h2cp definitions + processes

describe bubble sort

1. compare first and second values. if the first value is larger than second value, swap them around

2. compare the second and third values. if the second value is larger than the third value, swap them around.

3. keep comparing adjacent values, swapping them around if necessary, until the last two values in the list have been processed

4. when we have completed the first pass through the entire array, the largest value is in the correct position at the end of the array. the other values may or may not be in the correct order.

5. this repeats until no more swaps occur and the array is sorted.

   
   

   
   

describe insertion sort

1. To sort an array in ascending order using insertion sort, we start with the first element in the array. One element by itself is already sorted. 

2. Then we consider the next element in the unsorted array. If it is smaller than the first element, we insert this element to the left of the first element, else we place it to the right. Next, we consider the third element in the array. 

3. We make comparisons with the elements in the sorted array leftwards until we find the correct position to insert this element into the sorted array. 

4. Repeat this for the remaining elements, until the whole array is sorted.

describe merge sort

1. Merge sort recursively divides an unsorted array into subarrays, until each contains one element. 

2. Thereafter, the algorithm repeatedly pairs up the subarrays and merge each pair into a new sorted subarray, using the above process of merging arrays, 

3. until there is only one subarray remaining and that will be the sorted array.

describe quicksort

1. A pivot element is selected from the array

2. The array is partitioned into two sub-arrays; elements less than the pivot, and elements greater than the pivot

3. The process is applied recursively to each sub-array

4. The base case is reached when a sub-array has zero or one element, and is thereby sorted

5. The sub-arrays are then concatenated in order – less than, pivot, greater than – to produce the final sorted array

differences between merge sort and quicksort

stability:

• merge sort — stable algorithm, it does not change the occurrence of similar elements (equal elements are ordered in the same order in the sorted list). this preserves the order which these elements appeared in the original array

• quicksort — unstable sorting technique, might change the occurrence of similar elements

memory:

• in-place quicksort — does not need additional memory space to sort

• merge sort — requires a temporary array to merge sorted arrays, hence needing additional memory space

time complexity:

• quicksort — worst time complexity O(n²)

• merge sort — worst time complexity O(nlogn)

linear search — pros + cons

pros

• easy and simple

• does not need a sorted array

cons:

• slow, O(n)

binary search — pros + cons

pros:

• fast, O(logn)

cons:

• needs a sorted array

closed addressing/open hashing/separate chaining — pros + cons

can be implemented using linked lists. when multiple elements are hashed to same index, these elements can be inserted into a singly-linked list

pros:

• easy implementation

• ensures all elements can be found in hash table, just as a key in the linked list

cons:

• inefficient use of memory as some addresses might never get used at all but still have memory allocated

• extra memory allocation to store elements as nodes in linked list

• worst time complexity O(n)

open addressing/closed hashing/linear probing — pros + cons

when inserting data into a hash table and collision is detected, search the next empty location by looking at next free index until empty location is found. hash table is full when there are no more locations

pros:

• highly efficient in memory as all allocated memory locations will be used. no extra memory is required

cons:

• due to clustering, searching is slower. worst case, when the hash table is full, all indexes need to be checked to see if the value exists or not

traits of a good hash table

• space of hash table — hash table should be 1.5x the maximum data size, this will reduce the number of collisions

• deterministic hash function — for same value, hash function should generate same hash value

• efficient to compute — hash function should be quick to compute hash value

• uniform distribution — hash function should provide a uniform distribution across the hash table to minimise collisions

• usage of all data inputs — ignoring parts of the data can lead to collisions

static vs dynamic memory allocation

Static Memory Allocation	Dynamic Memory Allocation
Memory is allocated at compile time.	Memory is allocated at run time.
In static memory allocation, the size and location of memory blocks are fixed and cannot be changed at runtime.	In dynamic memory allocation, the size and location of memory blocks can be changed according to the program logic and data size.
Allocated memory remains from start to end of the program.	Allocated memory can be released at any time during the program.
It is inflexible and wasteful if actual memory usage is less than allocated.	It is flexible and efficient as memory is allocated according to required usage
It is fast and simple, as there is no need to allocate or deallocate memory during execution, and it avoids memory fragmentation, as the memory blocks are contiguous and aligned.	It is slower and more complex, as you have to manage the memory allocation and deallocation yourself. It may also cause memory fragmentation.
Static memory allocation is preferred in an array.	Dynamic memory allocation is preferred in the linked list.

