has changed immediate sending
• The second version (RIP v2) is CIDR compatible

161
 EIGRP
Enhanced Interior Gateway Routing Protocol
• Developed and used by CISCO
• Routing update in every 90 seconds
• CIDR compatible
• Default metric is bandwidth
• Other metrics: delay, MTU, reliability, load
• Stores potential substitute paths

162
 Link-state routing
Operation:
1. Discover neighbors
2. Measuring the cost of accessing neighbors
3. Composing packets from measure results
4. Sending the packet to all routers
5. Routers knows the topology and can calculate the
optimal paths to all other router (by Dijkstra’s
algorithm)

Example: OSPF, IS-IS

163
 OSPF
Open Shortest Path First
• Link-state interior routing
• Default from 90’s
• Use ‚areas’ (smaller units than AS)
• Different classes of routers:
– IR: Internal Router (inside area)
– ABR: Area Border Router
– BR: Backbone Router
– ASBR: Autonomous System Boundary Router
• Multipath routing
164
 OSPF
Autonomous System OSPF
IR
ABR Area

ASBR
BR

OSPF
OSPF
Backbone

BGP
165
 Congestion control
• If too many packets are present in the subnet
the performance degrades.
• Network layer have to manage this situation
– Several algorithm is used
• Different of flow control in data-link layer
„perfect world”
maximum
capacity

„real world”

166
 QoS
Quality of Service
• Different services have different requirements
services/needs Reliability Delay Jitter Bandwidth
E-mail High Low Low Low
Web access High Medium Low Medium
Audio stream Low Low High Medium
Video stream Low Low High High
Telephony Low High High Low
Video conference Low High High High

• Solutions: buffering, resource reservation, traffic

shaping, leaky bucket, etc.
167
Transport layer

168
 Transport layer
Fourth layer of hybrid model (L4)
Reliable connection between software on two nodes.
Protocols may connectionless or connection-oriented.
Topics
• Error detection/correction
• Order guarantee
• Identifying programs on a node
• Flow control
• etc.

169
 Port
Problem:
• IP address (and DNS name) identifies the node only.
• A node has more different connections, it executes
more network applications.
• A program has to know which segment (data unit in
L4) belongs to it
Solution: port
• It identifies network programs or services on a node.
• It is a 16bits long number in decimal form.
• Range: 0 - 65535
170
 Port
Range: 0 – 65535 (since it is 16bits long)
• Well-known ports: 0 – 1023
Used by system processes that provide widely used
types of network services
• Registered ports: 1024 - 49151
• Private ports: 49152 – 65535
Used freely
Stored in files:
• linux: /etc/services
• windows: C:/WINDOWS/system32/drivers/etc/services
171
 Well-known ports
• 21: FTP (File Transfer Protocol)
used for down/up loading files
• 22: SSH (Secure SHell)
used for secure login to remote computer
• 25: SMTP (Simple Mail Transfer Protocol)
used for e-mail routing to mail servers
• 53: DNS (Domain Name System)
used for eg.: www.unideb.hu 193.6.128.25
• 67: DHCP (Dynamic Host Configuration Protocol)
automatic network configuration of host
172
 Well-known ports
• 80: HTTP (HyperText Transfer Protocol)
used by web browsers
• 110: POP3 (Post Office Protocol v3)
used for downloading e-mails from servers
• 143: IMAP (Internet Message Access Protocol)
used for downloading e-mails from servers (newer)
• 443: HTTPS (HyperText Transfer Protocol over SSL)
used by web browsers for secured sites
• 995: POP3 (Post Office Protocol v3 over SSL)
used for secured downloading e-mails from servers
173
 Transport layer protocols
UDP: User Datagram Protocol
• Connection free
• Non-reliable

TCP: Transmisson Control Protocol

• Connection based
• Reliable

174
 UDP
• The UDP (User Datagram Protocol) is the connection
free transport protocol of the TCP/IP protocol set.
• Transmission of datagrams without any guarantee
(without confirmation).
• Failure management is to higher level (applications)
protocols.
• The UDP protocol is suitable for applications which
do not need to concatenate sequences of segments.
E.g. DHCP, DNS .
• Short header, fast transmission.
175
 TCP
• The TCP (Transmission Control Protocol) is the
connection based transfer protocol of the TCP/IP
protocol set. It provides a reliable (receipted) bit
stream for applications.
• Before starting data transmission, the two nodes
build up a TCP connection (Three-way handshake).
• The destination node receipts the segment(s).
• If a segment is missing, the TCP protocol ensures
retransmission of the missing segment.
• Long header, slow transmission.
176
 Headers
UDP:

