CSE 3461 Midterm
point-to-point vs multipoint
point-to-point - 2 devices share a link multipoint - more than 2 devices share a link
Networking deals with?
technology and architecture of the communications networks used to interconnect communicating devices
attenuation distortion
attenuation is different for different frequencies
Communication infrastructure enables ___________ applications
distributed
Increasing data rate _______________ error rate
increases
Benefits of digital transmission?
- lost cost - data integrity - capacity utilization (high bandwidth) - security & privacy (encryption) - integration (analog and digital data treated similarly
Transport layer?
- reliable data exchange - indepdendent of network being used - independent of application
application lyaer
- support for different user applications (email, file transfer, etc)
Bipolar-AMI encodes data how?
0 is at 0 level 1 alternates between positive and negative voltage (helps with error detection) requires extra level for sending data
Shannon capacity
Determines the channel capacity given the bandwidth, noise, error rate, and data rate
What is a repeater
A repeater work with digital transmission it will extract the bit pattern and retransmit. In other words, noise is not amplified
Formula for gain in dB
10*log(Pout/Pin) log is base 10
What is circuit switching
A physical and continuous circuit that is established before, and maintained during communication. example would be a telephone network
Open Systems Interconnection (OSI)
A set of protocols allowing different types of computers to be linked together. - showed up too late, and never lived up to promises TCP/IP is major player
full duplex transmission
A type of data transmission in which data can move in both directions at the same time. example: telephone
Half-duplex transmission
A type of data transmission in which data can travel in either direction but only in one direction at a time.
How does a parity check work?
Add 1 bit to the end of the message, allows us to detect only odd number of errors There are two types of parity checks even and odd parity even parity means that the parity bit will be set to 0 or 1, whatever is needed to make the total number of ones even
What techniques can you use to encode digital data to an analog signal?
Amplitude shift keying (ASK) Frequency shift keying (FSK) Phase shift keying (PSK)
What type of errors can CRC detect?
Anything that is not divisible by the polynomial, We can detect all single bit errors as long as the polynomial is more than one nonzero term We can get all double errors as long as polynomial has 3 or more terms any odd number errors as long as the polynomial contains a (x+1) factor any burst error which is less than or equal to (n-k)
How do you detect errors?
Append some bits to the end that can assist in error detection. two common errors are Parity check and CRC (cyclic redundancy check)
ARQ
Automatic Repeat Request system used for error control - if a frame contains errors, didn't receive the frame, etc
Scrambling techniques?
B8ZS HDB3
Multilevel binary techniques?
Bipolar-AMI Pseudoternary
When can you lose synchronization with Bipolar-AMI? Pseudoternary encoding?
Bipolar-AMI - lose it with a long string of zeros Pseudoternary - opposite (when we have a long string of ones
What type of LAN Topologies are there?
Bus - A tree topology with no branches Tree - multiple lines that branch out and connect like atree ring - unidirectional travel of packets star - Star shape, there is a central hub/switch/repeater that rebroadcasts packets/frames, and sends forwards towards the destination
Pros and cons of Multilevel binary techniques?
Cons: Not as efficient as NRZ require 3 different levels require about 3dB more power for same probability of error Pros No DC component error detection capability synchronization is better lower bandwidth required
What is the three layer network?
Consists of network access layer transport layer application layer
How does Frequency shift keying (FSK) work?
FSK has two types FSK or BFSK basically represent two different values with two different frequencies, less error prone than ASK mutli FSK uses more than two frequencies - more bandwidth efficient - however more prone to error
Cyclic Redundancy Check works how?
For each block of k bits to transmit the transmitter generates n-k bit sequence called the CRC code to check if any errors basically the value transmitted should be divisible by some number, if it isn't then some transmission error occured
Common error correction process?
Forward error correction FEC encoder converts the k-bits of data into n-bits of what we call the codeword The code word is transmitted to the receiver a FEC decoder will decode the codeword into data, if no errors or they are correctable, we send the data on, otherwise we have to throw out
Routers do what?
Forward packets of data through network
LAN typically contains what protocols?
From outer protocol (most further out on "onion" design) MAC LLC - typically omitted IP TCP Application data
Say you want to send a message from your PC to a server, what is the simplified version of the path it will travel to get to the server?
From the workstation it goes to the modem, the modem will forward it through the network/internet until it reaches the modem that the server is on. The modem will forward it on to the server
Nyquist rate
Given the bandwidth of the message and the number of levels (for digital signal transmission), how much channel capacity do we need to exactly replicate a message?
How does NRZ-L encode data?
High voltage = 0 low voltage = 1
Increasing the frame size will help increase max utilization link, why can't we just increase frame size?
If multiple devices share the same transmission medium the large frame means it hogs the medium for longer Also higher chance for error, so more would need to be transmitted in the event of an error
What is the issue with a lot of daisy chaining in star topology?
