lecture 5 Flashcards
why do we need software abstraction
to make hardware convenient
how do we deal with complex problems
- by layering, each layer deals with a certain part of the problem (divide and conquer)
what is a protocol
- a set of rules/procedures that define how data is transmitted between different devices/routers
- helps sender to communicate with the receiver and vice versa
difference between interface and protocol
- interface is the protocol between corresponding layers (e.g layer 5 and layer 4) on the same stack
- protocol is between layers of different stacks
what happens in the sender layers
- checksums are added
- overheads are added
what happens in the receiver layers
- decryption
- removal of the checksums and overheads added in the sender layers
protocol stack
layer n protocol
what is a checksum
is a small-sized block of data added for the purpose of detecting errors
what is an overhead
is any combination of excess computation time, memory, bandwidth, or other resources that are required to perform a specific task
as a network designer you need…
- to decide on the amount of layers
- decide on the interfaces+protocol stack
the philosophy-translator-secretary architecture
- message from philosopher 1 is in a different language
- middle ground translators speak the same language
- from philosopher to translator
- translator to secretary
- secretary to secretary 2
- secretary 2 to translator 2
- translator 2 to philosopher 2
two fundamental concepts of network architecture
- message
- encapsulation
what is fragmentation
the process of splitting a big message into smaller parts
opposite of fragmentation is…
defragmentation
ethernet’s minimum and maximum frame size is
minimum: 64 bytes
maximum: 1518 bytes
what is a header
a header is supplemental data placed at the beginning of a block of data being transmitted/stored
what does header help with
- when a header is clear and unambiguous, it allows for parsing to happen
why do messages have a control and header (4)
- used to synchronise with the remote peer
- contain instructions that tell the remote peer what to do with the message (protocol)
- security purposes(protocol)
- error detection
what is encapsulation
it is when layer N-1 encapsulates and treats the data portion received from layer N as an entire integral unit
e.g layer 4 encapsulates data from layer 5 (transport encapsulate session’s)
what happens when the remote peer receives the encapsulated message
- it strips off the header, trailer, overheads, checksums off the data and passes it to the layer above
what is peer to peer process
when layer N of machine A communicates with layer N of machine B
design issues of layers
- identifying senders and receivers
- rules of data transfer (simplex, half duplex, full duplex)
- error control
- order and length of messages
- fast sender, slow receiver
what is addressing
number or address that is assigned to any device that seeks access to or is part of a network
2 types of layer services
- connectionless
- connection oriented
features of a connectionless service
- does not require protocols
- message delivery not guaranteed
- considers each packet of the same message independent of each other:
1. can have different addressing
2. different route taken - service interface is message based
features of a connection oriented service
- first establishes a connection between the source and destination
- authentication is needed
- error received if message delivery fails
- service interface is stream based
- reliable
the 5 service primitives of a connection oriented service (system calls)
- LISTEN
- CONNECT
- RECEIVE
- SEND
- DISCONNECT
meaning of: LISTEN
block waiting for an incoming connection
meaning of: CONNECT
establish a connection with a waiting peer
meaning of: RECEIVE
block waiting for an incoming message
meaning of: SEND
send a message to the peer
meaning of: DISCONNECT
terminate a connection
sequence of system calls
- request connection
- acknowledge connection
- request for data
- reply
- disconnect
- disconnect back
two kinds of protocol stacks
- OSI model
2. TCP/IP
7 layers of the OSI
- application
- presentation
- session
- transport
- network
- data link
- physical
what 3 layers of the OSI are in the subnet
- physical
- data link
- network
what communication technique does the OSI subnet use
- store and forward
- hop by hop
what is store and forward
a data communication technique in which a message transmitted from a source node is stored at an intermediary device before being forwarded to the destination node
the unit in the physical layer
bit
voltage and binary bits
- 1 is represented using high v
five problems with the physical layer
- how many volts represents a “1” or “0”?
- how initial connection is established?
- how long does a bit time last?
