The Internet

• When we think of networking, we often think of the Internet. The Internet is, indeed, a network, but there are many other networks besides the Internet.

• The Internet has a long history and did not spring to life fully-formed with the advent of the web in the mid-1990s.

• Despite much mythology to the contrary, the Internet and its predecessors were funded by public research funds for most of the history of the field.

• The Internet has evolved in ways quite different from those envisioned by its creators.

• The Internet was originally designed for experimental flexibility, not as reliable infrastructure.

• The Internet is a hugely complex system and is therefore subject to both emergent behavior and system failures. Fortunately, the designers understood these ideas better than most systems developers.

The Telegraph: The World’s First Digital Network

• Before we talk about the Internet, let’s spend some time talking about the world’s first network: the telegram system.

Invention and Early Development

• 1830s-1840s: Multiple inventors developed electromagnetic telegraph systems simultaneously
  • Samuel Morse (USA) created the most successful system with his Morse Code (1838)
  • William Fothergill Cooke and Charles Wheatstone (UK) developed competing systems
  • Carl Friedrich Gauss and Wilhelm Eduard Weber (Germany) built early experimental lines
• Key Innovation: Converted text into electrical pulses using binary-like encoding (dots and dashes)
• First successful message: “What hath God wrought” sent by Morse from Washington to Baltimore (1844)

Rapid Expansion and Infrastructure

• 1840s-1850s: Telegraph lines spread rapidly across continents
  • Major cities connected by overhead wires supported by telegraph poles
  • Railroad companies adopted telegraph for coordination and safety
  • Government and military recognized strategic importance
• 1851: First submarine cable successfully laid across English Channel
• 1858: First transatlantic cable attempt (failed after 3 weeks)
• 1866: Permanent transatlantic telegraph cable established, connecting Europe and North America

Global Network Formation

• 1860s-1870s: Telegraph became truly worldwide
  • Trans-continental lines: Connected East and West coasts of continents
  • Submarine cables: Linked continents across oceans (Atlantic, Pacific, Indian)
  • Colonial networks: European powers connected their global empires
• Major cable routes established:
  • London to India via Mediterranean and Red Sea
  • Australia connected to Asia and Europe
  • South America linked to North America and Europe

Impact as a Global Phenomenon

• Information Revolution: News could travel faster than any physical transport
  • Journalism transformed: International news became immediate
  • Financial markets: Stock prices and commodity rates synchronized globally
  • Diplomacy accelerated: International negotiations conducted in real-time
• Economic transformation:
  • Global commerce: Coordinated international trade and shipping
  • Time standardization: Led to creation of time zones for coordination
  • Business expansion: Enabled management of far-flung operations

The First Digital Network

• Digital encoding: Messages converted to electrical pulses (binary-like system)
  • Morse Code used combinations of short and long signals
  • Required encoding at transmission and decoding at reception
• Network protocols: Established standardized communication methods
  • Addressing systems for routing messages
  • Error correction and message verification procedures
  • Priority systems for urgent communications
• Infrastructure parallels to modern internet:
  • Switching stations: Early packet routing at telegraph offices
  • Network redundancy: Multiple routes for reliability
  • Global connectivity: Worldwide reach through interconnected nodes

Legacy and Decline

• Peak era: 1870s-1920s dominated long-distance communication
• Gradual replacement: Telephone and radio technology superseded telegraph
• Final chapter: Western Union sent last telegram in 2006
• Lasting impact: Established foundation for all future digital communication networks

Vannevar Bush and Hypertext

• One of the earliest proponents of developing a global information network of the form we have today in the web was Vannevar Bush, President Roosevelt’s Director of the Office of Scientific Research and Development in the United States. In a 1945 article in Atlantic Monthly entitled “As We May Think,” Bush anticipated many of the ideas that are central to the modern Internet, including the idea of hyperlinked documents.

Vannevar Bush and the Memex

• The following quote from Vannevar Bush is from “As We May Think,” 1945:

When data of any sort are placed in storage, they are filed alphabetically or numerically, and information is found (when it is) by tracing it down from subclass to subclass. It can be in only one place, unless duplicates are used; one has to have rules as to which path will locate it, and the rules are cumbersome. Having found one item, moreover, one has to emerge from the system and re-enter on a new path.

…The human mind does not work that way. It operates by association. With one item in its grasp, it snaps instantly to the next that is suggested by the association of thoughts, in accordance with some intricate web of trails carried by the cells of the brain. . . .

…Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and to coin one at random, “memex” will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.

