Telecom Midterm

Front 1

What are the units for transmission rate?

Back 1

Bits/second

Front 2

True or false: All devices connected to the internet are called server systems.

If false, what is the right answer?

Back 2

False: All devices connected to the internet are called hosts or end systems.

Front 3

Home networks are connected to _____________ ISPs and mobile networks are connected to ______________ ISPs.

Back 3

Home networks are connected to global ISPs and mobile networks are connected to regional ISPs.

Front 4

The network by which end systems are connected is made up these two things: ______________ & ________________.

Back 4

Communication links & packet switches.

Front 5

What does a packet switch do with arriving packets (basically, what is the purpose of packet switches/how do they work)?

Back 5

Packet switches take an arriving packet from an incoming link and forward it to its outgoing link.

Front 6

What are the two prominent types of packet switches?

Back 6

Routers (used in network core) and link-layer switches (used in access networks).

Front 7

What is the name of the sequence of switches a packet takes from end-to-end?

Back 7

A route or a path.

Front 8

How many GB of internet are used per month (globally)?

Back 8

1 billion GB.

Front 9

Short Answer: Use a transportation network with trucks as an analogy to describe how packet switch networks work.

Back 9

You subdivide cargo (the packets) into trucks (the loads). Each 'truck' travels independently to destination along roads (links) & intersections (packet switches). At the final warehouse (end system) the 'truck' (packet) is unloaded.

Front 10

What does ISP stand for?

Back 10

Internet Service Provider.

Front 11

_____________ controls the sending & receiving of information of internet components.

Back 11

Protocols

Front 12

What are the two most important types of protocols?

Back 12

TCP & IP

Front 13

What does TCP stand for; what does IP stand for?

Back 13

Transmission Control Protocol; internet protocol.

Front 14

The IETF (Internet Engineering Task Force) develop internet standards, written in documents called __________.

Back 14

RFCs (requests for comments).

Front 15

E-mail, Internet radio and VoIP are all examples of what kind of application?

Back 15

They are all distributed applications.

Front 16

Do internet applications run on end systems or packet switches?

Back 16

Internet applications run on end systems.

Front 17

What are APIs and what do they do?

Back 17

Application Programming Interfaces (APIs) specify how end system software asks the Internet to deliver data to other end systems, and are a set of rules that sending software must follow so that the Internet can deliver data to the destination.

Front 18

In terms of a protocol, what needs to be sent/received before making a get request and getting the requested file?

Back 18

The end system sends a TCP connection request and must receive a TCP connection response before requesting and getting a file.

Front 19

Define a protocol.

Back 19

Defines format and order of messages exchanged between two or more communication entities; governs how all network components communicate and behave with each other in delivering packets.

Front 20

What are the two categories of hosts? Give two examples of each.

Back 20

Clients, such as cellphones, PCs, and;

Servers, such as web pages, stream video.

Front 21

What runs on an end system and requests service from server programs?

Back 21

Client programs.

Front 22

List some programs using the client-server model.

Back 22

E-mail, file transfer, websites.

Front 23

True or false:

Server-side internet applications are distributed applications.

Back 23

False.

Client-side internet applications are distributed applications.

Front 24

What's the main difference between P2P (peer-to-peer) file sharing and client-server relationships?

Back 24

In P2P, end systems act as both clients and servers. For example, BitTorrent.

Front 25

What is an access network?

Back 25

Access networks are the link that connect the end system with the first (aka edge) router.

Front 26

Many access networks use local telco infrastructure where each residence has a twisted-pair copper link directed to where?

Back 26

Residents are linked by twisted-pair copper link to the nearby telco switch house in the central office (CO).

Front 27

What are the two types of broadband access?

Back 27

DSL & Cable

Front 28

What does DSL stand for?

Back 28

Digital Subscriber Line

Front 29

What does DSLAM stand for, where are they located and what do they do?

Back 29

DSLAM (digital subscriber line access multiplexer) are located in the central office and use existing telephone wire to exchange data with the DSL modem.

Front 30

Is there a difference between the frequency of downloading vs uploading (toward the customer computer vs away from the customer computer respectively)?

Back 30

Yes, it is high-speed downstream (downloading, toward the customer computer) at 50kHz to 1 MHz and medium-speed upstream (uploading, away from the customer computer) at 4kHz to 50kHz.

Front 31

What is required to connect a home phone and a PC to the DSLAM at the CO?

Back 31

The phone is connected to a splitter, which is connected to the DLSAM. The PC is connected to the same splitter through a DSL modem.

