1

(PAPER 3 NOTES BELOW)

Separate document with just A Level content here: CompSci Paper 3 Notes
Notes in bold and italic have not yet been asked on a paper

Paper 1

1.​ Information Representation

1.1 Data Representation

Benefits of BCD

●​ more straightforward to convert between BCD and denary

●​ less complex to encode and decode for programmers
●​ easier for digital equipment to use BCD to display information
●​ can represent monetary values exactly

Applications of BCD

●​ electronic displays e.g. calculators, digital clocks – only need to show individual
digits, conversion between denary and BCD is easier
●​ storage of date and time in BIOS of PC – conversion with denary is easier

Applications of Hexadecimal

●​ MAC addresses
●​ HTML colour codes
●​ can also be used to refer to memory addresses in assembly language and in
machine code

ASCII Representation

●​ each character has a unique code

●​ character is replaced by its corresponding code
●​ codes are stored in the same order as in the word

Character Set (e.g. ASCII, UNICODE, Extended ASCII)

●​ All of the characters that the computer can represent/use

●​ Each character has a corresponding unique binary number

Similarities
 2

●​ All can use 8 bits

●​ ASCII is a subset of Unicode/EA
●​ Each represent characters using a unique code

Differences

●​ Unicode can represent multiple languages and a wider range of characters than
ASCII
●​ ASCII – 7bit, Extended ASCII – 8bit, UNICODE – 16bit

1.2 Multimedia

Graphics **

Bitmap Graphic

●​ made up of pixels, each of a single colour (with each colour having a unique binary
value)
●​ stored as sequence of binary numbers (store binary value of each pixel)
●​ prone to pixelation when enlarged
●​ larger file size – data is stored about each pixel
●​ can be compressed significantly
●​ more difficult to edit – each pixel needs to be edited separately

Pixel

●​ The smallest addressable element in an image

File header

●​ Stores data (metadata) about the bitmap image (e.g. colour depth, image
resolution)
●​ Can store: file type, compression type, dimensions (width x height), file size

Image Resolution

●​ Total number of pixels in an image (number of pixels wide * number of pixels high)
●​ Increasing resolution – more pixels can be stored, image is sharper/less pixelated

Bit depth/Colour depth

●​ The number of bits used to represent each colour // number of bits per pixel
●​ Determines number of colours that can be represented in an image
 3

●​ Increase in bit depth – image has greater range of colours and is closer to original
(leads to increased file size)

Vector Graphic

●​ stores set of instructions about how to draw the shape

●​ does not pixelate when scaled/enlarged
●​ individual components of image can be edited
●​ smaller file size – contains just instructions
●​ do not compress well – has little redundant data

Drawing Object

●​ component of vector graphic created using a formula/command

Drawing Property

●​ contains data about the shapes

●​ defines aspect of appearance of a drawing object

Drawing List

●​ the list of shapes that make up an image // list of all drawing objects in image
●​ stores commands required to draw each object

​ Vector Graphic Representation

●​ encoded as series of geometric shapes

●​ stored coordinates of drawing objects in image
●​ contains drawing list – commands for creating each individual object (and their
attributes)

Sound **

Sound Representation

●​ amplitude is recorded a set number of times in a second

●​ each amplitude is given a corresponding unique binary value/number
●​ binary number is saved/stored in sequence

Sampling
 4

●​ taking measurements at regular intervals and storing the values

Sampling Rate

●​ number of samples taken per unit time // per second

Effect of increasing sampling rate

●​ sound is recorded more often // smaller gaps in sound wave & between samples
●​ reduces quantisation errors
●​ improves accuracy – digital waveform resembles analogue one more closely
●​ increases file size (increases number of total samples taken – so more bits needed
to store the data, takes up more space in memory)

Sampling Resolution

●​ number of bits used to store each sample

Effect of increasing sampling resolution

●​ increases number of bits used to store each sample // more bits per sample
●​ wider range of amplitudes can be stored // more amplitudes can be represented
●​ file size increases
●​ digital waveform is closer to original (improves accuracy)
●​ smaller quantisation errors

Analogue data

●​ A variable // data value that is constantly changing

1.3 Compression

Compression Reasons

●​ reduces file size – takes up less space in memory (leaves space for other files,
allows more images/files to be stored)
●​ faster download/upload rate (reduced transmission time) from/to web
●​ less bandwidth used to download/transmit
●​ original might be too large to send through email/as attachment

Lossy Compression

●​ Original data is lost/deleted (cannot be done in text file) & file cannot be
reconstructed
 5

●​ If used on text file, it would corrupt (all data is required, otherwise it would not
make sense)
●​ Videos have a lower resolution, will buffer less if being streamed real-time, lowers
bandwidth required to transmit
●​ Used when all data is not required, quality can be reduced without user noticing or
if a significant reduction in file size is needed

Lossless Compression

●​ Original data is preserved

●​ Used when all data is needed (e.g. file needs to be fully restored), a high quality
video/image is needed, or it is already a relatively small file

Compressing Files

Sound **

●​ reduces amplitude range (to only range used) – reduces bits needed to store each
sample
●​ run-length-encoding – consecutive sounds grouped (binary value of sound
recorded along with number of times it repeats)
●​ record only changes in sound (not actual sounds)
●​ RLE – consecutive sounds grouped (binary value of sound recorded along with
number of times it repeats)

Images

(Lossy)

●​ reduce bit depth – reduces number of bits used to store a colour (each pixel has
fewer bits)
●​ reduce number of colours – fewer bits needed to store each colour
●​ reduce resolution – fewer pixels altogether (less binary to store)

(Lossless)

●​ RLE – replaces sequences of the same colour pixel with colour code and number of
identical pixels

Run-length Encoding **

●​ Identifies groups/sequences of repeated characters and replaces them with a copy

of the character and number of times it occurred
 6

RLE Limitations

●​ Works by storing a colour and the number of times it occurs consecutively – there
may not be many sequences of same colour
●​ It would store each colour then the count, which would be 1 – only adds to file size
// adds to total data
●​ E.g. RBG would become R1 B1 G1
 7

2.​ Communication

2.1 Networks including the internet

LAN

allows…

●​ communication/sharing of data between devices on the network

●​ sharing of resources e.g. hardware/software applications
●​ central management (security, backup etc.)

●​ covers small geographical area, connection between devices is usually physical

●​ infrastructure is privately owned (not controlled by external organisations)
PRIVATE OWNERSHIP
●​ high data transfer rate
●​ protection is easier to implement, more secure than WAN

WAN

●​ covers a large geographical area, connection is often virtual

●​ low data transfer rate
●​ can have private or public ownership (likely to be controlled by external
organisations

Client-Server Model

●​ web pages/data saved on servers

●​ client sends request
●​ web servers process requests/performs requested tasks and returns results to
the client
●​ client displays result to user

●​ user’s computer is the client

●​ server can host shared files
●​ users can request a file from any client computer
●​ files can be accessed simultaneously by several users

Examples of Client-Server Models

 8

●​ sending and receiving an email

●​ using a print or file server
●​ a company/school centrally storing files

Features of a peer-to-peer network

●​ computers are of equal status

●​ each computer provides access to data and resources // data is distributed
●​ computers can communicate and share resources
●​ each computer responsible for its own security

Drawbacks of a peer-to-peer network

●​ reduced security (no central management of security) – each computer is at risk

from viruses from other computers
●​ no central management of back-up – if data from one computer is not backed up it
is lost to all of them
●​ no central management of files – hard to maintain consistency
●​ computers have slower response time – due to being accessed by other computers
●​ files may not always be available – not all computers are always switched on

Thick-Client

●​ server performs minimal processing for client

●​ most resources are installed locally – clients do most of their processing
independently

Thin-Client

●​ server performs all processes required by task/data storage

●​ clients only send requests to server and displays returned results

Star Topology

●​ devices only connected to central router/device (switch/hub etc.)

●​ each computer only connected to server
●​ fewer collisions – high performance as each device only connected to switch
●​ easily scalable – device directly connected to switch
●​ more resilient – not reliant on one single cable

How data is transmitted

●​ data from sending device is transmitted to the router

 9

●​ data has address of recipient

●​ router determines recipient’s destination address – using routing table
●​ router transmits data directly/only to recipient

Mesh Topology

●​ all computers connected to at least one other device

●​ multiple routes between devices
●​ computers can act as relays – forward packets to final destination

Advantages of mesh top.

●​ if one line goes down alternate routes are available

●​ improved security – not using one main line
●​ fewer collisions (more routes available)
●​ new nodes can be added without interfering with others

Cloud Computing

●​ accessing a file/service on a remote server

●​ Public cloud – services offered by a third party over public internet, available to
anyone with appropriate software/equipment (resources available on Internet,
available to anyone)
●​ Private cloud – services offered by a private internal network, only available to
select users (not general public), dedicated system only accessible from the
organisation

Advantages

●​ can be free
●​ saves storage on existing devices
●​ data can be accessed from any device (with internet access)
●​ data will likely be backed up/ higher chance of recovery
●​ better security
●​ scalable & easily shared

Disadvantages

●​ only accessible with internet access

●​ can take a long time to upload/download the data
●​ can be expensive (long term)
 10

●​ may have limited storage space for free

●​ may not have backup/recovery or security options
●​ you are reliant on someone else/dependant on third party (for security or backup)
●​ can’t access files if server goes down

Disadvantages of public cloud

●​ loss of control – data is stored on remote infrastructure, relies on external provider

●​ requires reliable internet connection to access data
●​ increased recurring costs – provides charger must be paid, LAN is one-time only

Advantages of a wired network (better performance, faster access)

●​ higher bandwidth – less latency (good for streaming larger files)

●​ more reliable/stable connection – less vulnerable to interference (distance/walls)
●​ more secure – confidential data can be transferred securely

Advantages of a wireless network

●​ freedom of movement (not fixed to a single location) – can move between rooms,
no need for physical connection // devices can be portable
●​ easily expandable/scalable if more devices want to join (ppl can join on multiple
devices)
●​ less cabling needed – cheaper setup
●​ allows access in remote locations (e.g. rural areas)

Disadvantages of a wireless network

●​ higher latency
●​ affected by weather
●​ slower transmission speed
●​ direct line of sight needed

Copper Cables

●​ data is transmitted through electrical signals

●​ lower transmission rate
●​ chance of interference and interception
●​ require repeaters over long distances
●​ more sturdy/reliable/flexible
 11

Fibre-optic Cables

●​ data is transmitted using light

●​ has a greater bandwidth and faster transmission speed
●​ smaller risk of interference
●​ can be used over long distances – needs less signal boosting
●​ more difficult to hack into
●​ more prone to damage, less flexible/can break when bent, more expensive to
install, difficult to terminate

Radio Waves

●​ carries data wirelessly in the form of electromagnetic waves

Satellite

●​ communication device in Earth’s orbit

●​ receives and transmits data

Switch

●​ allows communication between devices

●​ connects individual devices to each other
●​ receives transmissions and forwards them to their destination

Server

●​ manages access to a centralised resource (usually between devices on LAN)

Functions of WNIC

●​ provides interface/allows connection to wireless network as an antenna

●​ receives analogue waves and converts them to digital
●​ takes digital input and converts it into analogue waves -sends radio waves
through antenna
●​ encrypts and decrypts data
●​ provides MAC address to identify device on network

WAP

●​ hardware that provides radio communication from central device to nodes on a

network
●​ allows connection of devices using radio waves/signals/Wi-Fi
 12

●​ allows wireless enabled devices to connect to wired network

Bridge

●​ connects two LANs with the same protocol

●​ allows communication/data transmission between two networks with same
protocol

Repeater

●​ restores a digital signal so it can be transmitted over greater distances

Role of Router **

●​ receives packets from devices // external network/internet

●​ stores IP and MAC addresses of all devices attached to it
●​ maintains routing table
●​ routes/forwards packets to destination
●​ finds destination of a packet (using IP addresses)
●​ assigns private IP addresses to devices on a LAN
●​ finds most efficient path to destination
●​ can act as a firewall, gateway (+ perform protocol conversion/changes packet
format)

Ethernet

●​ a protocol
●​ used for data transmission over a wired network
●​ uses CSMA/CD
●​ data is transmitted in frames – each frame has source and destination address and
error checking data

CSMA/CD (protocol)

●​ used to detect and prevent collisions

●​ device/node that listens to a communication channel (scans voltage)
●​ data is only sent when channel is free/idle // line is empty

As there are multiple nodes on network/topology…

●​ data from two nodes can start to transmit simultaneously, causing collision
●​ if collision occurs, nodes send signal to stop transmitting
●​ waits a random time before attempting to send data again
 13

Bit Streaming

●​ Data is compressed before transmitting

●​ Video is transmitted continuously as series of bits
●​ On download, the server sends data to a buffer on the client computers
●​ Recipient receives bit stream from the buffer

Real-time

●​ Used when watching a live stream of events that are currently taking place
●​ Event is captured live with a video camera connected to a computer
●​ Media is sent to user’s device/buffer via bit stream directly as it is being recorded
●​ Cannot be paused or rewound

On-demand

●​ Video is already recorded/event has taken place

●​ Existing media is encoded to bit streaming format and uploaded to a server
●​ Can be watched at user’s convenience (can be paused/forwarded or rewound)

Difference between the WWW and the internet

WWW

●​ uses http/https protocol to transmit data

●​ collection of web pages

Internet

●​ uses TCP or IP protocols

●​ interconnected network of networks

Hardware Supporting the Internet

PSTN

●​ consists of many different types of communication lines

●​ allows for full duplex data transmission
●​ communication passes through different switching centres
●​ line remains active even during power outage
●​ dedicated channel used between two points for duration of phone call
 14