Increased distance between nodes means increased propagation time, which means more timeouts
LAN vs WAN
Local area networks, wide area networks self-explanatory
attenuation?
Loss of power in a signal as it travels from the sending device to the receiving device another definition: signal strength falls off with distance
Pseudoternary encodes data how?
Opposite of Bipolar-AMI 1 is at 0 level 0 alternates between positive and negative voltage
As an analogy, MAC address and IP address are like: ____________ like SSN ____________ like postal address
MAC address; IP address MAC - unique, never changes on a device IP - gives geolocation, varies based on where device is
What layer does LAN add?
MAC layer
Biphase techniques
Manchester Differential manchester
What is a MAN
Metropolitan Area Network - middle ground between LAN and WAN, sorta like internet connect for the city and its surrounding area.
How does B8ZS work?
NOTE: this works for bipolar-AMI When we have 8 zeros back to back, we substitute with the following patterns: 000VB0VB V = violation B = follows rules
How does HDB3 work?
NOTE: works on bipolar-AMI When we have string of 4 zeros substitute accordingly: number of pulses before is odd - 000V number of pulses before is even - B00V
What is delay distortion?
Only an issue with guided media propagation velocity varies with frequency, thus some components of one bit position spill over into another bit position, causing inter-symbol interference
Protocols are split into multiple layers, when 2 machines communicate, how do the layers communicate?
Peer layers communicate
How does Phase shift keying (PSK) work?
Phase of signal is shifted to represent data two-level PSK - two different phases represent two binary digits Differential PSK - phase shifted relative to previous transmission rather than some reference signal four-level PSK aka QPSK - ie we just shift pi/2 or 90 degrees. This allows each element to represent two bits PSK and QPSK are about 3dB superior to ASK and FSK
If both sides are sending info frames and we are using sliding window control, how can we acknowledge the other side?
Piggyback the acknowledgement in the header of the info-frame If no data to send just send ACK If data but no ACk to send, send last ACK number again
What is POTS
Plain Old Telephone Service
ARPANET
Predecessor to the internet, started in late 1960s Lead to IP and TCP
List 2 internet standards
RFC - request for comments IETF - Internet Engineering Task Force
Stop and wait ARQ
Recall the stop and wait method where we only send 1 frame at a time and wait for a ACK With ARQ, we need to number the frames we are sending with 0 or 1 Suppose the Info-frame is damaged - Sender has timeout for each info-frame - If no ACK within timeout just retransmit the frame Support he ACK is damanged - Info-frame sender will not recognize the ACK - sender will retransmit the info-frame - receiver gets two copies of the same info-frame - receiver has no idea it is a retransmit unless we number the frames
What is RNR
Receiver Not Reader RNR(x) - acknowledges up to x-1 frames, but tells the sender, that they are not read To resume send RR(x) to continue receiving
Protocol Data Units (PDU)
Refers to the data, headers, and trailers about which a particular networking layer is concerned. Basically at each layer (application, transport, network access layer) we use a protocol to communicate with peer layers on another device. the PDU may split up data into fragments. Each PDU also contains headers and other meta information for the protocol being used
Examples of communication nodes?
Routers, switches, bridges
How does a Ring topology in LAN work?
Say host C broadcasts a packet, it travels in 1 direction around the ring, if one of the hosts on the LAN recognizes that the message is for them, then copy the message and let it continue around the ring, the original sender of the message will absorb the returning frame after it makes the round trip around
Stop and wait flow control
Send only 1 frame at a time, once dest receives the frame, it must reply with an acknowledgement (ACK-frame) Source must wait for ACK before sending next frame basically only 1 message can ever be send or received at a point in time (pretty inefficient)
What is encoding (modulation)?
Taking data (digital or analog) and converting it to either a digital or analog signal
Go-back-N ARQ
The first packet with an error and all those that followed by others with errors. - after info-frame timeout, send RR with P bit set and after a response info-frame retransmission possible - negative acknowledgement, REJ control frame with N-bit receive sequence number field sending REJ(x) means, hey we got everything up to x-1, I need x again
How do you choose the polynomial for generating the CRC?
The highest power of the polynomial should be the desired length of the FCS, at a minimum, you should have x^(n-k) and x^0 be nonzero
Why are protocols split into multiple layers/modules
The logic is complex, by splitting into modules, we can swap out components without having to re-implement logic.
What is a LAN topology
The setup of the network (ring, star, branch, tree, etc)
Manchester coding encodes data how?