TCP:

Reserved

177
 Three-way handshake
TCP need to create a connection (session) before
transmission, in 3 steps:
1. Client to Server: (SYN)
„I want to talk with you.”
2. Server to Client: (SYN, ACK)
„Ok, I am ready to talk with you.”
3. Client to Server: (ACK)
„Ok, I have heard that you are ready to talk with me”
4. Client to Server:
„I want to say that…”
178
 Three-way handshake
client server

<SYN_SENT> <LISTEN>

<SYN_RECV>
time

<ESTABLISHED>

<ESTABLISHED>

179
 Sliding window
Sender: 11 sent and 5 receipted TCP segments, window size 10
Sliding window

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Sent and Sent, but not yet Not yet sent, Not sent, and not
acknowledged acknowledged ready to receive ready to receive

Sender: 12 sent and 7 receipted TCP segments, window size 10

Sliding window

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Sent and Sent, but not yet Not yet sent, Not sent, and not
acknowledged acknowledged ready to receive ready to receive
180
 Lost segment retransmission
sender receiver
time

1
X
1
time-out

1 3

1 3 4

1 3 4 5

1 2 3 4 5

1 2 3 4 5 6
order guarantee

181
 Flow control
establishing sender receiver puffer (2048B)
connection
empty
Application
sends segments Receiving data
(total size: 3072B)

Puffer full
time

Sender blocked
Application
reads 1024B data
Application
Application reads 512B data
sends segments
Application
terminating reads 1536B data
connection 182
 Connection termination
initiator receiver
<ESTABLISHED>
<FIN_WAIT_1>
<ESTABLISHED>
<CLOSE_WAIT>
time

<FIN_WAIT_2>

<LAST_ACK>
<TIME_WAIT>
<CLOSED>

<CLOSED>

183
 Use of TCP and UDP
TCP is used when the reliability is important
• We nees all bytes precisely even if the speed is slow
• Eg: downloading file, browsing web, reading email

UDP is used when the speed is important

• We need fast, continuous transmission even if some
segments are lost.
• Eg: IP phoning (eg: Skype), watching live video

184
 Super-server: inetd
• If all server programs (daemons) always listen to
packet, it is not efficient (too much processes).
• Incoming packets first goes to inetd
• inetd decide which server program belongs to this
packet (based on port number).
• inetd launches the demon (servicing program),
delivers the packet. FTP

ssh

inetd SMTP

HTTP

POP3 185
 Access control
tcpd: Access control for internet services
Operation
• Request arrives
• The inetd launches tcpd, not the service daemon
• The tcpd logs the request
• It checks the rights
– by pattern-based access control configuration files
• Either starts the requested daemon or don’t respond

186
 Configuration and commands
• /etc/protocols
• /etc/services
• /etc/inetd.conf
• /etc/hosts.allow, /etc/hosts.deny
• telnet
• netstat
• nmap
• netcat (nc)

187
Application layer

188
 Application layer
Top layer of OSI and TCP/IP models.
Interface between network and users.
Ensures the communication required by the users.
Contains protocols needed by end users.
Main topics
• Domain names (DNS)
• World Wide Web (www, HTTP, HTML, URL)
• E-mail (SMTP, POP3, IMAP)
• File transfer (FTP, BitTorrent)
• etc.
189
Network addresses and hostnames
Problem:
• Users like alphabetical names (texts)
instead of numbers.
• Computers identify each other by IP address
(which is numerical information).
• Need of decupling names and network addresses.

Solution:
• Mapping IP addresses to names
– Central hosts.txt file (ARPANET)
– Domain Name System (DNS)
190
 Domain Name System
• Hierarchical, domain-based naming scheme
• Implemented in distributed database system
• Client-server architecture
• Decentralization and scalability
• Platform independence
• General purpose realization
– Support latter applications
• Specified in RFC 1034 and RFC 1035 (etc.)
• In use since 1980s
• E.g.: www.unideb.hu 193.6.128.25
191
 Components of DNS
Domain namespace and resource records
• The names and information about them.
• Nodes of the graph represents resources.