circular vs linear queue

utilisation of space:

• circular — utilises array space efficiently. when the queue reaches the end of the array, it wraps around to the beginning, hence utilising unused slots at front of array

• linear — once a dequeue occurs, space previously occupied cannot be utilised.

speed

• circular — data packets can be enqueued as long as there is unused space

• linear — when queue is full, enqueuing must pause until queue is fully dequeued before it can be used again, slowing performance

deletion of a node with 2 children

• look for next larger value (in-order successor). next larger value can be obtained by finding the minimum value in the right child of the node

• replace the contents of the node with next larger value and delete node of next larger value

traversal orders

pre-order — current, left, right

in-order — left, current, right

post-order — left, right, current

time complexity of BST

best time complexity O(logn)

worst time complexity O(n) — when tree is unbalanced or skewed

recursion — pros + cons

pros:

• more elegant and less program code

• complex tasks can be broken down into simpler sub-problems

• when designing a solution to a mathematical problem that is recursive by nature, recursive solutions are easier to implement

cons:

• repeated recursive calls carry large amounts of memory usage and processor time from multiple function calls

• if recursion continues for too long, computer may run out of memory and the program will crash

use of stacks in recursion

• A certain amount of memory is set aside for a function to use (e.g. storing local variables)

• Function’s memory address + Contents of local variables → stored as a stack frame

• Each recursive call pushes another stack frame onto the stack, until the base case is reached.

• When the base case is reached, the top stack frame is popped, which restores the local variables and returns control to the previous caller.

• This popping process continues, destroying the stack frames one by one, until control finally returns to the original function caller.

data validation

a process of ensuring that the input data supplied to a system satisfies a set of rules such that it is sensible, complete, and within acceptable boundaries. Its purpose is to avoid data errors. It does not guarantee that data is accurate.

data verification

the process of getting the user to confirm that the data entered was what was intended to be entered.

transcription error

a single incorrect character being entered, such as a wrong password or an invalid email address.

transposition error

where two digits are accidentally swapped.

types of validation checks

Range check	Checks that a value falls within the specified range E.g. the marks of the A level H2 Computing Paper 1 must fall within the specified range of 0 to 100 and not anything less than 0 or more than 100 marks
Length check	Checks the data isn’t too short or too long E.g. a telephone number must have a length of 8 numbers
Presence check	Checks that data has been entered into a field E.g. ensure that an entry field is not left blank
Check digit	An extra digit added to the end of a code number, which has been calculated from the digits of the code number. This ensures the code number has been entered properly.

types of verification checks

Double entry — the process of entering data twice, with the second entry being compared with the first is to ensure that it is accurate. It is common in batch processing for a second data entry operator to key in a batch of data to verify it. Another example is when setting a password, a user would be asked to key in the password a second time to ensure that the passwords match up. 

Proofreading data — someone checking the data entered against the original source to ensure that the data match.

types of errors

• Syntax error — Does not follow the rules of the programming language

• Runtime error — Program crash when the code is running

• Logic error — Produce wrong output without crashing

types of test data

Normal / Valid	Typical data values that are valid and will be accepted by the system.
Abnormal / Invalid / Erroneous	Data values that the system should not accept or should be rejected.
Extreme / Boundary	Data values that are at the extreme end of the range of normal data that should be accepted.

encapsulation

the feature of combining attributes and methods together in a single class.

• Encapsulation supports information hiding through the combining of private properties and public methods into a class, ensuring private properties are only accessed/altered by calls to public methods. 