2.2 Networks including the internet continued

IPv4

●​ 4 groups each represented by 8 bits (32bit)

●​ Either denary or hexadecimal with numbers between 0-255 (in each group)
●​ Separated by full stops

IPv6

●​ 8 groups each represented by 16 bits (128bit)

●​ Hexadecimal numbers between 0 and FFFF
●​ Groups only contains 0 can be replaced with ::
●​ Separated using colons
●​ Used when the number or IP addresses needed exceeds number available using
IPv4

Subnetting benefits **

●​ improves security – data stays in its subnet, not all devices can access all areas of
network, devices do not receive unintended data
●​ allows extension of network/easier to expand – allows greater range of IP
addresses
●​ reduces amount of traffic in a network – improves network speed, data stays
within subnet and devices are not flooded with data
●​ easier maintenance/management – only one subnetwork may need taking down,
rest can continue *faults can also be isolated more efficiently

IP address in subnetwork

●​ made up of network ID and host ID

●​ each device on subnetwork has same network ID // each subnetwork has different
network ID
●​ every device in each subnetwork has different host ID but the same network ID
●​ host ID uniquely identifies device within the same subnetwork

Public IP address

●​ is visible to any device on the internet

●​ assigned to allow direct access to the internet
●​ allocated by ISP
 15

●​ unique throughout internet

Private IP address

●​ only visible to devices within the LAN

●​ used for internal LAN communication only
●​ allocated by router
●​ only unique within LAN

Dynamic IP address

●​ new one is reallocated each time a device rejoins a network

Static IP address

●​ does not change each time a device connects to the internet // is fixed

URL, WWW and DNS Use

●​ A URL is entered into web browser and parsed to obtain the domain name
●​ The Domain name is sent to DNS
●​ DNS has database of domain names and their corresponding IP Addresses
●​ DNS searches its database for given domain name
●​ If found, the IP addresses is returned to web browser, which displays the
resource
●​ If not found, the request is forwarded to a higher level DNS and the IP addresses
returned is added to the database of the lower level DNS
 16

3.​ Hardware

3.1 Computers and their components

Need for Secondary Storage

●​ to store files/data/software long-term

Need for Primary Storage

●​ to store files needed to boot system

●​ to store OS or any system software
●​ to store intermediate data/current data being processed

Embedded system **

●​ microprocessor within a larger system that performs a specific task

●​ has memory, input/output abilities and processor
●​ integrated into machine, not easily changed (by user/owner)
●​ e.g. a system in a washing machine that only controls cycle programs
●​ combination of hardware and software designed for a specific function
●​ does not have its own operating system
●​ does not require much processing power

Disadvantages

●​ difficult to change/update firmware by user // difficult to upgrade to take advantage

of new technology
●​ cannot be easily adapted for another task
●​ difficult to update or repair – usually thrown away instead

***no hardware operation questions have been repeated so far so chances are the other
devices we have to know (e.g. speakers, laser printers etc.) will likely come up***

Operation of Laser Printer

●​ contains a revolving drum that is given an electrical charge

●​ contents of page (provided by buffer/user) are drawn on drum as an electrostatic
charge by a laser beam that moves back and forth
●​ the drum is coated with oppositely charged toner which only sticks to areas
charged by the laser beam
●​ the drum then rolls over electrostatically charged paper, transferring the pattern
onto the page
 17

●​ paper is passed through a fuser/is heated to seal image

●​ the electrical charge is removed from drum, excess toner is collected

Operation of 3D Printer

●​ additive manufacturing
●​ uses digital 3D model or CAD file
●​ builds up model one layer at a time – starting from bottom, using xyz coordinates
●​ material is fused together layer by layer

●​ Fused Deposition Modelling – material is heated and pushed through nozzle

Use of temperature sensor

●​ prevents overheating // ensure material is hot enough

●​ identifies material of object/material being used

Microphone Operation

●​ has a diaphragm
●​ incoming sounds waves cause vibrations about diaphragm
●​ this causes coil to move past a magnet
●​ electrical signal is produced

Speaker

●​ an electric current is sent to speaker

●​ the electric current passes through a coil
●​ current in the coil creates an electromagnetic field
●​ an electromagnet is repelled by, or attracted to the permanent magnet based on
direction of the current in the coil
●​ movement of coil causes the diaphragm to vibrate – this vibration creates sound
waves

Magnetic Hard Disk

●​ has platters – divided into sectors and concentric tracks

●​ surface of disk can be magnetised
●​ has a read/write head mounted on an arm
●​ data is encoded as a magnetic pattern
●​ writing – variation in current in head causes variation in magnetic field on disk
 18

●​ reading – variation in magnetic field causes variation in current through head

​ Advantages

●​ costs less per unit storage (used when large storage capacity is required)
●​ has more longevity (used with devices that work all the time and have large
number of read write operations)

Solid State Memory

●​ uses a grid of columns and rows (arrays/blocks) that has two transistors at each
intersection:
●​ floating gate – stores voltage (represents either a 1 or 0)
●​ control gate – controls movement of charge/electrons during read/write
operations
●​ not possible to overwrite existing data (need to erase first then write data into
location)

​ Advantages

●​ no moving parts – more reliable

●​ faster data access times

Optical Disk Reader/Writer

●​ a rotating disk with concentric tracks made from a reflective metal layer
●​ data is read/written using laser light (either red or blue) that is shone onto the
disc
●​ data is stored in pits and lands on the track – sequences of amorphous and
crystalline states on the metallic layer (correspond to 0s and 1s)
●​ reading – reflected light from different states is encoded as a bit pattern
●​ writing – laser changes surface to crystalline or amorphous states based on the
bit pattern being stored
●​ read and write operation can occur simultaneously

​ Features/Uses

●​ used for transferring data between devices or as back-up systems

●​ can be read-only – used to distribute software, movies or games
●​ generally have lower storage capacity
 19

Resistive touch screen

●​ has two layers

●​ when user touches screen, layers touch and a circuit is completed
●​ processor determines horizontal and vertical point of contact
●​ will work if any object touches the screen

Capacitive touch screen

●​ has several layers

●​ when top layer is touched, there is a change in the electric current
●​ a microprocessor identifies the coordinates of touch

Virtual Reality Headset

●​ video/data is sent from a computer to the headset

●​ the video feed is sent to an LCD/OLED display
●​ there are two lenses placed between eyes and screen – allows for focusing and
reshaping of video for each eye, creating 3D effect
●​ uses an fps of 60 to 120
●​ sensors measure/track movements of user, allowing the video on the screen to
react to and mimic movements
●​ uses binaural (surround) sound so sound from speakers appears to come from all
directions
●​ can also use infrared sensors to monitor eye movement, allows depth of field on
screen to be more realistic

Purpose of a buffer

●​ To act as temporary storage – temporarily stores data until it is ready to be

transmitted to the device
●​ Stores data before it is used by receiving device
●​ Allows processes to operate independently of each other
●​ e.g. video buffer when streaming videos, printer buffer when data is transferred
from computer to printer

Process

●​ instructions and data are sent by computer to buffer

●​ data is transferred from buffer to device – allowing user to continue using
computer // allowing processor to continue processing
●​ when buffer is empty, an interrupt is sent to computer, requesting more data
 20

RAM

●​ primary memory
●​ stores currently running part of software/data/OS/programs/processes
●​ either static or dynamic
●​ can store data about I/O devices, contents of buffer or information about current
process

●​ SRAM – transistors arranged as flip-flops, more complex circuitry

●​ DRAM – uses (a single) transistors and capacitors, stores bits as a charge

SRAM Advantages/Disadvantages

●​ faster access time – because it does not need to be refreshed

●​ used on CPU for performance (used in cache memory)
●​ has lower data density
●​ does not need to be refreshed – consumes less power

DRAM Advantages/Disadvantages

●​ costs less per unit

●​ higher storage/data/bit density – more data stored per chip
●​ simpler design – uses fewer transistors
●​ needs to be refreshed – has higher power consumption
●​ slower access speed (used in main memory)

ROM

●​ primary memory
●​ stores start-up instructions/BIOS, firmware, any permanently required data
●​ stores the kernel of the operating system // parts of OS
●​ PROM – can be set once
●​ EPROM – erased using UV light, needs to be removed from device, can be
overwritten multiple times, must be entirely erased to rewrite
●​ EEPROM – erased using voltage (no additional equipment is needed), erased
within device, can be overwritten multiple times, does not have to be entirely
erased before rewriting, contents of firmware can be changed easily
 21

Use in embedded systems

●​ to store data that does not change

●​ data must be stored when device is powered off
●​ stores boot up instructions

Control system

●​ uses feedback
●​ produces an action

Role of an Actuator

●​ generates signal // converts electrical energy into mechanical energy

●​ to produce an action (be specific to situation)

Importance of feedback

●​ ensures system operates within given criteria

●​ allows system output to affect system input
●​ allows conditions to be automatically adjusted

3.2 Logic Gates and Logic Circuits

 22

4.​ Processor Fundamentals

4.1 CPU Architecture

Stored Program Concept

●​ instructions and data are stored in the same memory space/in main memory

Components in Von Neumann Architecture

●​ Buses
●​ Registers
●​ CPU
●​ CU
●​ ALU
●​ IAS
●​ System clock

General Purpose Registers

●​ Hold temporary data when performing operations

●​ Used for any purpose
●​ Can be used by most instructions

Special Purpose Registers (All registers we have to know are special purpose)

●​ Hold status of a program

●​ Specialised for specific use
●​ Can only be used by certain instructions

Program counter

●​ holds the address of the next instruction to be loaded

●​ is incremented once an instruction has been carried out

Memory Address Register

●​ stored address from memory location currently being read from or written to
(where data is being fetched from)

Memory Data Register

●​ holds the data fetched from address in MAR

 23

●​ data is copied to CIR

Current Instruction Register

●​ holds data from MDR

●​ holds instruction currently being decoded/executed

Index Register

●​ stores a value that is added to an address to give another address

Status Register

●​ stores flags – from results of logic and arithmetic operations // interrupt flags
●​ contains bits – can be individually set or cleared depending on operation

CU

●​ synchronises actions of other components in the CPU *based on pulses of system

clock
●​ sends/receives control signals along the control bus
●​ manages the execution of instructions – decodes an instructions opcode during FE
cycle
●​ controls communication between components of CPU
●​ types of signals it transfers: interrupt, timing, read & write

System Clock

●​ synchronises computer operations by creating time signals

●​ allows operations to be processed in the correct order / sequence
●​ keeps track of date and time

Immediate Access Store

●​ holds all the data/programs currently in use

●​ volatile memory
●​ has fast access times
 24

How data is transferred between components

●​ system clock gives out timing signals – sent on the control bus, this synchronises
the other system components
●​ CU initialises data transfer – generates signals that are sent on control bus to
other components

Role of Buses in Storing Data

●​ Address bus – carries address where data is being/going to be written to or read

from
●​ Data bus – carries data between the devices, buffer, or components

Performance **

Number of cores

●​ each core processes one instruction per clock pulse

●​ more cores mean that more sequences of instructions can be carries out
simultaneously
●​ multiple instructions can be executed per clock pulse (individua
●​ decreased time taken to complete a task

Why this may not increase performance

●​ software may not be designed for multiple cores – one core will be left idle
●​ memory access speed wont match speed of cores – causes delay
●​ may be other differences – e.g. amount of RAM

Bus Width

●​ allows transfer of more data each time/each transfer

Clock speed

●​ each instruction is carried out on a clock pulse

●​ clock speed dictates the rate at which instructions are being run
●​ faster clock speed = more instructions can be run per second/time period // more
FE cycles per unit time

Cache

●​ higher capacity means it can store a higher amount of frequently used instructions
for fast access
 25

●​ fast access memory (close to CPU)

●​ stores frequently used instructions/data
●​ more cache means more instructions can be transferred faster – less swapping
between RAM and cache
●​ prevents CPU idling while waiting for data

Quantity of RAM

●​ more applications can reside in main memory simultaneously

●​ saves/decreases disk access times

Ports

●​ purpose – to provide connection to peripherals & provide interface between

computer and other devices
●​ USB // COM port – fast data transfer speeds, universal standard
-​ 1 bit transferred at a time
-​ Can be asynchronous or synchronous
-​ USB-3, full duplex, otherwise, half-duplex
●​ HDMI
-​ allows video and audio to be transferred on one cable, convenient
●​ VGA
●​ DisplayPort

Fetch-Execute Cycle (Role of Registers)

●​ PC stores address of next instruction to be fetched + contents are incremented

each cycle
●​ MAR holds address where data is fetched from (from which data will be read
from/written into)
●​ MDR holds data at/from address in MAR (or data to be entered into it)
●​ Instruction (from MDR) is copied to CIR for decoding and execution

Register Transfer Notation **

[PC] ← [PC] + 1 – address in PC is incremented

MAR ← [PC] – contents of PC loaded to MAR

 26

MDR ← [[MAR]] – data located in address held in MAR is copied to MDR

CIR ← [MDR] – contents of MDR copied to CIR

Stages of the FE Cycle

●​ Next address to be fetched is stored in the PC

●​ This address is copied into MAR, using address bus
●​ Instruction located in that address is copied from main memory to MDR, using
data bus
●​ Instruction is copied from MDR to CIR
●​ Instruction is decoded by CIR (into opcode and operand)
●​ Processor executes instruction
●​ Address in PC is incremented

Purpose of an Interrupt

●​ to send a signal from a device/process

●​ seeking the attention of the processor

Causes of Software Interrupt **

●​ division by zero // runtime error

●​ attempt to access invalid memory location
●​ array index out of bounds
●​ stack overflow
●​ buffer overflow

