Computer Networking - Notes on Introduction

Here are my notes on Chapter 1 of Computer Networking: A Top Down Approach by Kurose and Ross. I’m going to bold their bolded terminologies that might be useful though in the intro a lot of these are basically common sense.


  • Hosts or end systems are the devices connected by networks
  • Est. 2 Billion Internet users
  • Communication links and packet switches connect the end systems together.
  • Transmission rate of a link is measured in bits/second
  • ** packets** are the segmented chunks of data sent through networks.
  • Two most prominent packet switches are routers and link-layer switches.
    • Link layer switches typicall used in access networks, while routers are typically used in the network core.
  • End systems are connected by ISPs including residential ISPs (telecom or cable companies), corporate ISPs, university ISPs, etc.
    • Lower tier ISPs are connected by national or internation upper-tier ISPs such as Level 3, ATT, Sprint, Pier 1
  • Principal protocols known as TCP/IP
    • Defined in the interenet standards in RFCs (Request for comments) by the Internet Engineering Task Force (IETF)
  • A protocol defines the format and the order of messages exchanged between two or more communicating entities, as well as the actions taken on the trans- mission and/or receipt of a message or other event.
  • two must prevalent types of residential access are digital subscriber line (DSL) and cable.
  • DSL typically obtained from telco and uses existing phone line (twisted pair copper wire) to exchange data with a digital subscriber line access multiplexer (DSLAM) located in the telco’s local central office (C).
    • DSL modem takes digital data and translates it to high-frequency tones for transmission over telephone wires to the CO; the analog signals are converted back to digital by the DSLAM
    • phone line carries telephone and data signals simultaneously, which are encoded at different frequencies
      • a high speed downstream channel, in the 50 kHz to 1 MHz band
      • A medium speed upstream channel, in the 4 kHz to 50 kHz band
      • An ordinary two-way telephone channel, in the 0 to 4 kHz band
    • On the customer side, a splitter separates the data and telephone signals arriving to the home and forward the data signal to the DSL modem
    • For telco, the DSLAM separates the data and phone signals and forwards data on.
    • Hundres or thousands of households per DSLAM
    • DSL define transmission rate of 24 Mbps down and 2.5 Mb up most recently (2003). Because they’re different, this said to be asymmetric access
    • DSL is designed for distances of only 5-10 miles between CO and residence
  • cable internet access makes use of television infrastructure.
    • Fiber optics connect the cable head end to the neighborhood-level junctions, from which traditional coaxial cable is used to reach individual houses and apartments.
    • Each neighborhood junction supports 500-5000 homes
    • Because both fiber and coaxial are used, often referred to as hybrid fiber coax (HFC)
    • Need a special cable modem on teh customer end and a cable modem termination system (CMTS), similar to a DSLAM
    • Also an asymmetric system
  • New tech fiber to the home (FTTH) connects optical fiber path from the CO to the home direct
    • Most common method is by having one fiber reach neighborhood and then split into various fibers
      • Do this through active optical networks (AONs) and **passive optical networks (PONs).
      • AON is essentially switched Ethernet (see Ch. 5)
      • PON is an optical network terminator (ONT) which is connected by dedicated optical fiber to a neighborhood splitter, the splitter combines a number of homes which connects to an Optical line terminator (OLT) in the CO. OLT provides conversion between optical and electrical signals, connects to the internet via a telco router.
  • LAN is like your wireless router at home
  • WAN is like 3G, LTE, 4G

Physical mediums

The mediums by which we propogate the signals.

Twisted Pair Copper Wire

  • Least expensive medium
  • Two insulated copper wires, about 1mm thick, arranged in a regular spiral
    • Twisted to reduce electrical interference from similar pairs close by
    • Bundled together in a cable by wrapping the pairs in a protective shield.
  • A Wire pair is a single comm. link
  • Unshielded twisted pairs (UTP) are commonly used for LANs
  • DSL and phone lines use this more

Coaxial Cable

  • This is what used to/is used for cable (the wire with the one thin wire sticking out the end.
  • Consists of two copper conductors in a concentric formation
  • can be shared among many systems

Fiber Optics

  • Conducts pulses of light, each representing a bit
  • High transmission rate (in the Gb)
  • Low signal attenuation up to 100 km and hard to tap
  • high cost

Terrestrial Radio Channels

  • Radio channels carry signals in the em spectrum
  • penetrate walls, no physical wire, mobile, long distance carriers
  • Environmental conditions determine path loss and shadow fading and interference
  • used for LANs, etc.

Satellite Radio Channels

  • A comm. satellite links two or more Earth-based microwave transmitter/receivers known as ground stations
  • Two types of satellites are used in comms:
    • geostationary
      • 280 ms propagation delay
    • low earth orbit
      • Move around earth and need a bunch to make a full coverage network

Network Core

Store and Forward

  • Most packet switches use $$store-and-forward transmission** at the inputs to the links.
    • This means the packet switch must receive the entire packet before it can begin to transmit the first bit of the packet onto the outbound link.
  • Total delay:
  • Where N links each with rate R and each packet consisting of L bits.