Front 32

If multiple households are connected to the same DSLAM, what is a possible issue that could arise?

Back 32

May be delays due to processing incoming data.

Front 33

What are two factors affecting upstream/downstream rates?

Back 33

The distance between the CO & home and the gauge of the telephone wire.

Front 34

What broadcasts TV channels, and how is it connected to the neighbourhood-level junction? How many homes can this junction support? What is the name of this network?

Back 34

The cable head end broadcasts TV channels and is connected to the junction through fibre optics. A junction may support 500-5000 homes.

This network is called HFC (hybrid finer coax).

Front 35

True or false:

Upstream & downstream channels are symmetric.

Back 35

False. They are asymmetric (downstream is typically a higher rate than upstream).

Front 36

What does an HFC need to connect to the Internet?

Back 36

A cable modem.

Front 37

Why would a packet channel fe faster when fewer people are around?

Back 37

Packets travel on the same channel and need to coordinate transmissions, this coordination slow with usage.

Front 38

True or False:

Cable provides faster transmission than DSL.

Back 38

True, if the HFC network is reasonably dimensioned.

Front 39

What does it mean for a network to be dimensioned?

Back 39

Dimensioned networks mean that the capacity of the network components are appropriate for the amount of usage. I.e. 500 homes in a junction is less dimensional than a network with 5000 homes.

Front 40

Which is fastest in terms of transmission rate: twister-pair copper wire, coaxial cable or fibre optics?

Back 40

Fiber optics.

Front 41

What does FTTH stand for?

Back 41

Fibre-To-The-Home.

Front 42

True or false: With direct fibre, fibre is shared until split to individual homes.

Back 42

True.

Front 43

What are the two optical-distribution network architectures that perform fibre optic splitting?

Back 43

AONs (active optical networks, a switched Ethernet) & PONs (passive optical networks, used in FIOS).

Front 44

What is an ONT and what does it do?

Back 44

An ONT is an optical network terminator and each home has one. An ONT is connected by optical fibres to the optical splitter.

Front 45

What does a splitter do?

Back 45

A splitter combines < 100 homes into a single fibre.

Front 46

What is an OLT and what does it do?

Back 46

An optical line terminator (OLT) is located at the CO and is connected by a splitter to ONTs at each home by optical fibres.

Front 47

What is a LAN and what does it do? What is the most common type of LAN?

Back 47

A local area network connects end systems to the edge router. The most common type of LAN is ethernet.

Front 48

List and describe the two types of internet access.

Back 48

Wireless LAN allows users to transmit/receive packets from an access point that is connected to the Internet within a reasonable distance to the access point. A wide-area wireless access networks allows packets to be transmitted to a base station over the cellular phone infrastructure.

Front 49

Whats the difference between 3G & 4G?

Back 49

3G (third generation) wireless provides packet-switched wide-area wireless at 1Mbps. 4G (LTE) does his at speeds of 10Mbps.

Front 50

What does LTE stand for, and what's another name for it?

Back 50

LTE, or 4G stands for long-term evolution.

Front 51

Draw a diagram for how a cellphone connects to the internet wirelessly.

Back 51

Front 52

What are the two categories of physical media (copper wire, cable, etc)?

Back 52

Guided media and unguided media.

Front 53

What is the difference between guided media and unguided media?

Back 53

In guided media, waves are guided along a solid media (cable/wire). In unguided media, waves propagate in the atmosphere (LAN/satellite).

Front 54

What is UTP and what is it used for?

Back 54

UTP or unshielded twisted pair is used for LANs.

Front 55

What's the difference between coaxial cable and the twister pair?

Back 55

In coaxial cable, the two copper conductors are concentric; in twister-pair they are parallel.

Front 56

Fibre optics works by conducting pulses of light. How many bits in a pulse?

Back 56

Each pulse is one bit.

Front 57

What type of channel carries signal in the electromagnetic spectrum, requires no physical wire and can penetrate walls?

Back 57

Terrestrial Radio Channels.

Front 58

What is shadow fading? What's multi path fading?

Back 58

Shadow fading, or path loss, is the term to describe how signal strength decreases over distance and obstructing objects. Multi path fading is signal reflection off interfering objects.

Front 59

What are the two types of satellite radio channels and what does a satellite radio channel do?

Back 59

A satellite radio channel links two or more ground stations. The two types are geostationary satellites and low-earth orbiting (LEO) satellites.

Front 60

Are LEOs used to provide internet?

Back 60

Not currently.