There is a clock transition in the middle of each bit period, which also serves as the clock and data Low to high transition = 1 high to low transition = 0
Data link control
Transmits data between adjacent nodes it is concerned with blocks of bits, called frames, assumes services of physical layer and provides lay of logic above it
Who developed IP/TCP
US Defense Advanced Research Project
Pulse Code Modulation
Used for voice encoding from analog data to digital signal The idea it is based on states that if we sample at a rate higher than 2x the highest signal freq, then the sample taken will contain all the info of the original signal
How do we signal the start and end of a frame that we are transmitting?
Using a flag sequence. Issue with this though is if the sequence shows up inside the frame solution: bit stuffing, for example suppose the flag is 01111110 if we have the pattern 011111, we insert a 0 after the sequence, the receiver will discard the 0
How does Amplitude shift keying (ASK) work?
Values represented by different amplitude of carrier For example, amplitude of 0 is just binary zero, efficient over optial fiber, but not copper Can get suddent gain changes, which could mess with the message sent
How does asynchronous transmission work?
avoids timing problems by sending small number of bits (for example transmit one character at a time In other words, timing only needs maintaining within each character, we resynchronize with each character
When you amplify an analog signal, what happens?
You amplify both the signal but also the noise
What are network protocols?
a defined format, order of messages sent, and received among network entities and actions taken on message transmission, receipt
What is packet switching?
a mode of data transmission in which a message is broken into a number of parts that are sent independently, over whatever route is optimum for each packet, and reassembled at the destination. key takeaways - small chunks of data at a time - packets passed from node to node between source and dest - data may be received out of sequence - used for terminal to computer and computer to computer communications
Sliding windows flow control works how?
allows multiple info-frames to be in transit in one direction Lets the transmitter send up to W unacknowledged info-frames W = max window size Frames are numbered as they are sent from 0, 1, 2, ... (2^N)-1 receiver must send ACK frames in response to the receiver
analog vs digital data
analog - continuous values within some interval digital - discrete values
analog vs digital signal?
analog - continuously varying EM waves that propagate over a medium (can be a variety different mediums) digital - sequence of voltage pulses transmitted over a wire medium. (Signal maintains a constant level for some time and then change to another constant level)
Digital transmission, the signal can be either _________ or _____________
analog or digital
List the types of transmission impairments
analog signal - degradation of signal quality digital signal - bit errors
what causes transmission impairment?
attenuation and attenuation distortion delay distortion noise
What is a RR frame?
called Receiver Ready frame It has the format RR(x) where x indicates that all frames (x-1) and lower are acknowledged
Technologies used in WAN?
circuit switching packet switching frame relay Asynchronous transfer mode (ATM)
2 types of Communication services
connectionless connection-oriented
Pro and Cons of Biphase tehcniques?
cons at least 1 transition per bit time and possibly two max modulation rate is NRZ *2 More bandwidth needed pros: synchronization due to mid bit transition no dc component error detection (missing expected transition)
What does the transmission system do in terms of communication tasks?
consists of a single line or complex network - it simply transmits the signal from point a to b (in simple terms)
NRZI encodes signals how?
constant voltage for duration of bit A transition (can be in either direction) denotes a binary 1 no transition at the beginning of a bit time denotes a 0
Protocols
control sending, receiving of messages TCP, IP, HTTP, FTP, PPP
What does a codec do?
converts analog data to digital data
What does the transmitter do in terms of communication tasks?
converts data into transmittable signals
What does the receiver do in terms of communication tasks?
converts received signal into data
What is the difference between data rate and modulation rate?
data rate is 1/T where T is the bit time Modulation rate is the rate at which signals are generated In Biphase Techniques, we have the a smaller signal element than bit duration. Suppose the signal element is 0.5 second, then the bit duration is 1 second. Refer back to how Biphase technique works
What is the signaling rate or modulation rate?
defined as the signal elements per second (band) it is the rate at which signal elements are transmitted (NOT the same thing as data rate, which is in bits per second)
Framing and link access provide what?
encapsulation of packets into frames implementing access to transmission medium
Data link flow control does what?
ensure the sending entity doesn't overwhelm the receiving entity
What is Asynchronous Transfer Mode (ATM)
evolution of frame relay little overhead for error control fixed packet (called cell) length designed to work in the range of 10s and 100s of Mbps
What is frame relay?
evolved from packet switching packet switching systems have large overheads to compensate for errors but modern systems are more reliable and errors can be caught by end systems most overhead for error control is stripped out user data rates up to 2 Mbps
What issues is the Network Access Layer (NAT) concerned with?
exchange of data between end system and attached network issues it is concerned with include: - destination address provision - invoking specific services like priority - access to & routing data across a network link - allows layers above to ignore link specifics
Network access layer (NAT)?
exchange of data between the computer and the network sending computer provides address of destination dependent on type of network used
What are the common type of communication links?
fiber, copper, radio, satellite
In the link layer what is the unit of data?
frame
What are the data link control functions?
frame synchronization framing and link access error detection flow control error control
What does the Source do in terms of communication tasks?