Causes of a Hardware Interrupt

●​ printer is out of paper

●​ keyboard key press
●​ power failure

Interrupts **

●​ Interrupt flag raised in interrupt register

●​ Register checked for at start/end of FE cycle (before it begins/after execution stage)
●​ Type and source of interrupt identified
●​ Priority is checked against the current process – if lower, FE cycle continues, if
higher…
●​ Contents of registers are stored in/moved to a stack
●​ Address of Appropriate ISR called to handle the interrupt (loaded to PC)
 27

●​ Once ISR finishes, check for further interrupts (if found, repeat cycle)
●​ Otherwise load data/contents from stack into registers and continue with previous
process

4.2 Assembly Language

Two-Pass Assembler

First Pass – create symbol table

●​ reads assembly language instructions, adds any new symbolic addresses to

symbol table
●​ removes comments and white space
●​ reads assembly language program one line at a time
●​ checks the opcode is in the instruction set
●​ adds labels to symbol table

Second Pass

●​ generates object/machine code

●​ reads assembly language program one line at a time

Instruction Groups **

●​ data movement – moves data between addresses or other locations

●​ input/output of data – takes input from user, outputs char of binary num
●​ arithmetic operations – perform addition or subtraction
●​ unconditional/conditional instructions – move to another instruction
●​ compare instructions – compares result to another value

●​ all types of addressing load contents of given/calculated address into accumulator

Immediate Addressing

●​ operand is the data

Direct Addressing

●​ operand holds memory address of data

●​ addresses address given by operand

Indirect Addressing
 28

●​ the address used is at the address the user gives

●​ operand holds memory address that stores the memory address of the data

Indexed Addressing

●​ forms address from given address/address in operand plus the contents of the
index register

Relative Addressing

●​ the address to be used is an offset number of locations away, relative to the

address of the current instruction
●​ allows for relocatable code – all target addresses can be specified by the base
address

4.3 Bit manipulation

Binary Shifts – moving bits in a register a certain number of places within a register

●​ Logical Shift – bits shifted out of register are replaced with 0s

●​ Arithmetic Shift – bits shifted out of register are replaced with 1s
●​ Cyclic Shift – no bits are lost, bits shifted out of one end appear at the other end
●​ Left Shift
●​ Right Shift

Bit Masking

●​ AND is used to check is bit has been set

●​ OR is used to set the bits
●​ XOR is used to clear a bit that has been set

5.​ System Software

5.1 Operating Systems

Purpose of an Operating System

 29

●​ to provide a user interface

●​ to provide a platform for software to run
●​ to hide complexities of hardware from user

Operating System – Management tasks

●​ memory management
●​ file management
●​ security management
●​ hardware management
●​ input/output management
●​ process management
●​ error checking and recovery

Memory Management

●​ controls movement of data between RAM/processor

●​ allocates memory to processes
●​ dynamically allocates memory to processes/programs

How MM Allocates RAM

●​ RAM is assigned into blocks

●​ dynamic allocation of RAM to programs
●​ reclaims unused block of RAM
●​ prevents two programs occupying same area of RAM simultaneously
●​ moves data from secondary storage when needed (manages paging & virtual
memory)

File Management Tasks

●​ space allocated to particular files

●​ maintains a directory structure
●​ provides file naming abilities
●​ implements access rights
●​ allows file sharing
●​ specifies tasks that can be performed on a file (copy, paste, delete, close…)

Security Management

●​ creates accounts/passwords
●​ provides firewall or anti-malware
●​ validates user and process authenticity
 30

Hardware Management

●​ receives data from input devices

●​ sends data to output devices
●​ operates/installs device drivers – installs programs for devices connected to
external ports
●​ allows communication between peripheral devices and computer
●​ handles buffers for transfer of data – ensures smooth transfer between devices
transmitting and receiving at different speeds
●​ manages interrupts from device

Process Management

●​ manages scheduling of processes – decides which process to run next // order of

processes
●​ manages resources the processes require – e.g. allocating memory
●​ enables processes to share/transfer data
●​ prevents interference between processes
●​ handles process queue
●​ supports/allows multitasking – ensures fair access, handling priorities/interrupts

Utility Software **

●​ help set-up/optimise/maintain the computer

●​ makes memory allocation more efficient
●​ checks the system for faults

Disk formatter – disk needs to be prepared/initialised for use

●​ Prepares/initialises a disk for storing files by partitioning it (generates new file

system)
●​ Can delete all data from disc
●​ Sets up file allocation table
●​ Needs to check disk for errors

Defragmentation – over time, saving and deleting small files fragments disk

●​ Moves/rearranges blocks of files so that each individual file is contiguous in

memory
●​ Moves free space together
●​ Less time taken to access files (less head movement as data is contiguous)
●​ Improves disk access times – no need to search for next fragment
 31

Disk repair – needed to optimise performance

●​ Scans for errors/inconsistencies in a disk and corrects them – prevents bad sectors
being used
●​ Reduces access times by optimising storage

Back-up (Software) – allows retrieval of data, provides security against loss

●​ Creates copy of data in case the original is lost (in regular intervals)
●​ Allows retrieval of data is any is lost/corrupted

Disk/System clean up

●​ Optimises storage by removing unwanted files

Compression software

●​ Reduces file size

●​ Saves storage and memory space
●​ Reduces transmission time

Virus Checker

●​ Frees up RAM – removes software that takes up memory/replicates, scans for

malicious program code

Program Library **

●​ contains pre-written functions and subroutines

●​ can be referenced or imported
●​ the functions/routines it has can be called in own program

●​ saves time as code does not have to be written from scratch

●​ more likely to work (it’s already tested)
●​ program updates automatically (if update in routine)
●​ can perform more complex calculations than programmer is able to do

Dynamic Link Library Benefits **

●​ requires less main memory as DLL is only loaded once when needed
●​ executable file is smaller (does not contain all library routines)
●​ no maintenance needed from programmer (DDL is separate from program)
 32

●​ no need to recompile program if changes are made to DDL (changes/improvements

to DDL file code are done independently of the main program)

5.2 Language Translators

●​ convert a high-level or assembly programming language to a different form
(usually machine/object code)

Assembler

●​ translated assembly code into machine code

Compiler **

●​ translate high-level language (entire code is translated then run)

●​ used after program is completed
●​ produces error report after translating
●​ creates executable file – can be run without source code

After Testing

●​ produces executable file

●​ user cannot access/edit/sell the code
●​ users do not need translator to run program
●​ can be compiled for different hardware specifications – generates more income
●​ program can be tested multiple times without need to retranslate

Drawbacks when testing

●​ code cannot be changed without recompilation

●​ program will not run if there are any errors
●​ errors cannot be corrected in real-time
●​ one error may result in false errors being reported
●​ cannot easily test individual sections of code / cannot test if unfinished

Advantages when testing

●​ can debug multiple errors simultaneously

●​ produces executable file
●​ developer can test program multiple times without recompiling

Interpreter **

●​ translate high-level language line by line (is translated then run)

 33

●​ used while writing a program for testing and debugging

●​ errors can be corrected in real time – stops when an error occurs and displays
position

When Testing

●​ allows developer to make real-time changes that can be seen immediately

●​ program can be debugged at each stage
●​ developer can test when incomplete – small parts can be tested individually, if one
section does not work others can still be tested
●​ to avoid dependant errors

Partial Interpreters/Compilers

●​ can be used on different platforms as they are interpreted when run

●​ code is optimised for CPU as machine code is generated at run time
●​ source code does not need to be recompiled – more efficient to run

(programs may not need to be compiled if the software is already an executable file, has
been pre-compiled/built using compiler or if the source code has not been provided)

IDE Features/Tools **

Coding

●​ context sensitive prompts – as the code is being written, displays

predictions/options to complete the statements, suggests additions
●​ auto-complete – helps programmer figure out what to type next
●​ auto-correct

Error Detection

●​ dynamic syntax check – underlines/highlights syntax errors as code is being

entered/in real-time

​ Presentation

●​ pretty printing – helps identify key terms

●​ auto indentation
●​ expand/collapse code blocks

Debugging
 34

●​ single stepping – allows programmer to run the code one line at a time (breaking
in between) so effects of each statement on values can be seen/checked
●​ breakpoints – stop the code executing at a set line (to check current
values/progress)
●​ report windows – output contents of variables and data structures (see how
variables change)
●​ variables & expressions

6.​ Security, Privacy and Data Integrity

6.1 Data Security

Data Security

●​ protects data against loss/corruption

 35

●​ ensure recovery

Data Privacy

●​ ensuring data is protected against unauthorised access

Data Integrity

●​ ensures consistency/accuracy of data and whether it is up to date

●​ e.g. validation/verification rules, cascading update/delete

Why Data needs to be kept secure

●​ to protect against someone deleting/modifying or stealing it

Why the computer system need to be kept secure

●​ to protect against someone installing malware or damaging the system/accessing

data on it

Measures Used to Protect Computer Systems

Two factor authentication

Strong username and password

●​ prevents unauthorised access

Biometric passwords

Digital signatures

●​ user puts message through hashing algorithm to produce a digest

●​ digest encrypted with sender’s private key (creates digital signature)
●​ message and signature sent to receiver
●​ receiver decrypts signature to reproduce digest – same hashing algorithm used on
received document to create second digest
●​ signature digest and receivers digest are compared (document is authentic is they
are the same)
●​ digital signature can only be decrypted using matching sender’s public key

Firewall – accepts or rejects incoming/outgoing packets based on criteria

 36

●​ monitors incoming and outgoing traffic/packets – compares it to criteria set by user

(checks against whitelist/blacklist/blocked IP addresses)
●​ blocks/rejects files or transmissions that do not match criteria, accepts ones who
do

Up-to-date Anti-Malware – scans files on hard drive for malicious software

●​ regularly scans computer for viruses, checks against stored database of viruses
●​ database needs to be updated regularly
●​ if virus detected, it is quarantined or deleted
●​ compares downloaded files to database of known viruses – prevents download
from continuing

Anti-spyware

●​ scans computer for spyware, checks against stored database of spyware

●​ database needs to be updated regularly
●​ if spyware detected, it is quarantined or deleted
●​ compares downloaded files to database of known spyware – prevents download
from continuing

Backups

●​ data can be restored in case it is damages/lost due to malware

Encryption

●​ converts data into cipher text

Access rights

Threats Posed by Internet/Networks

Virus/Malware

●​ malicious software that replicates itself

●​ downloaded/run without user’s knowledge
●​ runs in background
●​ can pretend to be legitimate
 37

●​ can damage/delete/corrupt data

●​ does not send data out of computer

Spyware

●​ malicious software
●​ downloaded/run without user’s knowledge
●​ runs in background
●​ can pretend to be legitimate

●​ secretly records/collects user’s data/actions

●​ sends data/activity logs to third party
●​ does not replicate itself

Hacking/Hackers

●​ illegal/unauthorised access used to delete/damage/collect data

●​ done with malicious intent

Phishing – REQUIRES USER ACTION

●​ email pretends to be from official body

●​ persuades individuals to disclose private information
●​ requests authentication by redirecting to unofficial website

Pharming – AUTOMATIC

●​ redirects user to a false/fake website

●​ prevention – check URL validity, ensure connection is secure (https)

Security Methods to Protect Security of Data

Access Rights

●​ give certain users access to different elements

●​ uses different accounts/logins which have different levels of access
 38

●​ specific views can be assigned

Encryption

●​ data is converted into cipher text // data is encoded

●​ use – cannot be understood if intercepted without a decryption key

6.2 Data Integrity

Validation

●​ checks data is reasonable/sensible

●​ e.g. checking data is given in the right data type

​ when explaining use, refer to specific checks…

●​ format check: makes sure data is in required format

●​ length check: makes sure data contains correct number of characters
●​ type check: ensures only correct data type e.g. non-numeric data is entered
●​ existence check: makes sure data is already present in the system

Verification

●​ checks that data matches the original/is accurate

●​ e.g. visual check (manual comparison with source document/material), double
entry (enter twice, computer system compares)

Reasons why data might still be incorrect

●​ data on original document may not be correct – verification only checks if it

matches
●​ validation doesn’t check accuracy of data, only if it’s reasonable

Parity Check (used during transfer)

●​ type of parity is decided upon before transfer (even or odd)

●​ each byte contains a parity bit

In parity blocks…

●​ an additional parity byte is sent with vertical AND horizontal parity

●​ each row/column must have an even or odd number of 1s
●​ receiver counts the number of 1s and 0s in byte, allowing errors to be identified
●​ the intersection is the error
 39

●​ *an error cannot be detected if an even number of bits has been changed, as they
could cancel each other out*

Checksum (used during transfer)

●​ checksum value calculated from the data before transmission

●​ value is transmitted along with data
●​ receiving computer recalculates checksum from the received data
●​ the received checksum and recalculated checksum match, no error has occurred
●​ if they don’t match, an error has occurred

7.​ Ethics and Ownership

7.1 Ethics and Ownership

Benefits of Joining an Ethical Body e.g. BCS, IEEE

●​ has set ethical guidelines to follow – clients/staff know standards, no

decision-making required
 40

●​ professional integrity - clients/staff know he is reputable (shows skills/knowledge)

●​ they provide help and support (legal advice)
●​ training courses – helps to keep skills up to date

Reasons to Act Ethically

●​ make sure team members feel valued

●​ maximise quality of work
●​ promote teamwork and cooperation
●​ enable creation of best product for customer

How to Act in Best Interest of Client

●​ keep client’s personal data private

●​ involve client in development // communicate with client
●​ provide solutions the client asked for
●​ keep project on schedule and stay within given budget
●​ keep client informed of any problems/delays

How to Act Ethically

●​ be truthful – ask for help on how to use a program

●​ perform your own research
●​ ask for additional training/mentor

Being Unethical (Not acting in the best interests of…)

