Chapter 1: The Internet

Question 1

How do Digital Subscriber Lines work?

Answer 1

Internet access is provided by the same telco company as the telephone access to the home.
The signal is sent from the modem to a splitter, which uses the phone line to send it to the DSLAM at the central office.

Question 2

How does cable access work?

Answer 2

It uses the existing television HFC (Hybrid Fibre Coaxial) cables. The signal is sent from the modem to a splitter, which uses the phone line to send it to a CMTS (Cable Modem Termination System). It is shared access.

Question 3

What does asymmetric access mean?

Answer 3

The downstream rate is different from the upstream rate.

Question 4

Give an advantage of TSL and one of cable access.

Answer 4

TSL:
- Generally larger upload bandwidth
- Cheaper
- Large number of downloading users is unaffected.

Cable:
- Much faster download speeds
- less latency

Question 5

What is the relationship between packet transmission delay (Dt), length per packet (L) and transmission rate (R)?

Answer 5

R = L/Dt

Question 6

What is the difference between guided and unguided media? Give an example of each.

Answer 6

Guided: Signals propagate in solid media, such as copper wire, coaxial cables and fibre optics.

Unguided: Signals propagate freely, such as radio signals.

Question 7

What are the three main types of cables?

Answer 7

Unshielded Twisted Pair copper cables, coaxial cables and fibre optic cables.

Question 8

Give a description and use of unshielded twisted pair cables.

Answer 8

Two insulated 1mm copper wired are spiralled to reduce electrical interference. It is used as the dominant solution of high-speed LANing and used in DSL technology.

Question 9

Give a description and use of coaxial cables.

Answer 9

Two concentric copper conductors, with shielding on the outside. Common in cable television systems, and used as a guided shared medium.

Question 10

Give a description and use of fibre optics.

Answer 10

A thin flexible medium that conducts pulses of light, with each pulse representing a bit. It supports incredibly high bit rates, is immune to electromagnetic interference, have low signal attenuation and very hard to tap - meaning it is used in overseas links and the backbone of the internet. It is too expensive to be used in all short-haul signal transfer.

Question 11

Give a description and use of radio signals.

Answer 11

An EM wave that can carry signals for long distances. Personal devices like wireless headsets, LAN technologies and cellular access technologies.

Question 12

What types of satellites are used in radio communications?

Answer 12

Geostationary and Low-Earth Orbiting.

Question 13

What is store-and-forward transmission?

Answer 13

A type of packet switching where the packet switch must receive the entire packet before it can begin transmitting.

Question 14

In store-and-forward transmission, what is the total transmission delay between a source and destination, if the packet length is L bits and the transmission rate is R bits/sec? Assume there is no propagation delay.

Answer 14

Source -> Switch is L/R sec.
Switch -> Dest is L/R sec.

Total = 2L/R secs.

Question 15

In store-and-forward transmission, if there are P packets and N links, what is the end-to-end delay for a packet length of L and transmission rate of R?
Assume there is no propagation delay.

Answer 15

Each packet takes L/R to travel across each link, so NL/R total delay in transmission.

Packet P is released once P-1 has been sent from router 1, so after (P-1)L/R.

In total, it takes NL/R + (P-1)L/R = (P+N-1)L/R.

Question 16

Explain how queuing delays arise in switches.

Answer 16

For each link, the switch has an output buffer, which stores the packets to be sent to that link. When packets arrive, they must wait in the queue, causing a queuing delay.

Question 17

Summarise the method of forwarding packets across the Internet.

Answer 17

Each packet has a destination IP address, and each router has a forwarding table that maps destination IP addresses to the router’s outbound links. The next router replicates this, iteratively honing in on the destination.

Question 18

What is circuit switching?

Answer 18

Resources are reserved for the duration of the communication session between the end systems, analogous to a restaurant that requires booking in advance. Commonly used in traditional telephone networks.

Question 19

Compare Frequency Division Multiplexing (FDM) and Time Division Multiplexing (TDM).

Answer 19

Both used in circuit switching to divide up resources.

In FDM, the frequency spectrum of a link is divided up, with the width of the band being termed the bandwidth. FM radio stations share the radio spectrum like this.

In TDM, time is divided up into frames, and frames subdivided into time slots. Each connection can use one time slot in each frame to send data.