Name servers
• Store resource records.
• Answers queries.

Resolver applications
• Ask name servers,
for example if IP address is needed, but a name is given.
192
 Domain Namespace
root

countries generic

TLD uk hu de us com gov org

unideb bme facebook google ieee

www inf law vik www mail www

zone

www irh mail

193
 Domain Namespace
Tree graph, where each node represents a set of
resource (e.g. computer).
Each node has a label (a kind of name).
• Subset of ASCII (a-z, A-Z, 0-9, -)
– Internationalized characters (Punycode)
• Max length of labels is 63 characters.
• No case sensitivity.
• No equal labels with same parent node.
• Label of root is a string with length 0 (null label).

194
 Domain Namespace
Fully Qualified Domain Name (FQDN)
• Nodes can be identified by the series of labels from
the node to the root.
• Absolute domain name.
• User representation (max 253 characters):
irh.inf.unideb.hu.
• Binary representation (max 255 bytes):
3irh3inf6unideb2hu0
Partially Qualified Domain Name
• Relative domain name.
195
 Domain Namespace
Zone
• Administrative unit of domain namespace.
• A contiguous sub-graph
– May consist of a domain and sub-domains.
• Zones does not overlap.
• Belongs to organizations/institutions responsible for
a set of domain names.
• Contains name servers.
• Referred by its ‚highest’
domain name.
196
 Reverse lookup
The namespace has a special subgraph
• under the in-addr.arpa. domain.
• for mapping IP to domain name.
• its subdomains belongs to bytes of IP addresses.
• its Resource Records contains domain names (PTR).
Example:
• 25.128.6.193.in-addr.arpa. refers to the
domain name of node has IP address 193.6.128.25
(www.unideb.hu)

197
 Reverse Namespace

uk hu org arpa
unideb bme in-addr

www inf 0 193 255

irh 0 6 255
193.6.135.80
0 135 255
A
irh.inf.unideb.hu. 193.6.135.80
PTR
0 80 255
80.135.6.193.in-addr.arpa. irh.inf.unideb.hu. irh.inf.unideb.hu. 198
 Resource Record
A domain name specify a node of the graph.
A node related to resource set.
Information resources are stored in resource records.
Resource records (RR) stored in zone file.
The order of RRs is not important.
Examples of resource records:
• What is the IP address of a computer given by name?
• Which computer is a name server in a zone?
• Which computer is a mail-exchanger?
• etc.
199
 Resource Record
Structure:
[domain_name] [TTL] [class] type data
• domain_name: domain to which this record applies
• TTL: how ‚stable’ is the record (or validity in seconds)
volatile low value, quasi constant high value
• class: practically always IN (Internet)
• type: what kind of information is stored in data field
• data: value with type specific format/content
In case of blank optional field last record or zone file
directives are used.
 Resource Record
Frequently used RR types:
• SOA: authoritative information about the zone
• NS: authoritative name server of the domain
• A: network address of the domain (hostname)
• AAAA : IPv6 address of the domain
• MX: mail exchanger (or MTA) of the domain
• CNAME: alias name of the canonical domain
• HINFO: info about the host hardware/operating system
• PTR: pointer to reverse DNS lookup
• TXT: arbitrary human-readable text about domain
201
 Resource Record
Values of different types:
• SOA: complex record
(primary name server, email of responsible person,
serial number, timing details of refreshing)
• NS: domain name of a host
• A: IPv4 address (if class is IN)
• AAAA : IPv6 address of the domain
• MX: priority and a domain name of mail server
• CNAME: a (canonical) domain name
• PTR: domain name of a host
202
 Example zone file
$TTL 43200 ;default TTL
$ORIGIN example.org. ;base name
@ IN SOA dns1.example.org. root.example.org. (
2009100501 ; serial <2009-Okt-05, update 1>
86400 ; refresh <1 day>
3600 ; retry <1 hour>
1209600 ; expire <2 weeks>
10800 ) ; minimum TTL <3 hours>
example.org. 86400 IN NS dns1.example.org.
example.org. 86400 IN NS dns2.example.org.
example.org. 86400 IN MX 10 mail.example.org.
dns1.example.org. IN A 192.168.0.1
dns2.example.org. IN A 192.168.0.2
mail.example.org. IN AAAA 2001:503:ba3e::2:30
server.example.org. IN A 192.168.0.4
host.example.org. IN A 192.168.0.101
e2.example.org. IN A 192.168.0.102
ftp.example.org. IN CNAME server.example.org. 203
 Resolver
A software, which means interface between user
network applications and name servers.
Client side of the DNS (usually platform dependent).
If a program needs IP address but domain name is
given address resolver do the address mapping.
It sends a request to name server and gives the reply
based on resource records to the user application.
Results:
• an RR-based answer (www.unideb.hu 193.6.128.25)
• name error
• data not found
204
 Name server
The name server is a software on a computer, which
• stores resource records of a zone (zone file)
• knows connections to neighboring zones
• temporarily stores some RRs of other zones
– Cache: based on TTL fields of RRs
• replies to resolver query.
Each zone has name servers
• primary name server
authoritative zone file managed by administrator
• secondary name server
automatic copy from primary NS (see SOA record)
205
 Primary and secondary servers
primary NS expire
refresh refresh retry retry retry
secondary NS
client