●​ product – might fail if error isn’t reported

●​ client – code might not work, lets down client
●​ profession – failing in his duties as an employee
●​ company

Copyright

●​ The formal and legal rights to ownership // Intellectual property rights

●​ Protects against unauthorised reproduction of work
●​ Allows for legal right of redress

​ Preventing Illegal Copies

●​ encryption
●​ use product key
●​ Compile source code (distribute as .exe file)
 41

Software Licencing

Free Software Foundation & Open Source Initiative

●​ user can edit/improve source code (must be released under same conditions as
original software/same legal rights apply)
●​ user can redistribute the software

Reason for an open source licence

●​ allows user to customise code

●​ allows errors to be reported
●​ allows additional features to be added to code
●​ allows for collaboration

​ Reasons against an open source licence

●​ requires you to release source code

●​ allows anyone to edit/modify and share the source code/program
●​ doesn’t allow you to profit off the created program

Shareware

●​ enables program to be copyrighted – user cannot legally modify it, control over
product is maintained, protects intellectual property rights
●​ user always gets a trial period – can also gain income if bought afterwards
●​ user cannot redistribute the software
●​ cannot be edited – developed keeps control
●​ more people might try it – free trial allows more people to experience it, meaning
they are more likely to buy it (also allows people to give feedback based on free
trial

Commercial Software

●​ user must pay before being able to legally use the software
●​ cannot be redistributed or edited

Advantages of commercial software

●​ enables program to be copyrighted – user cannot legally edit it, control over
product is maintained
●​ protects source code / prevents changes being made
●​ a fee can be charged for program – programmer gains income
 42

●​ prevents illegal copies being made – action can be taken if this occurs
●​ for customer – likely to have fewer bugs, redress available if game is broken,
potentially better support as a fee is being charged

Applications of AI **

●​ police identifying wanted people – image/facial recognition (can identify

features/patterns in an image and match them to a person/object)
●​ natural language interfaces – speech recognition (can identify language/words
being spoken, can learn accents, matches words to database, generates most likely
sentence)
●​ self-driving cars – can detect position on road (self-parking, collision avoidance,
following a route)
●​ game playing – models characters in computer games (allows comp characters to
react according to player’s movements)

●​ can start recording to secondary storage when person is detected

●​ can identify direction of movement and then move the camera accordingly
●​ scans scene in real time

●​ facial/image recognition – takes each frame individually, analyses pixels

●​ camera focuses on pattern identified

Social impacts

●​ privacy issues – people do not like their data being stored

●​ incorrect recognition leads to mistakes (systems may be denied)
●​ individuals will feel safer – reduction in crime
●​ helps in catching criminals (links to reduction in crime)

Economic Impacts

●​ reduces costs – less time taken for tasks to be carried out

●​ increases profits – leads to more efficient work performance (redundant tasks done
by AI)
●​ decreases cost for customer
●​ decreases profit margins – program may be expensive to maintain/buy/update
 43

8.​ Databases

8.1 Database Concepts

Drawbacks/Limitations of a file-based approach **

●​ more data redundancy – same data is stored many times

●​ more data inconsistency (worse data integrity) – data is not updated across the
whole program and there may be duplicates that are stored differently
●​ hard to perform complex queries – new program has to be written each time
●​ lack of privacy – user views/access cannot be implemented

Advantages of a relational database (compared to file-based approach) **

 44

●​ reduced data redundancy – each data item is only stored once (due to linked
tables)
●​ maintains data consistency (data integrity) – changes in one table automatically
update in another
●​ complex queries easier to run
●​ can provide different views – user can only see specific aspects/parts of the
database (improved privacy)
●​ program-data independence – programs do not need to be rewritten if data is
changed (data is separate from software)

Entity

●​ object which data can be stored about

Field

●​ column/attribute in a table

Tuple (Record)

●​ a single row of data in a table (about one instance of an object)

Primary Key

●​ unique attribute – used to uniquely identify a record/tuple

●​ can be used as foreign key in another table // form link between the tables

Candidate key

●​ an attribute that could be a primary key

Secondary Key

●​ alternative/additional key used along with primary key to locate specific data
(candidate key that has not been chosen as primary key)

Foreign Key

●​ a field in one table that links to a primary key in another

Importance of referential integrity

●​ makes sure data in the database is consistent and up-to-date

 45

●​ ensures that every foreign key has a corresponding primary key

●​ prevents records being added/modified/deleted incorrectly
●​ makes sure any changes made to data in one place are reflected in all related
records
●​ ensures any queries made return accurate and complete results
●​ makes sure non-existent data cannot be referenced

Normalisation

1NF

●​ no repeating groups of attributes

2NF

●​ has no repeated attributes

●​ no partial key dependencies (all attributes fully dependant on primary key)

3NF

●​ no transitive dependencies
●​ all attributes/fields fully dependant on primary key AND NO OTHER ATTRIBUTES

0NF to 1NF

●​ remove/identify any repeating groups of attributes

●​ ensure each field is atomic (e.g. Name should be split into FirstName and
LastName)
●​ identify primary key

1NF to 2NF

●​ remove any partial key dependencies

2NF to 3NF

●​ remove any non-key dependencies

8.2 Database Management Systems (DBMS)

DBMS
 46

●​ provides data management

Data Dictionary Contents (metadata about database) // data abt data in a database

●​ table name // field name

●​ data types
●​ type of validation used // validation rules
●​ primary and foreign keys
●​ relationships between elements

Logical Schema – shows structure of database, its relationships (e.g. E-R diagram)

●​ overview of a database structure

●​ models the problem, using methods such as ER Diagrams
●​ independent of any particular DBMS // model of a database that is not specific to
one DBMS
●​ describes the relationship between data and its structure

Security in a DBMS

●​ authentication
●​ backup/recovery procedures – automatically creates copies of database and stores
it off-site on a regular basis, allows data to be recovered if lost
●​ access rights – users are given different access permissions to different tables,
read/write, read only etc.
●​ views – different user able to see different parts of database, only see what’s
required
●​ encryption – data is turned into ciphertext, cannot be understood without
decryption key
●​ record and table locking – prevents simultaneous access to data, so data is not
overwritten

Software Tools in DBMS

●​ provides a developer interface

-​ allows user to create items such as tables, forms and reports
-​ creates input & output abilities through
menus/buttons/touchscreen/monitor etc.
●​ query processor
-​ allows user to enter criteria
-​ searches for data that meets the entered criteria
-​ organises results to be displayed to the user
 47

-​ software that that processes and executes queries written in SQL

8.3 DDL and DML

●​ DDL statements – always end with a semicolon!!!

CREATE DATABASE database_name;

CREATE TABLE table_name(

column_name1 DATATYPE

column_name1 DATATYPE NOT NULL

PRIMARY KEY(column_name)

FOREIGN KEY(column_name) REFERENCES external_database_name(column_name)

);

ALTER TABLE table_name

ADD column_name DATATYPE

ADD PRIMARY KEY(column_name)

ADD FOREIGN KEY(column_name) REFERENCES

external_database_name(column_name)

DROP COLUMN column_name

RENAME COLUMN old_column_name TO new_column_name;

●​ Data Types
-​ CHARACTER
-​ VARCHAR(n) – like a string, VARCHAR(255) allows 255 characters
-​ BOOLEAN
-​ INTEGER
-​ REAL
-​ DATE
-​ TIME
 48

Paper 2

9.​ Algorithm Design and Problem-solving

9.1 Computational Thinking Skills

Abstraction

●​ Used to filter out unnecessary information // Means only essential information is

included
●​ Simplifies the solution – makes it easier to implement/design
●​ System is tailored to need of user

Decomposition
 49

●​ Breaking down a complex problem into smaller problems/subproblems which are

easier to program
●​ Makes problem easier to solve
●​ Leads to concept of program modules – subproblems can be assigned to individual
teams

​ Benefits

●​ Makes task easier to understand/solve – smaller problems are easier to solve

●​ Smaller problems are easier to program, test and maintain
●​ Sub-problems can be given different teams // can be solved separately

9.2 Algorithms

●​ algorithm – sequence of defined steps that describe how to solve a problem

Stepwise Refinement

●​ process of developing a modular design by splitting a problem into sub-tasks

●​ the sub-tasks are repeatedly split into smaller sub-tasks until each is just one
statement/element from which the task may be programmed

could include statements such as…

●​ set total/count to 0
●​ input a number
●​ check if…
●​ repeat from step # a set number of times // repeat for total of # iterations
●​ output the total/count/element/string
●​ calculate the rounded value of…
●​ assign value to an element
●​ use function … to return a value
●​ open file in read/write mode
●​ loop through all lines in a file
●​ loop ends when…

(obviously, be specific to the question, for example, say how many times it repeats and
from which step)

9.3 Algorithms Continued

 50

10.​ Data Types and Structures

10.1 Data Types and Records

Defining Record Structures (using arrays of given records and type)

●​ declare a record – containing all data items required // containing items of different
data types
●​ declare an array of the given record – each array element represents data for one
instance of an object (e.g. would store one customer order)

Advantages of Records

●​ a set of data of different data types is stored under a single identifier

●​ allows multiple instances to be referenced using the single identifier

Uses of BOOLEAN

●​ to terminate a conditional loop when value has been found

 51

●​ when the variable can only take one of two possible values

10.2 Arrays

●​ simplify an algorithm, makes it easier to amend or add data

●​ are easier to understand, test and debug
●​ it is possible to iterate through the values using a loop – make data organisation
easier
●​ arrays of records can store multiple data types

10.3 Files

Advantages of storing data

●​ Data in file is saved after computer is switched off (stored permanently)

●​ Data does not need to be manually re-entered when program is re-run

* WRITE mode overwrites a file completely/overwrites any existing data in file

How to store a set of data in a file

●​ Data items are combined to form a single string (set of data is saved as a single
line in file)
●​ Items are separated by a special character

10.4 Abstract Data Types

Linked Lists

●​ Consists of nodes – each node contains data and a pointer to the next node
●​ Has a pointer to the start of list (start pointer)
●​ Last node in list has a null pointer (indicated there are no further nodes in list)
●​ Data is added by manipulating pointers (Data in nodes does not need to be moved)
●​ Nodes are traversed in sequence (based on pointers)
●​ Unused nodes are stored on free list

​ Adding/inserting a new node

●​ check for a empty node

●​ assign data item to first empty node in free list (from now on referred to as X)
●​ set the pointer of X to point to the node that will come after it after insertion (Y)
●​ set pointer of node that previously pointed to Y to point to X
 52

●​ set pointer of free list to point at next empty node

​ Advantages

●​ It is easier to add/delete data in a linked list

●​ Only the pointers need to be changed if data contents are changed (determine
ordering of data)

​ Disadvantages

●​ Pointers need to be stored as well as data

●​ More complex to implement/set up

​ Implementing a Linked List

●​ Declare two 1D arrays – one for data, one for pointers

●​ Elements from same index represent one node (data in index position in array 1
maps to pointer in index position in second array)
●​ Declare variable for start pointer
●​ Declare variable for next free node pointer
●​ Define appropriate value for null pointer
●​ Use routines to add/delete/search within list

​ Using Record/based Implementation

●​ Define record type with fields for data and pointer

●​ Declare a 1D array of defined record type

Queues

●​ Each queue element contains one data item

●​ Has a pointer to the start and end of queue
●​ Works on a FIFO* basis
●​ May be circular
●​ Front of queue and End of queue pointers are equal when there is only one data
item in the queue

* first in first out – first data item put in queue is also the first one removed

​ Adding/inserting a value

●​ check queue is not full

 53

●​ increment variable/pointer that signifies end of queue

●​ increment variable storing the number of items in queue
●​ store data value being added in location pointed to by end of queue pointer
●​ in arrays – set element of array at end of queue to value/item being added

​ Removing a value

●​ value pointed to by the front of queue pointer is removed first – can be assigned to
a variable
●​ front of queue pointer is incremented

​ Required Operations

●​ adding an item – a check needs to be carried out to ensure queue is not full
●​ removing an item – a check needs to be carried out to ensure queue is not empty

​ Implementing a queue

●​ Declare 1D array of required/specified size

●​ Declare integer variable for front of queue pointer
●​ Declare variable for end of queue pointer
●​ Declare variable for size of queue to limit max number of items allowed
●​ Initialise variables e.g. the size of the queue

Stacks

●​ Used to store string data which needs to be accessed in several modules within a
program

​ Implementing a Stack

●​ Declare a 1D array of type string

●​ Number of elements in the array corresponds to the size of the required stack
●​ Declare variable for a stack pointer
●​ Declare variable for size of stack/max value of pointer
●​ Use stack pointer as index to array
●​ Initialise pointers and variables to indicate empty stack
●​ Store each item on stack as one array element
●​ Push and Pop routines used to operate stack (need to check if stack is full or empty
first)
 54

​ Why files may not appear correctly if using a stack (from O/N 2023 paper 22)

●​ If a multiple lines are stored at once, lines transferred to file will appear out of
sequence as stacks operate on a FILO basis
●​ Stack is full – not all lines can be stored on stack, resulting file will be missing
original lines
●​ Stack is empty – stack is being read faster than being written into, so blank lines
may be inserted into file

11.​ Programming

11.1 Programming Basics

Features that made codes understandable

●​ Indentation
●​ White space
●​ Comments
●​ Meaningful variable names, use of camelCase
●​ Capitalised keywords
●​ Standard way to indicate special cases (such as unused array elements) – allows
recognition when processing, means there is no unexpected data

​ Importance of ‘Good Practice’

●​ makes code easier to understand, describes purpose of sections of code and of

identifiers
 55

●​ makes code easier to debug/test and maintain

Constants

●​ used for values that are only entered once

●​ avoids input error or accidental change of value
●​ easier to maintain program when constant has to change
●​ make program easier to understand