Queueing Delays and Packet Loss

  • For each attached link a packet switch has an output buffer or output queue which stores packets that the router is about to send into the link.
  • So, not only do packets suffer from store-and-forward delays but they also suffer from queueing delays since the amount of buffer space is finite.
    • If the buffer is completely full with other packets waiting for transmission packet loss will occur

Forwarding Tables and Routing Protocols

  • Each packet-switch (router) has a forwarding table that maps destination addresses (or portions of the IP) to that router’s outbound links
  • Packet arrives, router looks up address, sends on proper link
  • Routing protocols are used to automatically set the forwarding tables

Circuit Switching

  • A circuit in a link is implemented with either frequency-division multiplexing (FDM) or time-division multiplexing (TDM)
    • With FDM, the frequency spectrum of a link is divided up among the connections established across the link.
    • FM radio stations also use FDM to share the frequency spectrum between 88 MHz and 108 MHz, with each station being allocated a specific frequency band.
    • TDM link, time is divided into frames of fixed duration, and each frame is divided into a fixed number of time slots. When the network establishes a connec- tion across a link, the network dedicates one time slot in every frame to this connec- tion. These slots are dedicated for the sole use of that connection, with one time slot available for use (in every frame) to transmit the connection’s data.
    • prponents of packet switching argue circuit switching is wasteful because if for example someone is connected on the phone but completely silent it still needs to transmit this data, or if you need to send data then you need to establish a connection with all the overhead even if the user sends like 1 bit.

Network of networks

  • tier 1 ISPs interconnect regional ISPs which interconnect smaller access ISPs, this is a simplified model but fairly representative
    • You also have:
      • PoPs (Points of presence)
        • A group of one or more routers (at the same location) in the provider’s network where customer ISPs can connect into the provider ISP. A customer network can connect to a PoP via a leased telecom cable
      • ** Multihome**
        • Where any ISP except for a tier 1 ISP can connect two or more provider ISPs.
        • If an ISP multihomes, it can continue to send/receive packets ecen if one of its providers has a failure
  • Pairs of nearby ISPs at the same level of the hierarchy can peer together to avoid having to use higher tier ISPs to network
  • Third party companies can create Internet Exchange Points (IXP) which is a meeting point for multiple ISPs to peer together at
  • Content Provider Networks are basically companies like Google which interface with all of these via private lines and peer with their own global datacenters to bypass paying tier 1 networks for a lot of bandwidth. See below figure.

Delay, Loss, and Throughput in Packet-Switched Networks


  • At each node along a traveling path, a packet suffers from:
    • nodal processing delay
    • queueing delay
    • transmission delay
    • propagation delay
  • Which give total nodal delay
  • Nodal processing delay
    • The time required to examine the packet’s header and determine where to direct the packet among other factors like verifying the check bits
      • In high-speed routers are typically on the order of microseconds or less
  • Queueing Delay
    • The time a packet waits in a queue to be transmitted
      • variable according to traffic, but can be between microseconds to milliseconds
  • ** Transmission Delay** is the amount of time to transmit all of the packet’s bits into the link and is governed by $L/R$ where L is the number of bits in the packet and R is the bandwidth ($bits/second$)
  • ** Propagation Delay** is the delay it takes to propogate from the beginning of the link to the end
    • depends on phyiscal medium and is in the range of
      • $2 * 10^8 \text{meters/sec}$ to $3*10^8 \text{meters/sec} \le \text{speed of light}$
    • $\text{delay} = \text{distance of link}/\text{propagation speed}$

Queueing Delay and Packet Loss

  • Let’s first declare some variables
    • $a$ denotes the average rate at which packets arrive at a queue (packets/s)
    • $R$ is the transmission rate (bits/s) pushed out of the queue
    • $L$ number of bits per packet (assumed for simplicity to be distributed equally)
  • Average rate of arrivate is $La \text{ bits/second}
  • If you assume that the queue is infinitely big, $La/R$ is the traffic intensity
    • If the traffic intensity is greater than 1 ($La/R \gt 1$), then the average rate that bits arrive in the queue exceeds transmission rate and the queueing delay will approach infinite
    • So, design the system so that $La/R \le 1$
  • Here you can see the average queueing delay as a function of traffic intensity (exponential approaching 1)

Packet Loss

  • Realistically a queue has a finite length so rather than the queueing delay going up, you actually lose packets because the queue drops them


  • I mean, it’s traceroute… Special packets are sent which cause routers to send back info on themselves to the program. Just do

or something in the terminal and you can see it.

Other delays

  • There can be purposeful delay because of protocols like in VoIP where you have to fill a packet with encoded voice data.

Throughput in Computer Networks

  • instantaneous throughput is the rate at which the receiver of a file or something is actually receiving that data on the other end (think torrents down rate)
  • average throughput of a file transfer is $F/T \text{bits/sec}$ where $F$ is the number of bits in a file and $T$ is the seconds it takes to transfer a file
  • In an ideal scenario the throughput of a network is the transmission rate of the bottleneck link

Protocol layers and their service models

Protocol Layering

  • Each layer has a service model in that it uses the services of the layer below it and provides services to the layer above it
  • Drawbacks of layer: duplication of functionality (i.e. in error handling)
  • Protocol stack consists of 5 layers from the top down:
    • Application Layer is where network applications and their protocols reside. HTTP, SMTP, FTP, DNS
      • pass messages
    • Transport layer transports application-layer messages between application endpoints.
      • TCP & UDP
      • Each transport-layer packet is referred to as a segment in this book
    • Networking Layer is responsible for moving datagrams from one host to another. Essentially this is what does the routing and navigation of sorts
      • IP Protocol
    • ** Link Layer** is the protocol by which delivery is made (Ethernet, WiFi, and cable access network’s DOCSIS protocol)
      • Link layer packets known as frames (at least in this book)
    • Physical layer moves individual bits from one node to the next
      • Twisted copper wire, etc.
  • Each layer encapsulates the message with its own headers. So each packet on every layer had header fields and a payload field comprised of the packet from the layer above.
Written on October 9, 2014