time

206
 Query
• Structure of query and answers are the same
Ethernet header
IP header
UDP header
Header
opcode=standard response; …

APDU (DNS response)

Question
name=www.freemail.hu; class=IN; type=A
Answer
name=www.freemail.hu; class=IN; type=A; TTL=3112; address=84.2.43.64
Authority
name=freemail.hu; class=IN; type=NS; TTL=2540; name server=ans0.t-online.hu
Additional
name=ans0.t-online.hu; class=IN; type=A; TTL=83410; address=195.56.77.76
Ethernet tail
207
 DNS lookup process
• Client’s resolver asks the name server
(send a query in UDP segment, destination port 53)
• Server looks for answer
– in temporary memory (cache)
– in own Resource Records
– ask other servers
• Iterative query
• Recursive query
• Other servers’ answer is stored in cache for a while
• Name server response to client
208
 Iterative and recursive query
IP of „www.unideb.hu”? „root”
DNS sever
Client iterative
method

„hu” TLD
recursive DNS sever
ISP
method DNS sever

„unideb.hu” zone Authoritative

DNS sever
www.undeb.hu

209
 Iterative and recursive query
Iterative method
• Easy implementation on server
• Implemented on all name server
• Answer can be a reference to other servers
Recursive method
• Easy implementation on client
• Must be implemented on both side
• Special flag bits in query/response header
• Answer allways the asked information (or error)

210
 Configuration and commands
• /etc/hosts
192.168.0.23 RedLaptop
• /etc/nsswitch.conf
host: dns files
• /etc/resolv.conf
domain unideb.hu
nameserver 193.6.128.5
• nslookup
– Interactive mode
– Non-interactive mode
• host
211
 World Wide Web
The most widely used and most quickly spreading part
of Internet.
Concept: Tim Berners-Lee (CERN, 1989)
We can navigate among websites by hyperlink.
Based on:
• URL (Uniform Resource Locator)
• HTML (HyperText Markup Language)
• HTTP (HyperText Transfer Protocol)

212
 URL
Known as web address.
All webpage can be refered by URL.
Its parts:
• Scheme (protocol)
• Domain name or IP address
• Port number
• Path and name of file on server
• Query string
• Fragment identifier (bookmark)

213
 URL examples
• http://www.example.org:80/index.html?lang=eng#top
• http://www.unideb.hu
• ftp://152.66.115.246/.banner
• http://neptun.unideb.hu/?page=studhun
• https://hu-hu.facebook.com/login.php
• http://en.wikipedia.org/wiki/HTML#History
• mailto:[email protected]
Legend:
• Scheme • Path
• Domain name • Query
• Port • Fragment ID 214
 HTML
A desription language to create websites.
Standardized by W3C (World Wide Web Consortium).
Websites are text-based files (contains only characters)
which is represented (in visual form) by browsers.
Popular browsers:
• Internet Explorer • Opera
• Mozilla Firefox • Safari
• Google Chrome • Konqueror
• Netscape Navigator • etc.