Library Routines/Program Libraries

●​ More complex functions can be used in code through use of a program library

​ Benefits during development

●​ tried and tested free from errors

●​ perform function you may not be able to program yourself
●​ readily available and speed up development time

​ Use

●​ Used for tasks that are performed/repeated in several places within the code
●​ When part of an algorithm performs a specific task
●​ Reduces complexity of program
●​ Testing/debugging is easier

11.2 Constructs

Selection constructs

●​ IF statement
●​ CASE statement (from O/N 2023 paper 21)
-​ consists of clauses that are checked in sequence
-​ if a value satisfies the first clause, other clauses will never be tested
-​ the “otherwise” clause may never be performed if all possible values are
addressed via the other previous clauses

Iterative constructs

●​ WHILE pre conditional loop – used when number of iterations is not known
●​ REPEAT post conditional loop – used when number of iterations is not known
●​ FOR count-controlled loop – used for known number of iterations
 56

Sequence construct

11.3 Structured Programming

Functions/Procedures

●​ A function/procedure interface refers to the parameters and return value – e.g. a

function takes two integer parameters and returns a boolean value
-​ it provides a mechanism to allow calling a program to pass data
-​ defines/provides parameters, giving their data type and order

​ Benefits

●​ Can be called repeatedly when required

●​ Is only designed and tested once (then used repeatedly)
●​ Any changes to function code need to be made once only (easier to maintain)

●​ A function returns a value, a procedure can output values

12.​ Software Development

12.1 Program Development Life Cycle

Benefits

●​ makes projects/programs easier to plan and manage

●​ clear deliverables produced at end of each stage (can show prototypes to client)

Stages

Analysis

●​ Documents produced include: the problem definition, client requirements,

documentation related to current system e.g. ER diagrams
●​ Developer discusses program requirements with customer/client

​ Design
 57

●​ An identifier table is produced

●​ Data structures and choice of programming language are decided
●​ Includes algorithms, programs and pseudocode
●​ User interface

​ Coding

●​ Syntax errors can occur

​ Testing

●​ A trace table is produced

Waterfall Design

●​ each stage completed before new one is begun (linear)

●​ full documentation, lots of planning involved
●​ client only involved at start and end of process

​ Benefits

●​ easy to manage, stages do not overlap and are completed one at a time
●​ each stage has clear/specific deliverables (detailed documentation of project)
●​ leaves little room for client to change mind and the project is planned out
thoroughly, meaning a smoother development process (client also has better
idea of the finished product when designing it)

​ Disadvantages

●​ no working program/software until late in life cycle – slower to market than

competitors (does not allow early versions)
●​ more difficult to cope with changes in requirements
●​ needs high amount of feedback/involvement from client

Iterative Design

●​ development cycle is run repeatedly until full program has been developed
●​ split into stages which are repeated

​ Benefits

●​ working programs developed early on in cycle, produced at each iteration

 58

●​ easier to test and debug

●​ more flexible to changes in client requirements
●​ customer involved during all stages/at each iteration

​ Drawbacks

●​ a lot of planning needs to be done in order to section project into clear

stages/iterations (whole system needs to be defined at start)
●​ makes it easy for client to change mind often about end result

Rapid Application Development (RAD)

●​ modules are developed in parallel as prototypes

●​ minimal planning is carried out – allows for changes to requirements
●​ flexible development process
●​ used for time critical/sensitive development
●​ client involved during all stages of development

​ Benefits

●​ quicker development possible – multiple areas worked on simultaneously

●​ prototype produced in early stages
●​ easier to change requirements
●​ early review possible

​ Drawbacks

●​ difficult to estimate cost/time needed to complete project

●​ makes it easy for client to change mind often about end result
●​ documentation often omitted

12.2 Program Design

Structure Chart

​ Purpose

●​ To module relationships and see how a problem is broken down

●​ To determine whether a module is a function or procedure

Symbols

●​ Curved arrow – means iteration/looping

 59

●​ Downward arrow – result from one stage is input/passed to next stage

●​ Upward arrow – more work required on previous stage to complete the current
stage

State-Transition Diagrams

●​ arrows represent transitions between states

●​ each arrow can have an input or an output – always in format of i|o
●​ X signifies start
 60

12.3 Program Testing and Maintenance

Types of Maintenance

Adaptive

●​ Accommodates legislative changes or user requirement changes

●​ Allows program/software to be updated if new technology or library routines are
made available

​ Perfective

●​ Changes are made to program after it has been made available to public

​ Corrective

●​ Used when program does not operate as expected // contains a bug

Types of Testing

Beta Testing

●​ testing carried out by small group of potential users

●​ users check that program works as intended and identify any errors present
●​ users will provide feedback/suggestions for improvement
●​ problems identified are addressed before program is sold

Integration Testing

●​ each module in program is tested individually during development and is

debugged as necessary
●​ individual modules are combined into single program (or added to existing
program) and tested as a whole

Stub Testing

●​ used when program contains modules with errors/incomplete modules

●​ dysfunctional modules are replaced by dummy modules
●​ dummy modules return a known/expected result or output a statement to show
they have been called
 61

Walkthrough Method

(Required Information)

●​ the program code/specification – structure/design of the program needs to be

known
●​ test data and expected results – so all paths through program can be tested

​ How Errors are Identified

●​ program is checked by creating a trace table, going through program one line at a
time
●​ records/checks variables as they change
●​ error may be indicated when…
-​ variable is given an unexpected value
-​ unexpected path through program/faults in logic of program

White-box Testing

●​ detailed testing of how each procedure works

●​ tests logic of all possible paths through the program module

Black-box Testing

●​ tests a module’s inputs and outputs

Alpha Testing

●​ testing carried out by development team

●​ check program works as intended and address/identify any errors within the
program before releasing it to the public

Types of Errors

Syntax Errors

​ (Errors cannot be detected by examining a single line)

●​ Selection and iteration

●​ Incorrect block structure – missing keywords like ENDIF, ENDPROCEDURE
●​ Data type errors
●​ Incorrect parameter use
 62

●​ Incorrect brackets or misspelt keywords

​ A program with no syntax errors…

●​ Obeys the rules/grammar of the programming language used

●​ The program will run without an error being flagged

Run-time Errors

●​ Program performs and invalid operation

●​ E.g. division by zero or endless loop

Test Data

●​ Normal – value within an acceptable range (should be accepted)

●​ Abnormal – value outside acceptable range (should be rejected
●​ Boundary/Extreme – minimum/maximum acceptable value (should be accepted)

Pseudocode

General Operations

INPUT variable_name

OUTPUT variable_name1 , variable_name1 , “...”

* comments added using //

* <= and >= used

* <> used for not equal to

10. Data Types and Structures

10.1 Defining Records

 63

TYPE record_name

DECLARE field_name1 : DATA TYPE

DECLARE field_name2 : DATA TYPE

DECLARE field_name3 : DATA TYPE

ENDTYPE

Reading From Records

*data from a record can be transferred to an array…

​ DECLARE array_name : ARRAY[1:30] OF record_name

*referenced similarly to 2D arrays… name of the column in record is used as 2nd index

​ array_name[Index].column_name e.g. staff[1].EmployeeName

Saving Data to Records

​ DECLARE variable_name : record_name

​ array_name[Index].column_name ← *value*

10.2 Arrays

​ DECLARE 1Darray_name : ARRAY[x:y] OF DATA TYPE

​ DECLARE 2Darray_name : ARRAY[x:y , x:y] OF DATA TYPE

10.3 Files

​ OPENFILE file_name FOR WRITE

​ WRITEFILE file_name, *data being inserted*

​ OPENFILE file_name FOR READ

READFILE file_name, *variable data from file is being stored in*

 64

​ CLOSEFILE file_name

* file name should always end in .txt and be written in quotation marks e.g. “filename.txt”

* in exams, if given a file name as a variable, assume that is it already in the correct format
and includes the .txt portion

11. Programming

11.1 Basics

​ CONSTANT constant_name : *value*

* use constants with variables that do not change throughout code, you get marks for
using them instead of variables

* mark scheme tends to make constant names capitalised

DECLARE variable_name : DATATYPE

* when declaring multiple variables of the same data type, commas can be used and only
one declaration statement has to be made (e.g. DECLARE index, total, count : INTEGER)

​ variable_name ← *value* e.g. Counter ← Counter + 1 or Counter ← 0

11.2 Constructs (all have a statement indicating end of loop/condition e.g. ENDIF)

●​ if statements

IF <condition> THEN

*enter code to be executed*

ELSE

​ *enter code to be executed*

ENDIF​
 65

* elif does not exist in pseudocode

●​ case statements

​ CASE OF variable_name

​ ​ condition1 : *enter code to be executed*

​ ​ condition2 : *enter code to be executed*
​ ​ condition3 : *enter code to be executed*

​ ​ OTHERWISE : *enter code to be executed*

​ ENDCASE

●​ count-controlled loops (for loop)

FOR variable_name ← *value1* TO *value2*

​ *enter code to be executed*

NEXT variable_name

●​ post-condition loops (repeat loop)

​ REPEAT

​ ​ *enter code to be executed*

​ UNTIL <condition>

●​ pre-condition loops (while loop)

WHILE <condition>

​ ​ *enter code to be executed*

 66

​ ENDWHILE

11.3 Structured (Functions and Procedures)

●​ procedure

​ PROCEDURE procedure_name( parameter : DATA TYPE)

​ ​ *enter code to be executed*

​ ENDPROCEDURE

* if more parameters, use a comma to separate them

* there can also be no parameters, in which case leave brackets empty e.g.
parameter_name()

​ CALL procedure_name() or CALL procedure_name( *value* ) if parameters

●​ function

FUNCTION function_name( parameter : DATA TYPE) RETURNS DATA TYPE

​ ​ *enter code to be executed*

​ ENDFUNCTION

* return is used to output values e.g. RETURN variable_name

Pass by Value

●​ a copy of the variable is passed into the module, the variable is not changed
outside of the module

PROCEDURE name(BYVAL parameter : DATA TYPE)

Pass by Reference

●​ any changes made to variable within the module are permanent

PROCEDURE name(BYREF parameter : DATA TYPE)

 67

— Paper 3 —

13.​ Data Representation

13.1 User-defined Data Types

Purpose of user defined data types

●​ To create a new data type

●​ To allow construction of data types not available in a programming language
(extends flexibility of a programming language)
●​ Constructed by the programmer

Composite

●​ Collection of data that consists of multiple elements of different (or the same) data
types which are grouped under a single identifier
●​ Can be user-defined or primitive
 68

●​ Contain more than one data type in their definition

●​ Includes a record, set and class/object

Record

●​ Uses other data types in its definition to form a single new one
●​ Data types referenced can be primitive or user-defined
●​ Includes related items and a fixed number of items

TYPE record_name
DECLARE field_name1 : DATA TYPE
DECLARE field_name2 : DATA TYPE
DECLARE field_name3 : DATA TYPE
ENDTYPE

Set

●​ Includes a list of unordered elements

●​ Theory operations (such as intersection or union) can be applied to its elements
●​ Includes the data type it stores ias part of its definition
●​ All elements are of the same data type

​ TYPE <identifier> = SET OF DATA TYPE

​ DEFINE <set_name> = (value1, value2, value3, …) : <identifier>

Non-Composite

●​ Can be defined without referencing another data type

●​ Can be a primitive type available in a programming language or a user-defined
type
●​ Contains only one data type in their definition
●​ Includes a pointer and any primitive/enumerated data type
➔​ Enumerated Data Type – has an ordered list of all possible values
➔​ Pointer Data Type – used to reference a memory location, stores
addresses/memory locations and indicates the type of data stored there

(Pseudocode for Non-Composite Data Types)

●​ Enumerated
 69

TYPE <identifier> = (value1, value2, value3, …)

TYPE Season = (Spring, Summer, Autumn, Winter)

●​ Pointer

TYPE <identifier> = ^DATA TYPE

TYPE IntPointer = ^INTEGER

13.2 File Organisation and Access

File Organisation Methods

Serial Files

●​ Records are stored one after the other in chronological order

●​ New records are appended to the end of the file
●​ When searching – every record needs to be checked until the record is found or all
have been checked
●​ Examples include…
➔​ Creating unsorted/temporary transaction files
➔​ Creating data logging files

Sequential Files

●​ Files are stored and addressed one after the other

●​ A new version of the file has to be created to update it
●​ Files are stored with ordered records – records are stored in order of the key field
●​ New records are inserted in the correct position
●​ When searching – the key field is compared and every record is checked until it is
found, or the key field of a current record is greater than the one being searched
for

Random Files

●​ Records are stored in no particular order within the file (there is no sequencing)
●​ There is a relationship between the record key and its location within the file – the
location of the record is found using a hashing algorithm
●​ Updates to the file can be carried out directly

File Access Methods

 70

Sequential Access Process (For Sequential and Serial Files)

●​ Records are checked linearly until the desired record is found/end of file is reached
●​ Starts searching for records one after the other from the physical start of the file
until the record is found or the end of the file is reached
●​ Most suitable when data is stored in a certain order based on a field – e.g. bank
stored data records in ascending order of account number

Direct (For Sequential and Random Files)

●​ Most suitable when a record is referenced by a unique address

●​ Allows a record to be found in a file without other records being read – records are
found by using the key field of the target record (the record’s location is found
using a hashing algorithm)

●​ Sequential Files

➔​ an index of all key fields is kept – the index is searched for the address of
the file location where the target record is stored
●​ Random Files
➔​ a hashing algorithm is used on the key field of the record to calculate
address of the memory location where the target record is expected to be
stored
➔​ Linear probing or Search overflow can be used to find a record if it is not at
the expected location