• It also supports implementation independence, which is the use of methods of a class without needing to know how it is implemented. Even if the underlying implementation were to change, it does not matter to the user.

inheritance

• Attributes and methods in a subclass are acquired from a base class (aka superclass/parent).

• This reduces the need for duplicate code and promotes software reuse as the inherited attributes and methods need not be written again.

polymorphism

• when subclasses define methods with the same name as the methods in base class, but with different implementations.

• This allows the method in the derived class to override the one in the base class, so that each derived class can behave differently even when using the same method name.

• Polymorphism enables code generalization by allowing objects of different classes to be treated uniformly, providing flexibility and extensibility in code design.

class diagram layout

unicode vs ascii

amount of characters

• ASCII — only handles 128 basic English characters and symbols.

• Unicode — supports over 149,000 characters across more-than 150 scripts

primary key

a field or a set of fields in a table whose values uniquely identify each record in a table and should not change over time.

secondary key

an additional key, or alternate key, which can be used in addition to the primary key to locate specific data.

composite key

a combination of two or more fields in a table that can be used to uniquely identify each record in a table. Uniqueness is only guaranteed when the fields are combined.

foreign key

an attribute (field) in one table that refers to the primary key in another table. It links to a primary key in the other table and forms a relationship between the tables.

prevents invalid data from being inserted into the foreign key column, because it has to be one of the values contained in the referenced table

data redundancy

refers to the same data being stored more than once

Having to enter data multiple times means the database is at an increased risk of having inaccurate data.

It would be hard to update because every occurrence of the data item will need to be changed. This can lead to data inconsistency.

normalisation

1NF	All columns must be atomic. This means there can be no multi-valued columns – i.e., columns that would hold a collection such as an array or another table. In other words, the information in each column cannot be broken down further.
2NF	In 1NF Every non-key attribute must be functionally dependent on the entire primary key. Attribute Y is functionally dependent on attribute X (usually the primary key), if for every valid instance of X, the value of X uniquely determines the value of Y. This means that for every non-key attribute, the entire primary key uniquely determines the value of that attribute, i.e. no attribute can depend on only part of the primary key.
3NF	The table should be in 2NF The table should not have transitive dependencies – all fields must only be determined by the primary/composite key, not by other non-key attributes A functional dependency X => Z is said to be transitive if there exists an attribute Y such that X => Y and Y => Z Note that X => Y does not necessarily imply the converse Y =>X.

sql vs nosql

	Relational Databases (SQL)	NoSQL Databases
Language	Use of structured query languages to perform operations.	Use of a dynamic schema to query data. Also, some NoSQL databases use SQL-like syntax for document manipulation
Data Schema	Fixed, predefined schema where data must fit into tables with specific columns and data types. This rigid structure ensures consistency, and it works well for applications with stable, well-structured, and predictable data requirements.	Adopt flexible data models, allowing for dynamic and non-schematic data storage. This flexibility enables developers to insert data without a predefined schema. NoSQL databases are most useful in scenarios where data structures may be undefined, not fully known in advance, or are subject to frequent changes.
Scalability	Typically rely on vertical scaling, which involves improving and adding resources, such as faster processors and more memory, to the same server to handle increased load. Such high-performance components can be expensive, and upgrades are limited by the capacity of a single machine.	Typically use horizontal scaling, which is achieved by adding more servers or nodes to a distributed system, which then helps increase capacity. The nodes communicate with each other and distribute the load, so adding more nodes helps increase the overall capacity of the system.  This is a more scalable and cost-effective solution for managing a growing database and increasing database traffic.
Support for Big Data	The vertical scaling makes it difficult for relational databases to store very big data.	The horizontal scaling and dynamic data schema make NoSQL suitable for big data.
Properties	Use the ACID (Atomicity, Consistency, Isolation, Durability) property. ACID properties ensure immediate and strict consistency in the database. SQL queries guarantee that either all or none of the changes made during a transaction are committed to the database and have rules for how to handle concurrent transactions and unexpected events.	Settles for eventual consistency. They emphasize scalability and distributed architectures.  Eventual consistency acknowledges that, in a distributed system, it may take some time for all nodes to converge to a consistent state after an update. While NoSQL databases sacrifice immediate consistency for scalability and fault tolerance, they ensure that, given enough time, all replicas of the data will eventually converge to the same state.