Front 61

What are the similarities and difference between packet switching and circuit switching?

Back 61

Both are approaches to moving data. Circuit switching reserve resources in telephone networks which are dedicated for a duration of a session. Packet switches are not reserved but are rather used on demand.

Front 62

Is the internet packet switched or circuit switched?

Back 62

The internet is packet switched.

Front 63

Explain how circuit switching works when a call is made.

Back 63

In circuit switching, there is an end-to-end connection established between hosts. The network resources are divided into pieces, where a piece is allocated as information is passed when a call is made. When a call is made, a circuit is established and the link bandwidth is divided into pieces.

Front 64

Define FDM and TDM and their differences.

Back 64

FDM (frequency-divisor multiplexing) and TDM (time-division multiplexing) are both ways of implementing a circuit. In FDM, frequency is divided amongst connections. In TDM, time frames are divided into slots and each connected is allocated a slot.

Front 65

True or false:

FDM and TDM can be used in tandem.

Back 65

True.

Front 66

Which has better flexibility: TDM or FDM? Latency?

Back 66

TDM has better flexibility but FDM has better latency.

Front 67

Define silent periods and why they are an issue with circuit switching?

Back 67

Silent periods are when dedicate circuits are idle (someone stops using the network but the resource cannot be used by someone else).

Front 68

How long does it take to sent a file of 640,000 bits from host A to host B using circuit switching if the link speeds are 1.536 Mbps, the links use TDM with 24 slots/sec and it takes 500 msec to establish end-to-end circuit.

Back 68

The time for one file is 10 second (because 1.536 Mb/s * 1 channel/24 slots per sec * 10^6 bits/1MB * 1 file/640000 bits = 10 seconds). Also need an addition .5 seconds to establish the connection. Therefore, the answer is 10.5 sec/file.

Front 69

What does store-and forward transmission mean? What is the store-and-forward delay?

Back 69

In packet switching, the entire packet must be received before it can be forwarded which is called store-and-forward transmission. The time spent waiting for the entire packet is called the store-and-forward delay.

Front 70

If a packet is L bits, how long does it take to send it with Q links between two hosts if each link has a rate of R bps?

Back 70

L bits * [1 sec/(R bits/secs per link)] * Q links = LQ/R sec

Front 71

What is an output buffer?

Back 71

And output buffer or an output queue stores packets when an arriving packet finds the link busy.

Front 72

What happens if too many packets are waiting in an output buffer?

Back 72

Buffer space is finite, and if too many packets are waiting they will drop (packet loss).

Front 73

How do routers determine the next hop (forwarding router)?

Back 73

Routers examine the packet destination and determine the next hop using a forwarding table.

Front 74

What tier are internet backbone networks and what are its characteristics? What's the difference between this tier and the others?

Back 74

Tier 1 ISPs are directly connected to other tier 1 ISPs & have international coverage. Tier 2 ISPs have regional and national coverage and are only connected to a few tier 1 ISPs. Lower tier ISPs are connected to tier 2 ISPs.

Front 75

What's a POP?

Back 75

A POP (Point of Presence) is the connection point from one ISP to another.

Front 76

There are five types of packet-switched delays. Name three.

Back 76

Total modal delays, nodal processing delay, queuing delay, transmission delay & propagation delay.

Front 77

What's the difference between a transmission delay and a prorogation delay?

Back 77

Transmission delay is the time required to push all the packets into the link, whereas prorogation delay is the time required to traverse the link itself.

Front 78

What is a service model?

Back 78

Protocols are organized in layers. A service model is the service one layer offers to the layer above.

Front 79

What are the five layers in the protocol stack?

Back 79

From top-to-bottom, the layers are: application, transport, network, linke & physical.

Front 80

What is a datagram?

Back 80

A network-layer packet is called a datagram.

Front 81

Give two types of malware and how they work.

Back 81

Viruses require user interaction (opening an email attachment), worms enter the device without explicit interaction (using a website through which malware is sent) and trojan horse which are hidden in otherwise useful software.

Front 82

What is a DoS attack?

Back 82

A denial of service attack prevents otherwise legitimate use of a network to take place.

Front 83

DoS has three categories. What are they and how do they work?

Back 83

A vulnerability attack sends a few messages to a vulnerable application in order to crash it. Bandwidth flooding sends a lot of packets to a targeted host so the host becomes clogged. Connection flooding establishes a large number of TCP connections to prevent the target from accepting legitimate connections.

Front 84

What is packet sniffing?