Hashing algorithm

●​ Used in direct access methods on random and sequential files

●​ Is a mathematical formula - performs a calculation on the key field of a record
●​ Result of calculation gives the address of the memory location where the record is
located/should be stored

Hash Value Duplicates (a calculated hash value is a duplicate of another value for a
different record key)

Collisions

●​ occurs when two values/data items in the key field for two records result in the
same hash value (when passed through a hashing algorithm)
 71

●​ Means the storage location identified by the algorithm may already be in use by
another record – two records cannot occupy the same address

Process of Collision Resolution

●​ When storing a record

➔​ Linear progression - record is stored is next free memory space after the
one identified by the hashing algorithm
➔​ Record is stored in the next free memory space in the overflow area
●​ When finding a record
➔​ search the overflow area linearly until the matching record key is found (if
not found, record is not in file)

13.3 Floating-point Numbers, Representation and Manipulation

Changing the bit allocation

●​ When the number of bits in the mantissa is raised, the precision/accuracy of the
represented number increases – when the bit number is lowered, the accuracy is
reduced
●​ When the number of bits in the exponent is raised, the range of possible numbers
can be represented is increased – when the bit number is lowered, the range
decreases

Why binary numbers are stored in normalised form

●​ To store a maximum range of numbers in a minimum number of bytes/bits

●​ Normalisation minimises the number of leading 0/1s represented (Numbers whose
mantissa begins with 10 or 01 are normalised)
●​ Maximises the number of significant bits – increases precision/accuracy when
storing very small/large numbers
●​ Avoids the possibility of many numbers having multiple representation

Storing floating point numbers

●​ Large numbers require a greater number of bits for the mantissa

 72

​ Overflow

●​ Occurs following an arithmetic/logical operation – result is too large to be precisely

represented in the available system
●​ Numbers cannot be stored accurately in certain computer systems if they require
more bits than is available

Underflow

●​ Occurs following an arithmetic/logical operation – the result is too small to be

precisely represented in the available system (number does not have enough bits
to be represented)

*always refer to the loss in precision and state the number/digits would be truncated

Why A Binary Number Is Sometimes An Approximation

●​ Real (decimal) numbers can have a fractional part

●​ Binary numbers have limited fractional representation (limited to powers of 2)
●​ Fixed length of storage means you can’t store very large/small numbers – it’s not
possible to store all fractions with the level of precision that is provided

Converting between denary and floating-point binary

M x 2E where M is the mantissa and E is the exponent

​ Both are always in two’s complement (There is a binary point following the -1 in
the mantissa)

​ Mantissa Representation:

-1 1/2 1/4 1/8 1/16 1/32 1/64 1/128

​ Exponent Representation:
 73

-128 64 32 16 8 4 2 1

●​ Floating-point binary → Denary

1.​ Convert the number in the exponent to denary

2.​ Shift the mantissa binary point to the right by the number held in the
exponent
3.​ Convert to binary, where numbers after the point are ½, ¼, ⅛ and so on…

E.g. 01011010 x 00000100

➔​ 00000100 = 4
➔​ 0.1011010 - point shifts 4 places right
➔​ 01011.010 = 1 + 2 + 8 + ¼ = 11.25

●​ Denary → Floating-point binary

​ 1. Convert the number into binary, separating fractions using a binary point
​ 2. Move the binary point left until only one digit comes before it, counting the
amount of places the point moves up (gives you the exponent)

Eg. 4.75

➔​ 4.5 = 4 + ½ + ¼ = 0100.11
➔​ 0100.11 → 0.10011 - point is shifted 3 places left, hence the exponent is 3
➔​ 3 = 0011, so the final number is 010011 x 0011

14.​ Communication and Internet Technologies

14.1 Protocols

Purpose of Protocols

●​ Provide a standard set of rules to enable successful data transfer

●​ Enables communication/compatibility between devices from different
manufacturers or platforms
➔​ makes communications independent of software and hardware
●​ If two devices with different protocols were sending messages between each
other, they would not be able communicate properly
 74

Purpose of SSL and TSL protocols

●​ Provide communications security over the internet/network (by providing

encryption)
●​ Enable two parties to identify and authenticate each other – allows them to
communicate with confidentiality and integrity

Protocol Suites

●​ The protocols in a stack determine the interconnectivity rules for a layered network
model such as the TCP/IP model

TCP/IP Protocol Suite

●​ Maintains connection between two hosts and ensures successful delivery

between them
●​ During data transfer, the self-contained modules (layers) are applied in order
using the device’s software
●​ Separate layers make hardware and software compatibility easier to implement
●​ Can be represented as a stack, with each layer having a separate functionality

Application
Transport
Internet
Link

1. Application Layer (Protocols Used)

●​ HTTPS – for sending/receiving/transfer of web pages and hypertext documents

●​ FTP – for sending and receiving files over a network/between devices (used to
transfer data from server to client on the network)
●​ SMTP – push protocol for sending/uploading emails
●​ POP3 or IMAP – pull protocols for retrieving/receiving emails
➔​ Keeps the server and client in sync by not deleting the original email
●​ BitTorrent – provides peer-to-peer file sharing over a network
➔​ Allows sharing of files between many users connected together over the
internet
 75

➔​ Allows more users to share files than a normal peer-to-peer network

would
➔​ Users share files directly with each other – there is no web server (all users
are of equal status)

Purpose of Application Layer

●​ Provides access to all programs that exchange data - interacts directly with the
user
●​ Used by web browsers or server software
●​ Enables data transfer to/from the Transport Layer - allows applications to access
services in other TCP/IP layers
●​ Defines the protocols that an application uses to allow the exchange of data

2. Transport Layer – handles packets

●​ Responsible for delivery of data from source host to destination host

●​ Breaks data into manageable packets (performs segmentation) and sends them to
the internet layer
●​ Adds a packet header and the sequence number to the header (sequences packets)
●​ Controls flow of packets
●​ Handles packet loss/corruption – ensures data arrives error free

3. Internet Layer – handles transmission of packets

●​ Identifies the intended network and host

●​ Transmits packets to the Data Link
●​ Routes packets independently – through optimum route
●​ Addresses packets with their source and destination IP addresses
●​ Uses an IP address and port number to form a socket

4. Link Layer (Network Access Interface) – Handles how data is physically sent

●​ Ensures the correct network protocols are followed

●​ Enables the upper layers to access the physical medium (allows communication
with the network layer)
●​ Responsible for transporting data within the network – formats data into frames
for transmission
●​ Maps IP addresses to MAC addresses

TCP/IP Layer Interactions

 76

●​ Each layer can only accept input from adjacent layer (next higher or next lower
layer)
●​ There is an interface between the adjacent layers which the only interaction
between them
●​ Interactions are carried out by installed software
●​ User interaction takes place at the application layer of stack through protocols
●​ Direct access to hardware takes place at the Link layer of the stack

14.2 Circuit Switching, Packet Switching

Circuit Switching

●​ Data is transferred using a dedicated circuit/channel and implemented at the

physical layer
●​ Circuit is established before transmission starts
●​ Circuit lasts for for entire duration of transmission and closes after it ends
●​ Data is transferred using the whole bandwidth
●​ All data is transferred over the same route
●​ Transmission is generally bidirectional
 77

Use

●​ When a dedicated path needs to be sustained throughout a call

●​ Where the whole bandwidth is required or real time communication is used
●​ E.g. standard voice communications, video streaming, private data networks
●​ Suitable for long continuous communication

Advantages

●​ No need to reassemble data - frames arrive in same order in which they are sent
●​ Entire bandwidth is available
●​ Simpler and fast method of data transfer – data is transmitted with a fixed data
rate and follows the same path (means no data is lost or disordered)
●​ Fast data transfer rate - suitable for real time transmission

Disadvantages

●​ No other transmission can occur when circuit is in use - wasted bandwidth (cannot
be shared)
●​ Not very secure - can be intercepted more easily as all data travels along same
route
●​ Significant cost and time required to establish dedicated connection between
stations
●​ Only one route is available - if an error occurs, transmission ends
●​ Can take time to set up circuit before start of transmission
●​ Circuit is always there, even it is not being used

Packet Switching

●​ Implemented at the network layer

Use

●​ On digital data networks such as the internet – for sending large files that don’t
need to be live streamed
●​ When it is necessary to overcome faulty lines through rerouting
●​ For secure communication and high volume data transmission
●​ When the entire bandwidth isn’t required
 78

●​ E.g. emails, text messages, documents etc.

Transferring Messages Across Internet

●​ Data/message to be transmitted is divided into equal-sized packets (consist of a

header, payload and trailer)
●​ A packet header is attached to each packet containing key information
(source/destination IP addresses, packet number etc.)
●​ Each packet is dispatched independently and given its own route - the routing for a
packet depends on the network traffic
●​ Routes are determined using a routing table - optimum route is always taken
●​ Packets usually arrive out of order - are reassembled in the correct order at
destination (using sequence number in header)
●​ If packets are missing/corrupted a resend request is sent

Advantages

●​ Packets are more likely to arrive as they can be rerouted or retransmitted if an

error occurs
●​ Bandwidth can be shared - packets from multiple messages can share paths
●​ Secure - packets travel via different routes
●​ High data transmission rate is possible

Disadvantages

●​ Time delay occurs - packets need to be reassembled at destination, channel

bandwidth has to be shared with other packets, packets might need to be re-sent
●​ Requires a complex algorithm to function
●​ Needs lots of RAM to handle large amounts of data

Function of a Router in Packet Switching

●​ Router examines the packet’s header – reads the IP address of destination

●​ Has access to a routing table – contains info about the netmask/gateway used,
available hops and the status of the routes along route
●​ It decides on the next hop/best route and sends the packet on its next hop

Benefits

●​ Accuracy – Ensures accurate delivery of message

 79

●​ Completeness – Missing packets can be easily detected and a resend request sent
to message arrives complete
●​ Router can detect changes in networks and send data another way, ensuring it
arrives
●​ Allows simultaneous use of channel by multiple users
●​ Better security – packets are hashed and send by different routes

Drawbacks

●​ Time delays – due to correcting potential errors in packets caused by network

problems
●​ Requires complex protocols for delivery
●​ Unsuitable for real time transmission applications

15.​ Hardware and Virtual Machines

15.1 Processors, Parallel Processing and Virtual Machines

RISC (Reduced Instruction Set Computers)

●​ Uses few instruction sets & addressing modes

●​ Uses fixed length, single-clock cycle instructions
 80

●​ Instructions only require one clock cycle (single-cycle)

●​ Uses many general purpose registers and a hard-wired CPU
●​ Makes use of pipelining – executes instructions in parallel, where the output of one
instruction is the input of the next
●​ Makes extensive use of RAM
●​ Design emphasis is on software

Pipelining

●​ Allows several instructions to be processed simultaneously – used to increase

instruction throughput during FE cycle

Process ​

●​ Instructions are divided into subtasks

➔​ Instruction fetch
➔​ Instruction decode
➔​ Operand fetch
➔​ Opcode/Instruction execute
➔​ Result store
●​ Each subtask (instruction stage) is completed during one clock cycle
●​ No two instructions can execute their same stage at the same clock cycle
●​ The second instruction begins in the second clock cycle, while the first instruction
has moved on to its second subtask etc. – while one instruction is being executed,
the next one can be fetched and so on…

CISC (Complex Instruction Set Computers)

●​ Uses many instruction formats and has a large instruction set

●​ Uses variable length, multi-clock cycle instructions
●​ Makes use of many different addressing modes
●​ Uses few registers and a programmable CPU
●​ Design emphasis is on hardware – requires complex circuits
●​ Makes frequent use of cache memory

Interrupt Handling (In both RISC and CISC)

1.​ The processor detects an interrupt at the start/end of the FE cycle

2.​ The current program is temporarily stopped and the status/contents of each
register are stored on a stack
3.​ The appropriate ISR routine is called and executed
 81

4.​ After the interrupt has been serviced the registers are restored to their original
status (data is restored from the stack)

Effect of Pipelining (only in RICS)

●​ Adds additional complexity – there could be a number of instructions still in the

pipelining when interrupt is received
●​ All currently operating instructions are discarded except for the last one – the
interrupt handling routine is applied to the remaining instruction
●​ Once interrupt has been serviced the processor can restart with the next
instruction in the sequence

Computer Architecture

SISD – single instruction, single data

●​ One processor executes a single instruction using the same data set – data is
taken from a single source and only a single instruction is performed on it

SIMD – single instruction, multiple data

●​ Processors execute the same instruction on multiple different data sets

simultaneously
●​ Instructions can be performed sequentially (Pipelining)
●​ Parallel computers with multiple processors

MISD – multiple instructions, single data

●​ Performs different instructions/operations on the same set of data

●​ Each processor works on the same data set independently
●​ Parallel computers with multiple processors

MIMD – multiple instructions, multiple data

●​ Consists of many processors that operate asynchronously/independently

●​ Any processor can execute different instructions on different data sets

Massively Parallel Computers

●​ kind of network infrastructure

●​ a large number of computer processors or separate computers connected together
– simultaneously performing a set of coordinated computations
 82

●​ communicate using a message interface (send messages between each other)

Virtual Machines

●​ Are an emulation of a computer system (and hardware/software) using a host

computer system
●​ Uses guest operating system for emulation

Benefits

●​ cost saving – new system can be tried on different virtual hardware without the
need to purchase the hardware
●​ different instruction set architectures can be emulated on a single computer
●​ the system can crash without affecting the host machine – virtual machine
provides protection to other software
●​ security – if a virus is downloaded on the emulated system, it only affects the VM
●​ more than one new computer system can be emulated – allows multiple new
systems to co-exist on a single computer (multiple VMs can be used on the same
computer)
●​ allows emulation of programs for new CS that are not compatible with the host
operating system – through the use of a guest operating system
●​ can emulate old software on a newer system using a compatible guest OS