215
Example HTML file

216
 Hyperlink
A (hyper)link is a reference to data that the reader can
directly follow (by a click).
A hyperlink points to
• a whole website or an element within a page,
• different media (picture, audio, video).
Hipermedia is a media with hiperlink.
• Media can be text, picture or video.
Hyperlink based on URL.
Example:
<a href=”http://www.google.hu”> Google </a>
217
 HTTP
HTTP is a request-response (client-server) information
transmission protocol of application layer. (RFC 1945)

Client: web browser which visualize web pages for user.

Server: computer (webserver) which stores webpages.

Usually it uses TCP connection (in Transport layer)

Safer solution: HTTPS (HTTP Secure)

HTTP over SSL/TLS protocol

218
 HTTP
• Request format
method sp resource sp HTTP version crlf
header field : value crlf
header
f u r field:
t h e r h e: a d e r value
f i e l d s …crlf
crlf

entity body

sp space character
crlf carriage return + line feed characters
: colon character

219
 HTTP
• Response format
HTTP version sp status code sp pharse crlf
header field : value crlf
header
f u r field:
t h e r h e: a d e r value
f i e l d s …crlf
crlf

entity body

sp space character
crlf carriage return + line feed characters
: colon character

220
 HTTP Status Codes
• 1xx: Request received, continuing process.
• 2xx: Indicates the action requested by the client was
received, accepted and processed successfully.
• 3xx: The client must take additional action to
complete the request.
• 4xx: In cases when the client seems to have erred.
• 5xx: The server failed to fulfill a valid request.

221
 Browsing web
1. The user gives the URL in the address bar of browser.
2. The web browser determines the protocol from URL
(eg. http://...).
3. It determines the (IP) address of web server from
domain name in URL via DNS (eg. www.unideb.hu).
4. It builds up a session with web server (via TCP
usually using port 80).
5. A request sent to HTTP server giving the name of the
folder (and the HTML file) containing the web page
(eg. /index.html).
222
 Browsing web
6. The server responds the request by sending to
client the text or other medias (pictures, sounds,
clips, etc.) defined in the HTML page.

7. The browser (client) composes files, displays the web

page to user, and closes the session.

223
 Cookie
• Name and value pair to ensure stateful operation
• Browser sends a usual request client server
• Server sends a „Set-Cookie”
header field
• Client saves the cookie (information)
• Later when the browser requests

time
the same site it sends the cookie.

later
• Server sends „personalized” site
based on the cookie value

224
 Browsing in terminal
linux$> telnet irh.inf.unideb.hu 80
Trying 193.6.135.80...
Command
Connected to erlang.inf.unideb.hu.
Escape character is '^]'.
GET /index.htm HTTP/1.1
Host: irh.inf.unideb.hu
Request

HTTP/1.1 200 OK
Date: Wed, 12 Feb 2014 11:26:45 GMT Reply header
Server: Apache/2.2.17 (Fedora)
Last-Modified: Sun, 20 Jan 2013 11:22:30 GMT
ETag: "1440c6d-135d-4d3b68f634980"
Accept-Ranges: bytes
Content-Length: 4957
Connection: close
Content-Type: text/html; charset=iso-8859-1