It consists of a desk, and while it can presumably be operated from a distance, it is primarily the piece of furniture at which he works. On the top are slanting translucent screens, on which material can be projected for convenient reading. There is a keyboard, and sets of buttons and levers. Otherwise it looks like an ordinary desk. . . .

….All this is conventional, except for the projection forward of present-day mechanisms and gadgetry. It affords an immediate step, however, to associative indexing, the basic idea of which is a provision whereby any item may be caused at will to select immediately and automatically another. This is the essential feature of the memex. The process of tying two items together is the important thing.

Ted Nelson and Hypertext

• Many of the notions of the modern web were anticipated by Ted Nelson, who introduced the word hypertext in the early 1960s.

• In 1974, Nelson published Computer Lib, a two-sided book that is more reminiscent of a graphic novel than one’s traditional conception of a book. In it, he describes how computers will empower people as technology develops.

Initial ARPANET Configuration (1969)

ARPANET Configuration (1971)

ARPANET Configuration (1975)

World Internet usage over time

Sending a message: a challenge

• Given a map such as the one shown below, how would you design a reliable communication strategy for sending messages between nodes in the network, given the following facts:

  • Nodes are unreliable.
  • Nodes are heterogeneous, in the sense that they come from many different manufacturers, with incompatible instruction sets and word sizes.
  • Connections are unreliable.
  • Connections are slow.
  • Network traffic is bursty, in the sense that there are usually delays between messages that are themselves relatively long.

Routing: How does each node know where to send each packet?

• Given our graph of nodes on the early Internet, how did the nodes know where to send their packets?
  • Each node kept track of its own routing table, which keeps track of how many “hops” it is to get to all nodes in the network, and which node it should send a packet to so it heads towards the destination node with the shortest number of hops.
  • Nodes are continually sending the current state of their routing tables to their neighbors, who update their tables. If a node or link goes down, a table will get updated at nodes that connect to the down (the neighbor node that can’t be reached will be dropped from a table). This gets propogated through to other nodes to update the state of the entire network.

Demo: Early Internet Routing

Today: IP Addresses

• Today, every computer has an IP Address that is used to identify where on the network it resides. Most computers and routers use a IPV4, which is a 32-bit addressing scheme, such as 172.64.15.16. Each octet of the number represents a byte.

• How many computers can be represented on the Internet in this way?

  Show the number of computers that can have IPv4 addresses

  $2^{32} = 4,294,967,296$

  • Guess what? There are many more than 4+ billion computers on the Internet, if we include routers, phones, IOT devices, etc.).

  • The way we handle this is by having private networks that are not accessible direcly on the web, and then by having routers translate between the private and public Internet. If you look at your own computer’s IP address, it is likely one of the following forms: 192.168.x.x or 10.x.x.x or 172.16.x.x.

• The Internet is slowly transitioning over to IPv6, which has 128-bit addressing, which allows for $340,282,366,920,938,463,463,374,607,431,768,211,456$ different computers, which is more than humans will ever have, certainly. If we added 10 billion addresses per year, it would take 10 trillion trillion times the age of the universe to use them all up. I think we’re safe from running out.

• You can (sometimes) use the following command to find out how many hops it will take to get from your computer to somewhere else on the Internet:

% ping www.pku.edu.cn
PING www.lb.pku.edu.cn (162.105.131.160): 56 data bytes
% sudo traceroute 162.105.131.160
Password:
traceroute to 162.105.131.160 (162.105.131.160), 64 hops max, 40 byte packets
 1  woa-srtr-vl615.sunet (10.108.40.1)  2.163 ms  0.726 ms  0.712 ms
 2  campus-nw-rtr-ae0-vl3806.sunet (192.168.188.25)  1.743 ms  1.038 ms  1.158 ms
 3  xb-nw-rtr-vl1005.sunet (171.64.255.203)  1.766 ms  1.164 ms  1.103 ms
 4  hpr-svl-rtr-vlan2.sunet (171.64.255.147)  1.833 ms  1.764 ms
    hpr-svl-rtr-vlan3.sunet (171.66.255.147)  1.680 ms
 5  hpr-svl-hpr3--stan-100ge.cenic.net (137.164.27.60)  3.091 ms  3.083 ms  3.188 ms
 6  hpr-internet2r&e--svl-agg10-100ge.cenic.net (137.164.26.127)  6.159 ms  6.730 ms  5.689 ms
 7  fourhundredge-0-0-0-1.4079.core1.sacr.net.internet2.edu (163.253.1.192)  25.098 ms  23.697 ms  24.373 ms
 8  fourhundredge-0-0-0-0.4079.core1.seat.net.internet2.edu (163.253.2.167)  23.865 ms  23.455 ms  24.364 ms
 9  fourhundredge-0-0-0-17.4079.core2.seat.net.internet2.edu (163.253.2.53)  25.084 ms  22.715 ms
    fourhundredge-0-0-0-16.4079.core2.seat.net.internet2.edu (163.253.2.55)  25.920 ms
10  tyo-mx2010-et0-1-1-v1468.jp.apan.net (203.181.248.142)  120.510 ms  119.736 ms  119.385 ms
11  hk-mx480-1j-et-0-3-0-v137.jp.apan.net (203.181.248.121)  172.609 ms  171.946 ms  171.645 ms
12  cngi-6ix.jp.apan.net (203.181.248.126)  181.540 ms  180.044 ms  180.150 ms
13  210.25.187.49 (210.25.187.49)  180.684 ms
    210.25.187.53 (210.25.187.53)  181.244 ms
    210.25.187.49 (210.25.187.49)  179.574 ms
14  210.25.189.238 (210.25.189.238)  180.203 ms  179.117 ms  178.925 ms
15  * * *
16  101.4.116.94 (101.4.116.94)  180.017 ms  182.201 ms  179.381 ms
17  * * *
18  202.112.41.181 (202.112.41.181)  180.544 ms  179.545 ms  179.513 ms
19  202.112.41.2 (202.112.41.2)  181.501 ms  180.817 ms  180.315 ms
20  * * *
21  162.105.253.58 (162.105.253.58)  181.197 ms  180.040 ms  180.254 ms
22  * * *
23  * * *
24  * * *

• This didn’t quite make it (for various reasons) to the main Peking University webpage at https://www.pku.edu.cn, but it got pretty close (it made it to a site at the university).

How do computers communicate across the oceans?

• There are many undersea cables that stretch across the ocean, on the sea floor. They cost millions of dollars to lay down, but carry many telephone and Internet communications.
• The first submarine cable was laid in 1854 and went from Ireland to Canada.
• The first official telegram message was from Queen Victoria of the United Kingdom to President of the United States.
• The connection had technical difficulties, and signals took hours to send (Queen Victoria’s message of 98 words took 16 hours to send!). This was because electrical properties of long cables was not understood well at that time.
• The cable only made it three weeks before it was destroyed.
• The technology was slowly improved, and many telegraph cables were laid down on the ocean floor over the next decades.
• Today, cables transmit almost all of the data between countries separated by oceans.
• Submarine cables can transmit incredible amounts of data: the MAREA cable between North America and Europe can handle up to 200 terabits per second (Tbps).

What about satellite communications – no need for thousands of miles of cables!

• Satellites have allowed communications between nations for decades, but there are significant issues with using them to transmit large amounts of data.
  • It is extremely expensive to put satellites into orbit, and to keep them funcitoning.
  • They are limited in bandwidth – some satellites can transmit up to 1 gigabit of data per second, but that is a thousand times less than the MAREA cable from above.
  • Many satellites are in geosynchronous orbit, meaning that they orbit above a fixed location above the Earth. But, to get into that orbit, they have to be placed 35,786 km above the Earth! At that distance, they orbit in 24 hours, which means that if they are placed above the equator, they will stay in the same position.
  • The problem with communicating via geosynchronous satellites is that the latency of the signal is high. It takes approximately $\frac{1}{8}$ of a second for a signal to travel 35,786 km, and then another $\frac{1}{8}$ of a second for a signal to travel back to Earth. This is because of the speed of light – nothing can travel faster, and that’s how fast radio signals travel! In practice, the latency is actually more like 500-600ms (half a second or more) because of additional processing, and the fact that a signal is not beamed straight up or straight down from a satellite.
  • The latency means that the delay for a signal to be sent and received is $0.25s$, which is quite challening to deal with when talking between two people, and also terrible for communications that rely on lower latency (e.g., playing video games). You don’t want your communications to go through a geosynchronous satellite if you can help it!
  • There are many communications satellites in low earth orbit (~550km for Starlink, for instance), which do not suffer from the latency issues discussed above. But, because teh satellites are not fixed above the earth, constant transfer between satellites is necessary for a signal to be sent, which is difficult and expensive to achieve.
  • Starlink is aiming to have 42,000 satellites in orbit eventually, which is very expensive, but it will allow for more communications from anywhere in the world.