Limitations

●​ cannot emulate some hardware – new hardware might have been developed after
the VM was
●​ using the machine means extra code has to be executed and more load is put on
the host computer – is less efficient, increases processing time, uses more RAM
(hence has poorer performance)
●​ increases the maintenance expenses as both host system and the virtual machine
must be maintained (is also more complex to manage & implement)
●​ VM may be affected by weaknesses of the host machine

Roles of Host Operating System

●​ Is the normal OS used by the host machine

●​ Has control of all the resources of the host machine/computer (and can access all
the physical resources)
●​ Provides a user interface to operate the virtual machine software
 83

●​ Runs the virtual machine software

Roles of Guest Operating System

●​ OS that runs within the virtual machine

●​ Controls the virtual hardware/software during the emulation – accesses actual
hardware through the virtual machine and host OS
●​ Provides a user interface for the emulated hardware/software
●​ Runs under the control of the host OS

15.2 Boolean Algebra and Logic Circuits

Combination Circuits – output depends entirely on the input values

Half-Adder Logic Circuit

●​ Used to perform binary addition

●​ Outputs a sum bit and a carry bit

INPUT OUTPUT

X Y sum carry
0 0 0 0
0 1 1 0
1 0 1 0
1 1 0 1

●​ Consists of a XOR and AND gate – XOR outputs sum, AND outputs carry

Full-Adder Logic Circuit

●​ Two half-adders combined

●​ Has 3 inputs (X, Y and Cin) and 2 outputs (S and Cout)
 84

INPUT OUTPUT

X Y Cin sum Cout

0 0 0 0 0

0 0 1 1 0

0 1 0 1 0

0 1 1 0 1

1 0 0 1 0

1 0 1 0 1

1 1 0 0 1

1 1 1 1 1

●​ Two half-adder circuits whose outputs are joined through an OR gate

Sequential Circuits – the output depends on the input value produced from a
previous output value

SR Flip-Flop
 85

●​ stores a single bit of data – 0 or 1

●​ used in memory to create storage cells within RAM – each flip-flop stores a data
bit (memory is made up of many SR flip-flops)

​ Problems of SR Flip-Flops

●​ potential for circuit to arrive in an uncertain state

●​ circuit can become unstable if inputs do not arrive simultaneously
●​ Consists of 2 inputs & outputs
➔​ “SET” labelled S
➔​ “RESET” labelled R
➔​ ‘Q’ and ‘NOT Q’ as outputs

●​ Made up of 2 cross-coupled NOR or NAND gates

1. Using NOR Gates

​ ​ ​ ​ ​ ​ ​ If both S and R = 1, an invalid condition

occurs as Q and NOT Q will be equal.

(both SET and RESET cannot be done at once)

Any other combination is valid. If both

S and R = 0, there is no change.

​ * when assigning values, S and NOT Q will be the same, and R and Q will be the
same (aim is to get the gate inputs to be either 00 or 11)

INPUT OUTPUT

S R Q NOT Q

1 0 0 1

0 1 1 0
2. Using NAND Gates

If both S and R = 0, an invalid condition

occurs as Q and NOT Q will be equal.

(both SET and RESET cannot be done at once)

 86

Any other combination is valid. If both

S and R = 1, there is no change.

​ * when assigning values, the same applies; S and NOT Q will be equal, R and Q
will be equal

​ (truth table is the same, only difference is when no change or invalid state occurs)

JK Flip-Flops

●​ improved version of SR flip-flops – eliminates invalid input combinations

●​ used to produce shift registers in a computer or a simple binary counter\

Advantages

●​ validates all input combinations

●​ avoidance of unstable states and increased stability

●​ Uses a clock pulse as an input (to synchronise inputs) and adds additional gates

When J and K = 0, there is

no change to Q.

When J and K = 1, the value of

Q toggles after each clock pulse.

(flip-flop changes between SET and RESET state)

* the output value of Q will always be equal to J, NOT Q will be equal to K

 87

Boolean Algebra

●​ form of algebra linked to logic gates

●​ based on TRUE and FALSE statements

​ Boolean Notation​

Logic Gate Notation

AND A.B

OR A+B

NOT

NAND

NOR

De Morgan’s Laws

Karnaugh Maps

●​ method of obtaining a simplified boolean algebra expression from a truth table

​ Advantages

●​ minimises number of boolean expressions

 88

●​ minimises number of logic gates – provides a more efficient circuit

Rules of Simplification

●​ works by grouping together ​adjacent cells containing 1s

➔​ groups cannot be diagonal – they can wrap around

➔​ only 2n cells in each group (e.g. only 2/4/8… 1s allowed in one group)
➔​ each 1 must be in at least one group – overlapping of groups is allowed
➔​ groups should be as large as possible
➔​ there should be the fewest amount of groups possible

Example 1:

Example 1:

16.​ System Software

16.1 Purpose of an Operating System

Maximising Use of Resources

 89

●​ implements process scheduling – increases efficiency of CPU

●​ manages main memory
●​ optimises input/output operations – direct memory access (DMA) allows
hardware to access main memory independently of the CPU, meaning the CPU is
not utilised and free to carry out other tasks during data transfer

User Interface

●​ hides the complexities of the computer/hardware/operating system from the user

●​ provides appropriate access systems for users with differing needs
●​ avoids complex commands involving memory locations or computer hardware
●​ e.g. a graphical user interface uses icons for navigation

Process Management

Multi-tasking

●​ Allows computers to (appear to) carry out more than one process at a time

​ Implementation

●​ Processor time, hardware and resources are shared between tasks

●​ Scheduling is used to decide which process is carried out next - endures
multi-tasking operates efficiently
●​ One task of a higher priority can interrupt another currently running task

Interrupt Handling

●​ transferring control to another routine when a service is required

Process States

●​ Ready – program is waiting to run on the CPU for allocated amount of time
●​ Running – program is currently running/being executed on the CPU
●​ Blocked – program execution is halted, waiting for an event to occur

Process Changes

●​ Running → Ready
➔​ The time slice of the running process expires
➔​ There is a process with a higher priority in the ready queue (the running
process gets preempted)
 90

➔​ An interrupt arrives at the CPU – the process running on the CPU gets
preempted
●​ Ready → Running
➔​ Process is selected by CPU to be executed – the OS scheduler has
allocated this given time to the process
●​ Running → Blocked
➔​ Process needs to carry out an input/output operation or wait for a
resource to be available
➔​ Process cannot proceed with execution before a specified event takes
place
●​ Blocked → Ready
➔​ Event or input/output operation has occurred, program can resume

Scheduling – managing the processes running on the CPU

●​ Allows more than one task to appear to be executed at the same time (enables
multi-tasking)
●​ Allows high priority jobs to be completed first
●​ Keeps the CPU busy at all times – this ensures all processes execute efficiently
and reduces wait times for all processes

Scheduling Routines

Shortest Job First

●​ Short processes are executed first and followed by longer processes (executed in
ascending order of CPU time required)
●​ Leads to an increased throughput – as more processes can be executes in a smaller
amount of time

Round Robin

●​ Each process is served by the CPU for a fixed time

●​ Starvation doesn’t occur – as each process is given a fixed time to be executed
every round robin cycle

First Come First Served

●​ No complex logic – each process is executed one by one

●​ Received processes are queued
●​ Starvation doesn’t occur – every process will eventually get a chance to run

Shortest Remaining Time

 91

●​ Processes are placed in queue (consists of a ready and blocked queue)

●​ If a process with a shorter execution/burst time is queued, the process currently
in line is preempted and the shorter process runs first
●​ Processes will run until executed or a process with a shorter execution time is
added
●​ Starvation may occur – if processes with a shorter burst time keep queuing, then
the longer processes may never run

How the OS Kernel acts as an Interrupt Handler

●​ Receives a signal when an interrupt is generated

●​ Checks priority of interrupt and reviews status of current interrupts
●​ Consults the interrupt dispatch table (IDT) - saves the contents of the registers on
the kernel stack
●​ Restored the register contents once the interrupt is serviced

Virtual Memory

●​ is created temporarily
●​ secondary storage is used to simulate additional main memory
●​ extends RAM – means the CPU appears to be able to access more memory space
than the actual RAM available
●​ only data in use needs to be in main memory – data can be swapped between
RAM and virtual memory as necessary

Reasons for Use

●​ when RAM is running low – e.g. when a computer is running many processes at
once
●​ for more efficient use of RAM – if programs are not immediately needed, they can
be moved from RAM to virtual memory

Paging – reading/writing blocks of data from/to secondary storage when required

●​ memory is divided into fixed size blocks

●​ dividing of memory into pages done by the operating system
●​ faster access times than segmentation

Segmentation
 92

●​ In segmented memory, the virtual address space is broken into varying sized
blocks (sections) - segment size is calculated by the compiler
●​ Each segment has a name, size and is numbered - its number is used as an index in
the segment map table
●​ During execution segments from virtual memory are loaded into physical memory
●​ Address is specified by the user - contains the segment name and offset value
●​ An offset value determines the size of the segment
●​ Segment map table maps virtual addresses to physical addresses (contains
segment number and offset value)
●​ Slower access times than paging

Page Replacement

●​ ‘Page Fault’ – an interrupt raised by hardware that occurs when a requested page
is not yet loaded into main memory
●​ The OS replaces an existing page with a new one – done by swapping pages
between secondary and main/primary memory

FIFO Algorithm (First In, First Out)

●​ OS keeps track of all pages in memory using a queue structure

●​ The oldest page is at the front of the queue and is the first to be removed when a
new page is added
●​ Page usage is not considered during replacement
●​ Suffers from Belady’s Anomaly – more page faults with an increasing number of
page frames

LRU Algorithm (Last Recently Used Replacement)

●​ Replaces the page which has not been used for the longest time
●​ Implemented using a linked list consisting of all pages in memory – most
recently used page is at the front, least recently used at the back

​ Clock Page Replacement

●​ Uses a circular queue structure – has one single pointer that acts as a head and
tail
●​ Has an R-flag which can be 0 or 1
➔​ If R=0 – the page being pointed to is removed and a new page is inserted
into its place , else a new page is inspected
 93

➔​ If R=1 – the next/following page is inspected, the pointer moving until a

page where R=0 if found

Disk Thrashing

●​ Problem that occurs when virtual memory is being used (due to frequent transfers
between virtual and physical memory)
➔​ As main memory fills up, more pages need to be swapped between virtual
and physical memory
➔​ This swapping leads to a very high rate of hard disk access - results in
excessive head movements
➔​ Latency increases as hard disk head movements take a relatively long time
➔​ Eventually, more time is spent swapping the pages/data than processing it -
program may freeze or not run

16.2 Translation Software

Interpreter

●​ examines source code one statement at a time

●​ checks each statement for errors
➔​ …if no error is found, the statement is executed
➔​ …if an error is found, it is reported and the interpreter halts
●​ interpretation is repeated for every iteration in loops
●​ interpretation has to be repeated every time the program is run

Compilation Stages

Lexical Analysis
●​ converting a sequence of characters into a sequence of tokens

Syntax Analysis

●​ uses parsing algorithms to interpret the meaning of a sequence of tokens

●​ checks code matches the grammar of the language
●​ syntax errors are reported
●​ a parse tree is produced

Code generation
●​ converting an intermediate representation of source code into an executable form

Optimisation
 94

●​ minimising a program’s execution time and memory requirement

Backus-Naur Form Notation

●​ Is a formal method of defining the grammatical rules of a programming language

​ <variable_name> ::= option1 | option2 | option3 … ;

➔​ <> is used to enclose an item

➔​ ::= is basically a = or ←, assigning options to a variable
➔​ | means OR – allows a variable to have multiple choices

​ Example
●​ To define a hexadecimal number, the following definitions can be made:
​ <digit> ::= 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 0 ;
​ <letter> ::= A | B | C | D | E | F ;
​ <hexnum> ::= <digit> | <letter> | <hexnum> <digit> | <hexnum> <letter>

➔​ This states that only digits between 0-9 and letters from A-F are valid
➔​ Many BNF definitions are recursive, allowing the hexadecimal number to
have multiple digits and letters

Reverse Polish Notation

●​ Used to carry out evaluation of expressions

➔​ Provides an unambiguous method of representing an expression
➔​ Reads from left to right
➔​ Doesn’t require brackets or rules of precedence (BODMAS)

Data Structures used to evaluate in RPN

●​ Stack – operands are popped from stack in reverse order to how they were pushed
●​ Binary Tree – allows both infix and postfix to be evaluated (tree traversal)

​ Infix vs Postfix

●​ Infix refers to regular algebraic equations e.g. 3 + 4 evaluates to 7

●​ Postfix is used in RPN e.g. 3 4 + evaluates to 7
➔​ When writing postfix notation, the operator comes after the two numbers it
is being performed on (2 4 * → 8 and 15 3 / → 5)
➔​ Brackets are not used – 5 3 + 7 2 - * evaluates to (5 + 3) * (7 - 2)
 95

➔​ Numbers are grouped with the operator to their right – 8 - 2 3 * is 8 - 2*3

RPN evaluation using stacks

●​ Expression is read from left to right, one item at a time

●​ Each element is checked to see whether it is an operator or a value
●​ Values are pushed onto a stack until an operator is found
●​ Operator is then applied to the last two values on the stack
●​ Result is pushed back onto the stack
●​ Process repeats until only one value remains (the result)

​ Eg. a b * 2 / c d / * (where a = 20, b = 3, c = 10 and d = 5)