<html><head>
<title>DE IK IRH</title> … Reply: requested page
225
 E-mail
Electronic mail (E-mail, email, eMail)
A method of exchanging digital messages from an
author to one or more recipients. (RFC 821)
E-mail address:
local_part@domain_part
user@provider
E-mail contains 2 sections
• Header:
It has several fields (sender, addressee, subject, …)
• Body:
The ‘message’.
226
 E-mail header fields
• From:
Sender’s e-mail address
• To:
The e-mail address(es) of the recipient(s)
• Subject:
Topic of the message
• Date:
The local time and date when the message was written
• Message-ID:
Automatically generated to identify the message
227
 E-mail header fields
• Cc:
E-mail addresses who will get copies of message.
• Bcc:
E-mail addresses of recipients who won’t see
eachother in the header of their message.
• Reply-To:
Address that should be used to reply to the message.
• Content-Type:
Information about how the message is to be displayed,
usually a MIME type.
• and much more…
228
 Body of e-mail
Originally it contains only characters (text).
Modern graphic email clients allow the use of either
plain text or HTML.
Multipurpose Internet Mail Extensions (MIME, RFC
2045) is an Internet standard that extends the format
of email to support:
• Text in character sets other than ASCII (eg.: áíűŁäšş)
• Non-text attachments (jpg, pdf, mp3, avi)
• Message bodies with multiple parts
• Example: text/plain, text/html, image/jpeg
229
 Mail servers and clients
• Programs used by users for managing e-mails are
called Mail User Agents (MUA).
• MUA submit the e-mail to Mail Submission Agent
(MSA) who will forward it.
• Messages are exchanged between hosts using the
Simple Mail Transfer Protocol (SMTP) with software
programs called Mail Transfer Agents (MTA).
• Messages are delivered to a mail store by programs
called Mail Delivery Agents (MDA). Users can retrieve
their messages from servers using standard protocols
such as POP3 or IMAP.
230
 Process of e-mailing
1. Sender composes the message and hits „Send”
button.
2. Senders MUA formats the e-mail and sends it to
MSA by SMTP.
3. MSA forwards the e-mail to recipients MDA (and
perhaps to some internal MTA) by SMTP.
4. The MDA delivers e-mail to the recipients mailbox.
5. Recipient presses the "get mail" button in own
MUA, which download the e-mail from MDA by
POP3 or IMAP.
231
 Process of e-mailing
MSA MTA MTA MDA

SMTP SMTP SMTP

Internet
SMTP POP3

MUA MUA

sender recipient
user user
232
 Process of e-mailing
1. Email header contains the destination email address
2. MSA asks the MX record of domain name after @
3. Name server answers the name of mail exchanger
server of the destination domain
4. MSA asks the IP address of mail exchanger from DNS
5. The email is sent to the port 25 of the given IP address
by SMTP
6. MDA receives the message and gets the username
(destination email address part before @)
7. MDA puts the mail to the user’s inbox mail folder
8. Recipient’s MUA download mails from MDA by POP3
233
 Connect to SMTP server
linux$> telnet mail.server.com 25
Trying 193.6.138.45...
Connected to delfin.unideb.hu.
Escape character is '^]'.
220 delfin.unideb.hu ESMTP Postfix (Ubuntu)
helo mail
250 delfin.unideb.hu
mail from: [email protected]
250 2.1.0 Ok
rcpt to: [email protected]
250 2.1.0 Ok
data
354 Enter mail, end with "." on a line by itself
Subject: test

This is a test e-mail.

.
250 2.0.0 Message accepted for delivery
Connection closed by foreign host.

234
 Connect to POP3 server
linux$> telnet freemail.hu 110
Trying 195.228.245.1...
Connected to freemail.hu.
Escape character is '^]'.
+OK <[email protected]>
USER proglabor
+OK
PASS proglabor
+OK
LIST
+OK
1 2442
2 12658
.
RETR 1
+OK
Message-ID: <[email protected]>
Subject: Important mail to you
From: "Dr. Varga Imre" <[email protected]>
To: [email protected]

Dear Friend, … 235

 FTP
• File Transfer Protocol
• Details in RFC 959
• Client-server architecture
• Down/upload files from/to servers
• 2 channels (Control & Data)
• FTP server codes (e.g. 220 Service ready for new user.)
• Anonymous FTP
• Browsers support it
• Much popular solution is the peer-to-peer BitTorrent

236
 Active and passive modes
• Client connects to port 21 of server (control channel)
client server
• Active mode: data control control data
3527 3526 21 20
– Client opens a port
(to data channel)

time
– Server connect to it

• Passive mode: client

data control
server
control data
– Server opens a new port 3527 3526 21 20
8415
(to data channel)
– Client connect to it