advantages of nosql

Flexibility	Having a flexible data model also means NoSQL databases can address large volumes of rapidly changing data, making them great for agile development, quick iterations, and frequent code pushes. Suitable when there are extra data requirements that are not clear or consistent New fields can be added without making changes to the existing schema
Cost-effectiveness	NoSQL databases are typically designed to scale out horizontally by using distributed clusters of hardware,  as opposed to scaling up by adding expensive and robust servers.
Fast queries	Queries in NoSQL databases can be faster than SQL databases.  Data in SQL databases is typically normalised, so queries for a single object or entity requires you to join data from multiple tables.  As the tables grow in size, the joins can become expensive.  However, the data in NoSQL databases is typically stored in a way that is optimised for queries.  Queries typically do not require joins, so the queries are simpler and are very fast. Ease of access to all data about a single apartment Better performance when handling simple queries
Replication	NoSQL replication functionality copies and stores data across multiple servers.  This replication provides data reliability,  ensuring access during downtime and protecting against data loss if servers go offline.

data privacy

the requirement for data to be accessed by, or disclosed to, authorised users only. In other words, it is about keeping data private rather than allowing it to be available in the public domain. The term may be applied to both individuals and organisations.

data integrity

refers to the accuracy and validity of data. Data integrity covers data in storage, during processing, and while in transit. Data integrity can be compromised in several ways and at different stages during data processing.

threats to data

1. Lapses in users’ behaviour — Users may not be careful when dealing with data. This includes making errors in data entry, using weak passwords, or revealing password/OTP to others.

2. Mismanagement by multiple users — Multiple users working on the same file may accidentally overwrite each others’ data, causing data to be inaccurate or invalid.

3. Natural disasters — Natural disasters can destroy the computer system and/or storage device physically.

4. Unauthorised intrusion into the system — If the system is vulnerable to hackers, data may be corrupted or lost. Other cyberattacks such as virus, worms and trojan are equally damaging to the system and its data.

5. Malicious software — Malicious software entering the computer system will cause harm to a system and the data stored in the system.

data backup

• Data backup involves creating copies of data and storing them separately. 

• It is intended to be used as a safety precaution or prevention against unexpected loss of data due to unintended data corruption or errors such as disk drive failing, files accidentally being deleted, or a data center going offline during a catastrophic event.

data archive

• Archiving data involves storing data that is not actively used but kept for historical references or auditing purposes. It ensures that important records remain available years after they were created. 

• In most cases, the purpose of data archiving is to meet legal and compliance requirements. 

• For example, a doctor’s office might be required to keep patient records for a certain period of time. Similarly, a bank may need to retain transaction records.

version control

• Version control is a system that records changes to a file or set of files over time so that you can recall specific versions later.

• E.g. Git

file naming convention

• File naming convention is a framework for naming files in a way that describes what they contain and how they relate to other files. 

• This helps to minimise the chances of files being misplaced or lost unintentionally due to poor organisation of files. 

• Such a convention also enables users to locate files quickly.

• Developing an FNC is done by identifying the key elements of the project, and the important differences and commonalities between your files.

disaster recovery

• Disaster recovery is a set of practices and technologies that determine how an organisation deals with a disaster, such as a cyberattack, natural disaster, or large-scale equipment failure.

• The disaster recovery process typically involves setting up a remote disaster recovery site with copies of protected systems and switching operations to those systems in case of disaster.

data encryption

• Data encryption alters data content according to an algorithm that can only be reversed with the right encryption key.

• Encryption protects your data from unauthorised access even if data is stolen by making it unreadable.

data erasure

• Data erasure is more secure than standard data wiping because it uses software to completely overwrite data on any storage device.