- how many wires does the transmission media have+require?
- are pulses electrical or optical or waves?
the unit in the data link layer
frames
what does the data link layer handle
- transferring the frames to the network layer free of errors
- handles damaged, lost and duplicate frames
- slows down a fast transmitter (flow control)
- creates and recognizes frames (physical layer just transmits bit stream but with no regard of meaning and structure)
two sublayers of the data link layer
- MAC (MEDIA ACCESS CONTROL)
2. LLC (LOGIC LINK CONTROL)
the unit in the network layer
packets
difference between frame and packet
- frame includes the source and destination MAC addresses (i.e., the physical address of the machine)
- packet includes the source and destination IP addresses
what does the network layer do (THE BOSS)
- controls operations in the subnet (communication between hosts)
- deals with routing
- handles congestion
- handles addressing
what is routing (network layer)
the process of transferring packets received from the Data Link Layer of the source network to the Data Link Layer of the correct destination network.
what does routing involve
involves decision making at each intermediate node on where to send the packet next so that it eventually reaches its destination
what is congestion
If the incoming rate of the packets arriving at any router is more than the outgoing rate
what causes congestion
- when packets arriving on an input line all need the same output line (queue)
- insufficient memory to hold all of them (packet loss)
- increased time which causes the packets that reached to the front of the queue to time out (duplicates will be sent, more congestion)
- slow processors
- low bandwidth
where is routing simple (network layer)
during broadcasting
unit of the transport layer
PDU (protocol data unit)
most important job of the transport layer
- process to process delivery
- makes sure data gets delivered to a specific process on a specific machine
the network layer does not guarantee
the packet will reach its intended destination
what does the transport layer handle
- retransmissions
- duplicates
- end to end protocol
- flow control
what does the session layer do
- dialogue control
- provide synchronization
- recovers failure from transport layer (e.g. you download a file but internet cuts off halfway, file down pauses and continues when internet is back)
what does the presentation layer do
- compress data
- encryption
- apply/encode syntax and semantics to the data
- format data (names, IDs etc)
what does the application layer do
- responsible for providing services to the user
- links a specific protocol to the service
e.g.
HTTP
SMTP
FTP
telnet
4 layers of the TCP/IP
- application layer
- host to host transport layer
- internet layer
- network interface layer
TCP/IP in relations to OSI
- physical and data link layer = network interface layer
- network layer = internet layer
- transport layer = host to host transport layer
- session, presentation and application = application layer
4 protocols used in the network interface layer
- ethernet
- token ring
- frame relay
- ATM (asynchronous transfer mode)
protocol used in the internet layer
- IP
protocols used in the host to host transport layer
- UDP (user datagram protocol)
- TCP (transmission control protocol)
protocols used in the application layer under TCP
- telnet
- FTP (file transfer protocol)
- SMTP (simple mail transfer protocol)
protocols used in the application layer under UDP
- DNS (domain name system)
- RIP (routing information protocol)
- SNMP (simple network management protocol)
TCP features
- connection oriented
- reliable
- used for flow control
- byte stream
- high quality
UDP
- connectionless
- unreliable
- does not do flow control
- datagram
- used for transmitting voice/speech/video
- faster delivery
TCP/IP layers and addresses
- network interface layer: physical address
- internet layer: logical address
- transport layer: port address
- application: specific address
sender physical address is on the
RIGHT (next to it is the logical address)
receiver physical address is on the
LEFT
size of physical address
6 bytes (12 hexadecimal digits)
why does the physical address change but the port and logic address don’t
because the physical address goes hop by hop
5 main differences between OSI and TCP/IP
- OSI focused on model before the protocol, TCP/IP focused on the protocols before the model
- OSI’s network layer is connection oriented and connectionless
- TCP/IP’s network layer is connectionless
- OSI’s transport layer is connection oriented ONLY
- TCP/IP’s transport layer is connection oriented and connectionless
which protocol stack is best for designing new networks
OSI model