generates data to be transmitted
What is a 2 dimensional bit parity?
given a matrix of bits we send, we can generate parity bits for both rows and columns, it allows us to correct only a single bit error
What is more important to ensure quality for guided vs unguided transmission?
guided - medium of transmission is more important unguided - bandwidth produced by antenna is more important
guided medium vs unguided medium
guided would be something like twisted pair, coax cable, optical fiber unguided medium - waves transmitted but not guided (air, water, vacuum
Increasing bandwidth will __________ data rate
increase
Physical layer is concerned with?
issues including: characteristics of transmission medium signal levels data rates other related matters
What does the sender window look like when it receivers RR(x)
it sets the window at x, x+1, x, ... x+W-1
Based on the idea that we only transmit the narrow band of frequencies where most of the signal energy is, what should the signal bandwidth be?
it should match the transmission bandwidth for the signal to be efficiently transferred
What are the listen, transmit, and bypass states?
listen state - host listens for a packet it will receive, small bit delay, message is copied transmit state - transmit a message to the ring bypass state - the message is completely bypassed, not read
NRZ-L loses synchronization when? NRZI?
long string of ones or long string of zeros NRZI - same cases
channel capacity
max rate (bps) at which data can be transmitted over a given comm path or channel under given conditions
Frame synchronization?
mechanism to synchronize messages at the transmitter and receiver A common way to do this is by sending a start and end flag, we must use bit stuffing to ensure the start and end flags don't show up in the actual message though
Differential Manchester encodes data how?
midbit transition is for the clock only transition at start of bit period means we have a zero no transition at start of bit period means we have a one
Two-level start topology
multiple central hubs that are connected together aka daisy chained
effective bandwidth of signal?
narrow band of frequencies containing most of the signal energy
Why do we need IP and port numbers?
need a way to address each computer (IP), and we need to be able to distinguish each application on a computer (port)
Asymmetric digital Subscriber line (ADSL)
new modem technologies over telephone wire increases speeds of upload and download speeds
Can you correct bit errors with just a single parity bit?
no
NRZ techniques?
nonreturn to Zero-level (NRZ-L) non return to zero inverted (NRZI)
Simplex transmission
one-way transmission
What is the spectrum of signal
range of frequencies contained in a signal
What is error rate
rate at which errors at receiver occurs (bit errors that is)
Selective Reject ARQ
same as go-back-N ARG except SREJ is used instead of REJ, allows only rejected frames to be retransmitted rather than retransmitting all frames sent after the rejected frame
What is a communication protocol?
set of rules governing the exchange of data between two entities
Stop and wait flow control is best for __________ distances, with ______ frames
short, long
What do we take into consideration when encoding/modulating, some data into a signal to send to the receiver?
signal spectrum clocking (ie synchronizing the Tx and Rx) error detection signal interference and noise immunity cost and complexity (certain encoding techniques require higher signal rate than data rate
How do you calculate the net gain or loss in a cascaded transmission?
simply add the gain/loss of each transmission path
WAN characteristics
spread over a large geographical area crossing public rights way rely in part on common carrier circuits consists of a number of interconnected switching nodes
In asynchronous transmission how do we know when transmission starts and stops?
start is signaled with a start bit. We already predetermine how many bits per character are going to be set, so just send those bits, then we get a stop element Stop element is transition from 1 to 0, after stop element there is about 1 to 2 bit timing before the next character can start
Protocol architecture is a?
structured set of modules that implements the communication function
Central hub/node that broadcasts a frame vs a switch?
switch is smarter- knows where the receiver is located and doesn't rebroadcast a message to every node
Synchronous transmission vs asynchronous
synchronous allows for frame synchronization of larger blocks of data, more efficient than asynchronous
What are the key elements of a protocol?
syntax - concerns the forma tof data block semantics - includes control info & error handling timing - includes speed matching & sequencing
What does the destination do in terms of communication tasks?
takes incoming data
A transmission system includes what components?
transmission medium amplifiers or repeaters
Data communications deals with?
transmission of signals in a reliable and efficient manner
List factors/tasks that need to be taken into consideration during communication
transmission system utilization - manage when a device can send a signal addressing - how to get from host a to b interfacing routing signal generation recovery synchronization message formatting exchange management security error detection and correction network management flow control
What is High level data link control (HDLC)
uses both Go-back-N and Selective reject ARQ, flow and error control piggybacked on info frames
LAN spreads over what geographical area?
usually small area, could be 1 building, a small campus, etc
Travel speed for light in vacuum, fiber, and electricity over copper?
vacuum - 300 m /microsec fiber - 200 m / microsec copper - 250 m/microsec
Absolute bandwidth of signal is?
width of signal spectrum
What are the end systems of a computer network?
workstations, servers, smartphones, etc