• It verifies that the data is unrecoverable, limiting liability by deleting data that is no longer needed.

• This can be done after data is processed and analysed or periodically when data is no longer relevant. 

• Erasing unnecessary data is a requirement of many compliance regulations

backup vs archive

	Backup	Archive
Data Storage	The original data remains in place, while a backup copy is stored in another location.	Archived data is moved from its original location to an archive storage location.
Data State	Backed-up data is constantly changing.	Once you create an archive, you do not modify it.
Data Retention Policy	You periodically delete or overwrite data backups that are too old to be useful.	Data archives are designed for long-term storage.
Storage Type	Hot cloud storage or easily accessible local storage locations	Cold cloud storage or tape archives.
Data Scope	All of your data, with the exception of unimportant information like temporary files.	Specific files that you must retain for compliance purposes.

obligations

1. Accountability Obligation

• Organisations must take responsibility for protecting personal data

• Make information about your data protection policies, practices and complaints process available upon request

• Designate a data protection officer (DPO) with his business contact information available to the public

2. Notification Obligation

• Notify individuals of the purposes for which your organisation is intending to collect, use, or disclose their personal data.

3. Consent Obligation

• Only collect, use or disclose personal data for purposes which an individual has given their consent to.

• Allow individuals to withdraw consent with reasonable notice, and inform them of their likely consequences of withdrawal.

4. Purpose Limitation Obligation

• Only collect, use or disclose personal data for the purposes that a reasonable person would consider appropriate under the given circumstances and for which the individual has given consent.

• Cannot require individuals to consent to the collection, use or disclosure of their personal data beyond what is reasonable to provide that good or service.

5. Accuracy Obligation

• Make reasonable effort to ensure that personal data collected is accurate and complete, especially if it will affect individuals

6. Protection Obligation

• Make reasonable security arrangement to protect personal data

7. Retention Limitation Obligation

• Cease retention of personal data or dispose of it in a proper manner when it is no longer needed.

8. Transfer Limitation Obligation

• Transfer personal data to another country only according to the requirements prescribed under the regulations, to ensure that the standard of protection is comparable to the protection under the PDPA , unless exempted by the PDPC (Personal Data Protection Commission)

9. Access and Correction Obligation

• Upon request, provide individuals with access to their personal data as well as information about how the data was used or disclosed within a year before request.

• Correct any error or omission in an individual’s personal data as soon as practicable and send corrected data to other organisations to which the personal data was disclosed to, within a year before the correction is made.

10. Data Breach Notification Obligation

• In the event of a data breach, take steps to assess if it is notifiable. If the data breach likely results in significant harm to individuals, and/or are of significant scale, organisations are required to notify the PDPC and the affected individuals as soon as practicable.

11. Data Portability Obligation

• At the request of the individual, organisations are required to transmit the individual’s data that is in the organisation's possession or under its control, to another organisation in a commonly used machine-readable format.

ethical principles

PRIC

→    Integrity

• Act with complete honesty and transparency in all professional dealings.

• Do not misrepresent capabilities, products or findings.

• Disclose known risks, vulnerabilities or limitations to relevant parties, even when inconvenient.

• Example breach: A data security company discovers a vulnerability in a client's database but hides it to avoid panic and protect their image — this is a breach of integrity.

→    Responsibility

• Accept accountability for one's work and its consequences.

• Adhere to professional and client standards; do not defer problems or cover mistakes.

• Ensure data is entered accurately and reported promptly.

• Example breach: A developer knowingly ships buggy code and blames tools, avoiding accountability for the defect.

 

→    Competence

• Only undertake work within one's area of expertise.

• Continuously update knowledge and skills; seek guidance when outside one's competence.

• Do not claim proficiency in skills or technologies you have not demonstrated.

• Example breach: A programmer claims expertise in a programming language they have never used in order to win a contract.

 

→    Professionalism

• Act in the best interests of clients and the public, not solely for profit.