3 2 10 10 2

20 20 60 60 30 30 30 30 60

17.​ Security

17.1 Encryption, Encryption Protocols and Digital Certificates

Key Cryptography

●​ Ensures a message is authentic/from a trusted source

●​ Ensures message has not been altered during transmission
●​ Makes sure only intended receiver is able to understand a message
●​ Non-repudiation – neither sender or receiver can deny the transmission happened

Asymmetric Encryption
 96

●​ Provides better security – by using a pair of different keys

●​ One of the keys is used to encrypt the message, the matching one is used to
decrypt it
●​ Only the public key is available to everyone, the private key is kept secret
●​ Is a longer process, as it is more complex
●​ Length of keys is longer (usually 2048 bits)

Process

●​ The receiver’s computer uses an algorithm to generate a matching pair of keys

(public and private)
●​ The public key is sent to the sender’s computer
●​ The sender encrypts the document/file/data using the public key to create cipher
text
●​ The sender’s computer sends the cipher text to the receiver’s computer – can only
be decrypted using the receiver's private key

Detecting Alterations

●​ The message, together with the digital signature, is decrypted using the receiver’s
private key
●​ The digital signature received is decrypted with the sender’s public key to recover
the digest sent
●​ The decrypted message received is hashed with the agreed hashing algorithm to
reproduce the message digest received
●​ The two digests (received and reproduced) are compared – if they are the same,
then the message has not been altered

Symmetric Encryption

●​ Uses a single key which is used/shared by all to encrypt and decrypt messages
●​ Simple process that can be carried out quickly - risk of compromise is higher
●​ Length of keys is shorter (usually 128/256 bits)

Quantum Cryptography

●​ Protects security of data transmitted over fibre optic cable

●​ Is a virtually unhackable encryption system

Benefits
 97

●​ Detects any eavesdropping (due to change in photon properties)

●​ Once transferred, the integrity of the key can be guaranteed – it cannot be copied
or decrypted later
●​ More secure, longer keys can be used

Limitations

●​ Limited range – works only over relatively short distances

●​ Requires a dedicated fibre-optic line and specialist hardware – is expensive
●​ Polarisation of light may be altered during transmission through fibre-optic cable
●​ Lacks many vital features and has high error rates (still new and being developed)
– e.g. digital signatures, certified mail etc.

Private Key

●​ An unpublished/secret key that is never transmitted anywhere

●​ Has a matching public key
●​ Is used to decrypt data that was encrypted with its matching public key

Purpose and Example Situations of SSL/TLS

Secure Socket Layer

●​ Encrypts data being transferred over the internet – decides on which encryption
algorithms are to be used
●​ Performs data integrity checks and data compression

Transport Layer Security

●​ Provides encryption, authentication and data integrity in a more effective way
than SSL
●​ Ensures privacy and security of data between devices communicating over the
internet – provides third-party eavesdropping

Examples of where SSL/TLS would be used

●​ Online banking and online financial transactions

●​ Sending and receiving emails, sending software to a restricted list of users
●​ Using cloud storage facilities or a VPN

SSL/TLS use when Client-Server Communication is Initiated

●​ A SSL/TLS connection is initiated by an application which becomes the client

●​ The application which receives the connection becomes the server
●​ Every new session begins with a handshake
 98

●​ A digital certificate is requested/sent from the client/server

●​ The client verifies the server’s digital certificate and obtains the server’s public key
●​ The encryption algorithms are agreed upon by the server and client – the
symmetric session keys are then generated

Digital Certification

Digital Signatures

●​ An enquiry is made to Certificate Authority (CA)

●​ The enquirer sends their public key and all required information (e.g. to prove
identity) to CA
●​ The enquirer’s details are checked by the CA – If details are verified then the public
key is agreed upon
●​ The CA creates/issues a certificate that includes the enquirer’s public key
●​ Encrypting data is sent to the CA using their public key and sent by the CA using
their private key

How it’s produced before a message is sent

●​ Message is hashed using the agreed hashing algorithm to produce a digest

●​ The message digest is then encrypted with the sender’s private key to form the
signature

Digital Certificate

●​ An electronic/online document used to authenticate the identity of a

website/individual/organisation – typically issued by the CA
●​ Contains information for identifying an individual/website owner and a public key
●​ Provides the public key which can be used to validate the private key associated
with the signature
18.​ Artificial Intelligence (AI)

18.1 Artificial Intelligence

How graphs aid AI

●​ Artificial Neural Networks can be represented using graphs

●​ Graphs provide structures for relationships
●​ AI problems can be defined as finding a path in a graph
●​ Graphs may be analysed by a range of algorithms e.g. Dijksta’s algorithm

Purpose of A* and Dijkstra’s Algorithms

 99

●​ To find the optimal shortest and most cost-effective route between two nodes
based on distance/cost/timeArtificial Neural Network
●​ Part of AI that is meant to simulate the function of a human brain
●​ Key component of machine learning
●​ Have self-learning capabilities – enables production of better results as more data
becomes available
●​ Can solve complex problems humans cannot/find it difficult to

Artificial Neural Networks

Multiple Hidden Layers in an Artificial Neural Network

●​ Enables deep learning to take place

●​ Needed when the problem you are trying to solve has a higher level of complexity
(requires more layers to solve)
●​ Enables the neural network to learn and make decisions on its own
●​ Improve accuracy of results – more hidden layers means more complex learning
capabilities

How Artificial Neural Networks Enable Machine Learning

●​ They are intended to replicate the way human brains work

●​ Weights/values are assigned between nodes – adjusted through training to give
more accurate results
●​ The output layer provides the results
●​ Data is input at the input layer and is passed into the system
➔​ it is then analysed at each subsequent (hidden) layer where characteristics
are extracted/outputs are calculated
➔​ reinforcement learning takes place through repeating the training/learning
process
●​ Decisions can be made without being specifically programmed
●​ The deep learning net will have created complex feature detectors
●​ Back propagation (of errors) will be used to correct any errors that have been
made.

Deep Learning

●​ simulates data processing abilities of the human brain to make decisions

 100

●​ uses artificial neural networks which are modelled after the human brain –
structures algorithms in layers – an input layer, output layer and many hidden
layers
●​ uses large number of hidden layers to progressively extract higher level features
from the raw input (has more success)
●​ is a specialised form of machine learning
●​ trained using large amounts of unlabeled data

Reasons For Use

●​ makes good use of unstructured data

●​ outperforms other methods of machine learning if the data size is large
●​ enables machines to process data with a nonlinear approach
●​ is effective at identifying hidden patterns (ones that are too complex for humans to
spot/are undetectable)
●​ can provide more accurate outcome with higher numbers of hidden layers

Reinforcement Learning

●​ based on feedback – AI learns in an interactive environment through actions and

seeing the results of each action (works through trial and error, learns through its
own experiences)
●​ for each good action, the AI gets positive feedback, for each bad one it receives
negative feedback
●​ Node weightings are adjusted to achieve correct outcome – the AI uses feedback to
improve its performance in similar tasks

Reasons for Use

●​ enables autonomous learning using feedback without any labelled data

Machine Learning

Supervised Learning

●​ allows data to be collected/a data output to be produced from a previous

experience
●​ a known input and associated outputs are given – uses sample data with known
outputs (labelled input data)
●​ able to predict future outcomes based on past data

Unsupervised Learning
 101

●​ helps all kinds of unknown patterns in data to be found – enables learning by

allowing the process to discover previously undetected patterns on its
●​ only requires input data to be given
●​ uses any data – not trained using a right output (uses unlabeled input data)

19.​ Computational Thinking and Problem-solving

19.1 Algorithms

Binary Search

●​ Necessary condition – the elements in the list being searching must be

ordered/sorted in ascending/descending order
●​ The time to search a list increases with an increasing number of items in the list
●​ Starts in the middle of the array/list
●​ Works by finding the mid-point of an array/list and determines which side contains
the item to be found – it discards the half of the array/list not containing the search
item

Process
 102

●​ Find middle index/item

●​ Check the value of the middle index in the list
●​ The item searched for has been found if the index value is equal to it
●​ Otherwise, discard half the list that does not contain the searched item
●​ Repeat the previous steps until the item searched for is found, or there is only one
index left in the list and it is not the item searched for

Linear Search

●​ sequentially checks each element of the array/list until the matching element is
found, or the end of the array/list is reached.
●​ does not require the elements to be sorted.
●​ will usually do more comparisons of records/iterations against the target (before
finding it) than a binary search
●​ starts at the beginning of the array/list.

​ Binary VS. Linear Searches

●​ time to search increases linearly in relation to the number of items in the list for a
linear search and logarithmically for a Binary search
●​ time to search increases less rapidly for a binary search and time to search
increases more rapidly for a linear search

Big O Notation

●​ is used to indicate the time/space complexity of an algorithm.

●​ Linear search → O(n)
●​ Binary search → O(log 2n) / O(Log n)

●​ O(log n) is a time complexity that uses logarithmic time – the time taken goes up
linearly as the number of items rises exponentially

Performance of a sorting algorithm is affected by…

●​ Initial order of data

●​ Number of data items to be sorted
●​ The efficiency of the sorting algorithm

Queue
 103

●​ Uses FIFO data structure – data is removed in the order it is received

●​ Is of varying length
●​ Data is ‘enqueued’ and ‘dequeued’ at different ends
●​ Has two movable pointers

Stack

●​ Uses FILO data structure – data is removed in the reverse order to which is it
received
●​ Is of varying length
●​ Data is popped and pushed onto/off a stack from the same end
●​ Has one movable pointer

Uses of a Stack

●​ Recursion
●​ Implementation of ADTs
●​ Procedure calls
●​ Interrupt handling
●​ Evaluating an RPN expression

Linked List Implementation

●​ Is a dynamic data structure (not restricted in size)

●​ Has freedom to expand or contract – by adding/removing nodes as necessary
●​ Allows more efficient editing using pointers than an array

Array Implementation

●​ Is a static data structure, generally fixed in size

●​ When array is full, stack cannot be extended any further

19.2 Recursion

●​ A process using a function/procedure that calls itself

●​ Must have a base case (stopping condition)
●​ Must have a general case – which calls itself recursively, changes its state and
moves towards the base case

Why Stacks are Effective for Implementation

●​ Stacks have a LIFO data structure

●​ Each recursive call is pushed to the stack and is then popped as function ends
 104

●​ Enables backtracking/unwinding to maintain the required order

​ Winding and Unwinding

●​ Winding – the statements following a recursive function call are not executed
until the general case has reached base case
●​ Unwinding – occurs once the base case has been reached

​ Benefits of Recursion

●​ Can require fewer programming statements than an iterative approach (shorter

and more effective code)
●​ Can solve complex problems in a simpler way than an iterative solution could /
provides simpler solutions to problems

Drawbacks

●​ Very repetitive and requires more storage space (demanding use of stack),
meaning stack overflow can occur
●​ Infinite recursion can occur

Translating Recursive Programming Code

●​ The compiler must produce object code to

➔​ Push return addresses/values of local variables onto a stack with each
recursive call
➔​ Pop return addressed/values of local variables off the stack after the base
case is reached
20.​ Further Programming

20.1 Programming Paradigms

Programming Paradigm

●​ A set of programming concepts

●​ Defines the structure and capability of a programming language

Low-Level

●​ Uses the instruction set of a processor

●​ Makes use of different addressing modes – immediate, direct, indirect, indexed and
relative

Imperative
 105

●​ Uses variables – changed using assignment statements, rely on an

iterative/repetitive method
●​ Statements provide an explicit sequence of commands for the computer to perform
(in the order written) – each line of code changes something in the program being
run, commands update the program state

Object Oriented

●​ Uses the concept of class, inheritance, encapsulation and polymorphism

Object-Oriented Programming Terminology

Properties/Attributes

●​ The data items or data types defined in a class

●​ Private attributes
➔​ Enforce encapsulation (ensures they are hidden)
➔​ Ensure attributes are only accessible using the class’s own methods/within
the class

Methods
●​ the modules (e.g. functions) in a class implementing the behaviours that act on the
attributes (properties)

Inheritance
●​ methods and properties contained in one class are reused/made available to a
derived class

Polymorphism
●​ allows methods to be redefined (take on different behaviours) for derived classes

Encapsulation
●​ process of putting properties and methods inside a class (ensures sensitive data is
hidden) together as a single unit

Getter
●​ method that is used to return the value of a property

Setter
●​ method that is used to update/set the value of a property

Instance
●​ an occurrence of an object
 106

Structure/Contents of a Class

●​ Attributes/properties and their respective data types - these are variables bound to
the class
●​ Methods - subroutines that act upon the attributes
●​ Getters and Setters - methods that fetch/update the attribute contents
●​ Constructor - used to create instances of objects in the class

Object vs. Class

●​ The class is only defined once, but many objects can be created from it - acts as a
template from which objects are created
●​ Objects are allocated memory space when created
●​ Classes are not allocated memory space when defined
●​ A class cannot be manipulated, but objects can (a class is not available in memory)
●​ Classes are defined, objects are created/declared
●​ Classes can use inheritance, objects cannot

Declarative

●​ Instructs a program on what needs to be done instead of how to do it – specifies

desired result rather than method of acquiring it
●​ Uses facts, rules and queries to satisfy goals
●​ Can be logical (states program as set of logical relations) or functional
(constructed by applying functions to arguments)

20.2 File Processing and Exception Handling

Exception Handling Routine

●​ Responds to unexpected events while a program is running – prevents programs

from halting unexpectedly
●​ Traps runtime errors
●​ Produces meaningful error messages

Exception

●​ Unexpected event that occurs during the execution of a program

●​ Causes the program to halt execution
●​ Examples of Exceptions – end of file, programming/user error, hardware failure,
division by zero (runtime errors)