Back 84

Packet sniffing is when someone records all packets on a shared media (ethernet, cable, etc).

Front 85

What is IP spoofing? How can you circumvent it?

Back 85

IP spoofing is when you mask the source of an IP address. This can be circumvented using end-point authentication to verify source.

Front 86

What is a man in the middle attack?

Back 86

A man in the middle attack intercepts communication between two entities to inject, modify and delete packets.

Front 87

What is an always-on host?

Back 87

An always-on host is a server.

Front 88

What end systems use dynamic IP addresses, which have permanent IP addresses?

Back 88

Servers have permanent IP addresses, clients have dynamic IP addresses.

Front 89

What's one benefit of P2P? One disadvantage?

Back 89

P2P are highly scalable and don't need an always-on server, but they are difficult to manage due to arbitrary end systems.

Front 90

Who initiates communication: the client process or the server process?

Back 90

Client process.

Front 91

A socket is the interface between which two layers?

Back 91

The socket is the interface between the application layer and the transport layer?

Front 92

Given an example of an application that can tolerate some lot, and an application that require 100% reliable data transfer.

Back 92

Audio can tolerate loss but file transfer required 100% reliable data transfer.

Front 93

What is a bandwidth sensitive application?

Back 93

An application that requires a minimum amount of throughout in order to be effective is bandwidth sensitive.

Front 94

What's the difference between UDP and TCP?

Back 94

TCP (transmission control protocol) is reliable transport & UDP (user datagram protocol is unreliable data transfer). In TCP, there is set up required between the client and server and messages flow through 'handshaking'. With UDP, there is no connection setup, reliability, flow control, limiting, etc.

Front 95

If UDP is unreliable, then why bother using it?

Back 95

UDP has less overhead because it acts like IP.

Front 96

What else do you need other than the IP address of a host in order to identify a process running on it?

Back 96

You need the PID (process ID) & port number as well because there could be more than one process running on the same host.

Front 97

What's the difference between a host name and a path name in a URL? Explain using an example.

Back 97

In the URL www.someschool.edu/someDept/pic.gif ... www.someschool.edu is the host name and /someDept/pic.gif is the path name.

Front 98

True or false:

HTTP is stateless.

Back 98

True, a server maintains no information about past client requests.

Front 99

What's the difference between a persistent HTTP connection and a non-persistent one?

Back 99

In non-persistent, at most one object is sent over the TCP connection and each request/response pair is in a separate TCP connection ('handshake every time').

In a persistent HTTP, multiple objects are sent over a single TCP connection between client & server. The same TCP connection is used for all request/response pairs ('one handshake, send all objects').

Front 100

Total transmission time = transit time + x(RTT).

What is x?

Back 100

2

Front 101

Why would you choose a persistent connection instead of a non-persistent one? Why would you choose a non-persistent one instead of a persistent one?

Back 101

A persistent connection is preferred because it only requires one RTT for all objects whereas a non-persistent connection needs two per object.

A non-persistent connection may be preferred because it doesn't leave the connection 'open'.

Front 102

What does a 301 status code mean?

Back 102

301 Moved Permanently

Front 103

True or False:

Cache acts as a server.

Back 103

False. Cache acts as both a client and a server.

Front 104

Give three reasons why we use web caching.

Back 104

Web caching reduces response time for client requests, reduces traffic on an institution's access link and enables 'poor' content providers to effectively deliver content.

Front 105

How does a Conditional Get work with cache?

Back 105

If the cached copy of the object is up-to-date (if-modified-since: ) then the response from the server will contain no object.

Front 106

If a Conditional Get is used for cache, and the cache is up to date, what is the status code returned by the server?

Back 106

304 Not Modified.

Front 107

What does it mean that FTP maintains 'state'? What's one issue with this?

Back 107

It remembers current directory, earlier authentication, etc. The issue with this is that it may limit the number of simultaneous FTP connections open at once.

Front 108

What are the three major components of e-mail?

Back 108

User agents, mail servers, and simple mail transfer protocol (SMTP).

Front 109

Explain what a user agent is (for e-mail).

Back 109

It is aka the mail reader and is used for composing, editing, reading of mail messages. I.e. Outlook.

Front 110

Is STMP persistent or non-persistent?

Back 110

Persistent.

Front 111

Explain how POP3 protocol works.

Back 111

First, it connects to a mail server on port 110. Then, there is the first authorization phase (where the user enters their username/password). If the authorization goes through OK, next is the transaction phase with the list of message numbers and a retrieval message by number. The final phase is quit where messages marked for delete are deleted.