• Maintain confidentiality of sensitive information; protect data from unauthorised access.

• Do not exploit privileged access to systems or data.

• Example breach: A staff member at a vaccination centre leaks patients' personal health data to a third party.

ransomware

Blocks access to a victim's computer system until a sum of money, often in cryptocurrency, is paid

spyware

Secretly collects personal information, such as tracking websites visited or recording keystrokes (keyloggers) to steal passwords and credit card numbers

scareware

Attempts to frighten the victim with loud alarms or flashing images into buying fake antivirus software or handing over financial data

adware

Pushes unwanted advertisements to users while secretly collecting their information

fileless malware

Operates entirely within the computer's memory without downloading code, hiding in trusted applications to evade traditional virus scanners

trojan horse

Pretends to be a harmless or useful application but gives intruders unauthorized access to the computer when run

virus

Attaches itself to a normal program and modifies it, subsequently infecting other programs by attaching copies of itself when executed

worm

A standalone program that automatically spreads copies of itself over a network by exploiting system vulnerabilities or masquerading as an email attachment, which can consume bandwidth and overload servers

phishing

Uses fraudulent emails and fake websites that mimic reputable companies to trick users into revealing sensitive information

pharming

Intercepts a computer's request for a legitimate website and redirects the user to a fake website to steal their data

spamming

The mass distribution of unwanted messages designed to lure users into entering their personal information

cookie misuse

Attackers exploit the small data files websites leave on your browser to secretly collect your personal information

Distributed Denial of Service (DDoS)

A larger-scale attack that uses a "botnet"—a network of multiple compromised systems, such as malware-infected computers or Internet of Things (IoT) devices—to overwhelm the target's infrastructure with internet traffic

Denial of Service (DoS)

An attacker uses a single computer to flood a targeted website or network with requests, overloading it so that its performance degrades or it becomes completely inaccessible to regular users

ethical issues

SEEL

Category	Key concepts and exam triggers
Social	Digital divide, cyberbullying, social media mental health, facial recognition surveillance, job displacement communities
Ethical	Code of ethics (integrity/responsibility/competence/professionalism), AI bias, consent, accountability, transparency
Legal	PDPA (consent, purpose limitation, protection, access/correction), Computer Misuse Act (unauthorised access), copyright infringement
Economic	E-commerce growth, fintech inclusion, automation and job loss, monopolisation by big tech, digital divide → income gap

local area network (LAN)

a network of computing devices connected within a small geographical area, typically within the same building, such as a home, school, office or building

metropolitan area network (MAN)

a network of computing devices covering a larger geographical area (two or more buildings within the same town or city) than a LAN. A MAN is typically owned and operated by a large organisation such as a business, city or government body.

wide area network (WAN)

a network of computing devices covering a large-scale geographical area, typically across multiple geographical locations.

Consists of multiple smaller networks such as LANs or MANs.

intranet

a private network built within an organisation, like a company, school or government agency

internet

a global, public network accessible to anyone with an internet connection. It is a network of networks linked by a broad array of electronic, wireless, and optical networking technologies.

Media Access Control (MAC) address

a unique identifier assigned to a network interface controller (NIC) in a device.

Internet Protocol (IP) address

a unique numerical label assigned to devices connected to a network that uses the Internet Protocol for communication.

The Internet Protocol is a set of rules for data transmission which are agreed by sender and receiver.

• To identify and locate devices on a network.

• Can be assigned dynamically or statically.

Domain Name Service (DNS)

a hierarchical distributed database that maps human-readable domain names to IP addresses. Eliminates the need to memorise IP addresses.