Question 20

Give an advantage and a disadvantage of circuit switching over packet switching.

Answer 20

Advantage: In busy networks, data is transferred without unreliable latency, so useful for calls.

Disadvantage: Silent periods are wasteful, as no other person can use the connection.
Complicated to coordinate the switches across the network.
Less efficient.

Question 21

How long does it take to send a file of 640,000 bits over a circuit-switched network with:
- TDM with 24 slots
- Bit rate of 1.536 Mbps
- 500msec to establish an end-to-end circuit?

Answer 21

We can use 1 slot, which will have a transmission rate of 1536000/24 = 64000 bits per second.
Thus, it takes 640,000/64000 + 0.5 = 10.5 seconds to send the file.

Question 22

What is the role of IXPs in the Internet?

Answer 22

They act as a method of connections between competing ISPs.

Question 23

What is the role of regional ISPs in the Internet?

Answer 23

They connect nets of access ISPs to a tier-1 ISPs.

Question 24

What is the role of a content provider in the Internet?

Answer 24

They are private networks that connects to the Internet, often bypassing Tier-1 ISPs.

Question 25

What is the role of Points-of-Presence in the Internet?

Answer 25

They connect up two ISPs to reduce the number of intermediate links needed to transmit between them.

Question 26

What are the four main delays occurring in packet-switched networks?

Answer 26

Nodal processing delay
Queuing delay
Transmission delay
Propagation delay

Question 27

Define nodal processing delay.

Answer 27

The time required to examine the packet header and determine where to direct the packet.

[May also include error checking, and is normally in microseconds]

Question 28

Define propagation delay.

Answer 28

The time taken for a packet to be transmitted across a wire. The speed is dependent on the physical medium, and is just less than the speed of light.

Propagation delay = distance/speed_in_medium

[Generally milliseconds.]

Question 29

Define traffic intensity and explain what happens if it is more than 1.

Answer 29

The ratio between the rate of bit arrival and the transmission rate:

Ti = (L * a)/R

where:
- Ti is the traffic intensity in packets
- L/R is the transmission delay in sec
- a is the average packet arrival rate in packets/sec.

When Ti > 1, then packets arrive faster than they can leave, so the queuing delay tends to infinity.

Question 30

Assume the traffic intensity is less than or equal to 1. If N packets arrive every (L/R)N seconds, what is the average queuing delay?

Answer 30

1st packet = 0 secs
2nd packet = L/R secs
3rd packet = 2L/R secs
:
nth packet = (n-1)L/R secs

Then, average is L/R* (1 + … + n-1)/n = L/R*(n-1)(n-2)/2n secs

Question 31

If there are N-1 routers between the source host and the destination host on an uncongested network, what is the total end-to-end delay?

Answer 31

d = N(d_proc + d_trans + d_prop)

Question 32

How does Traceroute work?

Answer 32

Three packets are send with TTL 1, causing the first link to return a message with it’s IP address. By timing this, we can average the RTT to that node.
The TTL is incremented and resent.

If asterisks appear, there was no response.

Question 32

If Server -> Router -> Client is a simple network with end-to-end throughputs of Rs and Rc for each hop respectively, what is the throughput?

Answer 32

min(Rs, Rc).

[In other words, it is the bottleneck link’s throughput].

Question 33

What is meant by throughput?

Answer 33

The rate at which the file is received. This can be instantaneous or average (file_size/transfer_time).

Question 34

You are downloading an MP3 file of 32 million bits from a server with transmission rate of 2 Mbps and your access rate being 1 Mbps.
How long does the download take?

[This is purely an estimate; there’s many other delays that would exist.]

Answer 34

The transmission rate of the server is irrelevant, being limited by the 1 Mbps. Thus, we have 32,000,000/1,000,000 = 32 seconds.

Question 35

What is the formula for the bandwidth-delay product? (Think the total amount of data in the connection)

Answer 35

Bandwidth-delay = bandwidth * RTT.

Question 36

What layers are present in the network stack?

Answer 36

Physical, Link, Network, Transport, Application

Question 37

What is packet encapsulation?

Answer 37

When data is moved down a layer, it is placed inside the header of the lower layer in turn, creating a single packet with a header and payload. As the packet moves down the stack, it is more and more encapsulated.