Front 112

"Download and Keep" means copies of e-mail messages are saved on different ______________.

A. Clients B. Ports C. Hosts

Back 112

"Download and Keep" means copies of e-mail messages are saved on different clients.

Front 113

Which protocol keeps the user state across sessions?

Back 113

IMAP

Front 114

DNS services translate _______________ to _____________.

Back 114

Host names to IP addresses.

Front 115

Which class of DNS server are TDL servers typically cached in?

Back 115

Local name servers.

Front 116

Name the four tuples of RR.

Back 116

Namee, Value, Type, TLL

Front 117

What is TTL and what does it do?

Back 117

TTL is time to live and it determines when a resource should be removed from the cache.

Front 118

If a DNS' is type A, what is its name and value?

Back 118

The name will be a hostname and the value will be an IP address.

Front 119

In terms of P2P, what does it mean to "pull chunks"? In what order are chunks pulled?

Back 119

Since there are no always-on server, at any given time different peers will have different subsets of 'chunks' or pieces of files periodically. Pulling a chunk means asking each neighbour for a list of chunks they have. This is done by asking for missing chunks in order of rarest first.

Front 120

What does it mean to "optimistically unchoke"?

Back 120

In P2P, if you optimistically unchoke, it means that you're checking for other peers every thirty seconds and the newly chosen peer may join the top four peers.

Front 121

What data type are P2P databases stored in?

Back 121

DHTs (distributed hash tables).

Front 122

Define peer churn. How do circular DHTs deal with peer churn?

Back 122

Peer churn refers to the turnover involved in joining and leaving P2P systems. In a circular DHT, each peer knows the IP of its two successors and periodically pings them to check if they're still alive. If its immediate predecessor is dead upon ping, that peer sets the second predecessor to the primary one.

Front 123

True or False:

Skype is P2P.

Back 123

True.

Front 124

What are the two types of transport services via socket API?

Back 124

Unreliable diagram (UDP) and reliable (TCP).

Front 125

Which has no 'connection' / 'pipe' / 'handshaking' between the client and server: UDP or TCP?

Back 125

UDP.

Front 126

Do transport protocols run on the end system or the host?

Back 126

End system.

Front 127

What's the difference between the transport layer and the network layer?

Back 127

The transport layer is the logical communication between the processes running that the end host relies on, whereas the network layer is the logical communication between the end hosts themselves.

Front 128

True or False:

The network layer guarantees delivery, but not orderly delivery.

Back 128

False.

The network layer uses best-effort delivery service and there are no guarantees on delivery, orderly delivery or data integrity.

Front 129

What is multiplexing?

Back 129

Multiplexing means encapsulating chunks at source with header information to be used for demultiplexing, to create segments and passing the segment to the network layer.

Front 130

One reason for using UDP could be that it has an 8 byte header/segment whereas TCP has a 20 byte header/segment.

Back 130

True.

Front 131

What does UDP do to recover from an error?

Back 131

Nothing. It either discards corrupted segments or passes corrupted segments to the application with a warning.

Front 132

What's the purpose of ACKs and NAKs?

Back 132

ACKs are acknowledges to let you know that a packet was received OK and a NAK means it was received with errors.

Front 133

What is ARQ protocol?

Back 133

An ARQ (automatic repeat request) is a reliable transfer protocol based on retransmission based on ACKs.

Front 134

In the "stop and wait" protocol, what does the sender have to receive after sending one packet before sending the next?

Back 134

An ACK.

Front 135

Explain the rdt2.2 protocol.

Back 135

It is a NAK-free protocol, so the receiver of packets just sends back the last packet that had an OK ACK. The receiver must include the sequence number of the packet being ACKed.

Front 136

If rdt3.0 is being used, with a 1 Gbps link, 15 ms prorogation delay and 8000 bit packet, what is the transmission delay?

Back 136

L/R = 8 microseconds

Front 137

True or False:

rdt3.0 uses 'stop and wait' protocol.

Back 137

True.

Front 138

What does it mean if a protocol is using 'pipelining' ?

Back 138

Pipelining means that the sender allows multiple 'in-flight' / 'in progress' / 'yet to be acknowledged' packets.

Front 139

What is the name of the protocol also known as the 'sliding window' and how does it work?

Back 139

The go-back-N protocol means the sender can have up to N unacked packets in the pipeline. The receiver only sends cumulative acks and doesn't ack packet if there's a gap. The sender has a timer for the oldest unacked packet and if it times out, it resends all unacked packets.