System Type: Decentralised

• Installed on numerous domain name servers across the entire Internet

• Multiple servers handle requests and share information (not one central server)

• Root name servers at the core create a hierarchical structure supporting the whole Internet

• Although centralised at the top (root servers), the overall system is decentralised in distribution

DNS Resolution Process

What happens after you type a URL in the browser address bar:

1. Client checks local cache

• Your computer checks if it has previously looked up this domain name recently (local cache)

• If found → resolution ends here

2. Query sent to Recursive DNS Server

• Recursive server also checks its local cache of recently looked up domain names

• If found → results returned to client, resolutions ends (common for popular services like Google, Facebook)

3. Recursive server queries Root Name Server

• If not cached, Recursive server queries a Root Name Server

• Root server recognises the Top-Level Domain (e.g. .com, .sg) and redirects to correct TLD server

4. TLD Server returns Authoritative Nameserver Address

• TLD server holds records indicating where to find the authoritative DNS server

• Authoritative nameserver address returned to Recursive server

5. Authoritative server returns DNS record → Caching → Return result

• Authoritative nameserver returns the DNS record (depends on record type – see below)

• Recursive DNS Server caches the result locally for future requests (caching duration determined for TTL)

• Result relayed back to original client

Browser Action (after resolution of DNS)

Once domain name is resolved to IP address:

1. Browser connects to the remote server at the resolved IP address

2. Browser connects to the specific port number (e.g. port 8088)

3. Browser sends an HTTP request for the specified resource/path (e.g. blog/page-name)

4. Server responds with the requested resource

TCP/IP Model layers

Layer	Description	Protocols	Hardware/Software
Physical	Handles physical transmission of data packets over the network medium cables, wires, or wireless signals.	Ethernet	Network cables, Wireless adapters, NICs, Modems
Data Link	Packages the data into frames and ensures error-free transmission between devices on the same network segment.	Ethernet, Wi-Fi, PPP (Point-to-point protocol), ARP (Address Resolution Protocol)	NICs, Switches
Internet	Routes data packets across different networks, determining the best path to reach the destination device.	IP (Internet Protocol), ICMP (Internet Control Message Protocol)	Routers
Transport	Manages reliable data transfer between applications.	TCP (Transmission Control Protocol), UDP (User Datagram Protocol)	Not directly associated with specific hardware but relies on the functionality of the Network Access Layer
Application	Provides services directly to applications like web browsing, email, and file transfer.	HTTP, HTTPS, FTP, SMTP, DNS, POP3	Web browsers, Email clients, FTP clients, Operating Systems (for application support)

function of protocols

Protocols are essential for successful transmission of data over a network. Each protocol defines a set of rules that must be agreed between sender and receiver.

circuit switching

a communication method where a dedicated communication path, or circuit, is established between two devices before data transmission begins.

• No other devices can use the circuit for the duration of the session

• Commonly used: voice communication, some types of data communication

• Each data packet follows the same route to its destination

• The returning data follows the same route back to the source

packet switching

a communication method where data is divided into smaller units called packets and transmitted over the network.

• Packet = source & destination addresses + other info for routing

• May take different paths and transmitted out of order or delayed due to network congestion

• Reassembled in correct order to form original data upon reaching destination

packet vs circuit switching

Feature	Packet Switching	Circuit Switching
Data Transfer	Breaks data into packets, sent independently	Dedicated path established between sender and receiver
Routing	Packets can take different routes based on traffic	Dedicated path remains fixed for entire communication
Bandwidth	Dynamically allocated based on traffic	Guaranteed bandwidth for the communication
Efficiency	More efficient for bursty data traffic	Less efficient for bursty data traffic
Cost	Generally considered more cost-effective	Can be more expensive, especially for unused bandwidth
Applications	Ideal for data transfer (web browsing, email)	Ideal for real-time applications (voice calls, video conferencing)

causes of packet loss

Receive packets faster than they are able to route them on → buffering, high latency

• If severe, router may run out of memory and packets are discarded.