time
237
 FTP
Download the rfc0959.txt file which is in
documents/rfc folder of ftp.bme.hu server!
• In browser:
ftp://ftp.bme.hu/documents/rfc/rfc0959.txt
• In terminal:
linux$> ftp ftp.bme.hu
Name (ftp.bme.hu:user): anonymous
Password:
ftp> passive
ftp> cd documents/rfc
ftp> get rfc0959.txt
ftp> quit
238
 Connect to FTP server
Terminal 1 (Control channel) Terminal 2 (Data channel)
linux$> telnet ftp.bme.hu 21
Trying 2001:738:2001:2001::c1ca...
Connected to ftp.bme.hu.
Escape character is '^]'.
220--- Welcome to Pure-FTPd ---
USER anonymous
331- Welcome to ftp.bme.hu FTP service.
PASS
230 Any password will work
EPSV
229 Extended Passive mode OK (|||62282|)
RETR ReadMe.txt linux$> telnet ftp.bme.hu 62282
150 Accepted data connection Trying 2001:738:2001:2001::c1ca...
226-File successfully transferred Connected to ftp.bme.hu.
QUIT Escape character is '^]'.
221 Logout.
Connection closed by foreign host. This is the content of ReadMe.txt

Connection closed by foreign host

239
 ssh
Secure Shell
• Remote command-line login
• Encrypted data communication

linux$> ls
a.out Desktop prog.c program.log
linux$> ssh [email protected]
[email protected]'s password:
Last login: Thu Feb 13 12:49:32 2014 from
erlang.inf.unideb.hu
[remote]$ ls
Desktop inetd.conf readme.txt run.sh
[remote]$ exit
logout
Connection to irh.inf.unideb.hu closed.
linux$>
240
 Other parts of Application layer
• Remote login (telnet, ssh)
• Down/uploading files (scp, FTP, bittorent)
• Voice over IP (VoIP) (Skype, MSN)
• IPTV (UPC)
• Distributed databeses
• Online games
• etc.

241
 Mobile telephone systems
Cellular wireless network
• 1G: NMT (analog voice)
• 2G: GSM (ca. 0.01 Mbps)
• 2.5G: GPRS (ca. 0.1 Mbps)
• 2.75G: EDGE (ca. 0.5 Mbps)
• 3G: UTMS (ca. 1 Mbps)
• 3.5G: HSDPA/HSUPA (ca. 10 Mbps)
• 4G: LTE, WiMax (ca. 100 Mbps)
• 5G: coming soon… (ca. 1000 Mbps)

242
 GSM architecture
PSTN
M
B BSC ISDN
S T
M S
S
MSC
B
T VLR
S B
T BSC AuC
S EIR
HLR

B
T MCS
S B
T VLR
S BSC
M
S

243
 GSM architectures
• Mobile Equipment (ME)
Mobile station (MS)
• Subscriber Identity Module (SIM)
• Base Transceiver Station (BTS) Base Station
• Base Station Controller (BSC) Subsystem (BSS)
• Mobile Switching Center (MSC)
• Home Location Register (HLR)
Network Switching
• Visitor Location Register (VLR)
Subsystem (NSS)
• Authentication Center (AUC)
• Equipment Identity Register (EIR)

244
 Network
setup & commands
on Windows

245
 Network setup
4 necessary properties to use network on a computer:
• IP address
• Netmask
• Gateway
• DNS server
They are given by the Internet Service Provider (ISP).
Either the user do their setup or use DHCP (if possible).
The user can use ‘command-line’ or GUI (Graphical
User Interface) to do setup process.
246
 Network setup by Control Panel*
1. Go to Start Menu
2. ‘Control Panel’
3. ‘View network status and tasks’
in ‘Network and Internet’ block
4. ‘Change adapter settings’
5. Right mouse click on the adapter, choose ‘Properties’
6. Choose ‘Internet Protocol Version 4 (TCP/IPv4)’ and
push button ‘Properties’
7. Choose automatic configuration (DHCP) or give the
four datas.
* on Windows 7 operating system 247
 Useful network commands on Windows

• ipconfig: Shows the setups of network adapters.

• ipconfig /all: Shows the detailed setups of
network adapters.
• route print: Shows the routing table of the
computer.
• ping <node>: Check connection to other
computers.
• arp –a: Shows ARP table.
• tracert <node>: Show the hops to a computer.
• netstat –s: Shows network statistics (IP, TCP, UDP).
248
 Wireshark
• Free and open-source packet analyzer

249
 References & further readings
• Andrew S. Tanenbaum: Computer Networks,
Prentice-Hall, 2003
• Wikipedia,
http://en.wikipedia.org
• Béla Almási: Számítógép hálózatok, University of
Debrecen

250