• fix: TOS field in header → mark packets with a priority level

→ Request for special treatment

→ routers may ignore these requests

Packets sent to unreachable destination address

• Unaware routers route packets toward a default device → may also pass packet on → Loop forever between routers

• fix: Time to live (TTL) counter in header

→ Set when created, reduced by one every time it goes through a router

→ Counter = zero, packet is discarded

client-server architecture — pros + cons

• Server: Always-on host that services requests from many other hosts (clients)

  • Has a fixed, well-known IP address

• Client: Sends requests to the server

• Client requests data from server, server responds by sending data back to client

Advantages	Disadvantages
Centralised control of data and resources	Higher initial cost due to the need for specialised high-performance servers
Easy to schedule backups of all shared files at regular intervals	Administrative costs needed for the maintenance of servers and clients
Security may be enhanced with the use of specialised software or operating system features that are designed for servers

peer-to-peer

• Direct communication between pairs of intermittently connected hosts (peers)

• Peers – not owned by the service provider, but are desktops of users

Advantages	Disadvantages
Cheaper to set up as there is no cost related to dedicated servers; basic computers can act as servers to share resources	More effort is required to access back up resources as they are stored locally within each computer instead of centrally in a server
Easy to set up as no specialised software or operating system features are needed	Security is low as access rights are handled by individual computers; not administered by a central server
Storage of data is decentralised and can be carried out by individual users at each computer

native vs web

Native Applications	Web Applications
Developed for a specific operating system or platform. Can only be accessed from the machines they are deployed on.	Developed to be accessed via devices’ internet browsers. Can be accessed from anywhere, so there is no location constraint.
Need to be installed in the device to function. Native applications need to be developed separately for different platform machines.	No need to be downloaded or installed. Web applications are platform-independent, they can work on different types of platforms with the only requirement of a web browser.
Have access to system resources such as GPS, camera, etc.	No access to system resources such as GPS, camera, etc.
May be able to work offline. Native applications do not require the internet for their operations. Some applications just require internet connectivity at the time of update.	Need an active internet connection to work. Web applications rely heavily on internet connectivity for their operation.

usability principles in design of web applications

• Visibility of system status

  • E.g. Loading bar

• Match between system and the real world

  • E.g. Trash bin

• User control and freedom

  • E.g. Undo/redo, back/forward buttons

• Consistency and standards

  • Follow convention

• Error prevention

  • E.g. Confirm button

• Recognition rather than recall

• Flexibility and efficiency of use

  • E.g. Customisation options, action shortcuts

• Aesthetic and minimalist design

  • Focus on essentials

• Help users recognise, diagnose and recover from errors

  • E.g. Password requirements

• Help and documentation

GET vs POST

When sending data via GET, the data is transmitted in the URL itself as query parameters, visible in the address bar of the browser.

• (-ve) Limited amount of data that can be sent

• (-ve) Less secure

• (+ve) Able to be bookmarked or shared

When sending data via POST, the data is transmitted in the request body, making it more secure. 

• (-ve) Cannot be bookmarked or shared

• (+ve) More secure

• (+ve) Unlimited amount of data that can be sent

firewalls

• a filter that monitors access between an organisation’s internal network and the internet at large, allowing some packets to pass and blocking others.

• A firewall allows a network administrator to control access between the outside world and resources within the administered network by managing the traffic flow to and from these resources.

goals of firewalls:

1. Traffic Control — all traffic passes through a single choke point (easier to manage and enforce policy)

2. Authorised Traffic Only — blocks unauthorised access as defined by local security policy

3. Maintain Security — the firewall itself is resistant to attack

types of firewalls

• Packet Filter — at gateway router; examines each datagram based on administrator-set rules

• Stateful Packet Filter — tracks TCP connections; uses connection state to make filtering decisions

• Application Gateway — inspects actual application data (beyond IP/TCP/UDP headers); policy decisions based on application content

limitations of firewalls

• Cannot stop attacks from sources the user has explicitly allowed (e.g. user-granted exceptions bypass the firewall)

• Cannot protect against internal attacks — malicious traffic from inside the network may not pass through the firewall

• Single point of failure — if the firewall is compromised, the entire network becomes vulnerable

signature based IDS

• Maintains an extensive database of attack signatures (sets of rules describing known intrusion activity)

• Sniffs every passing packet; compares against database signatures

• Generates alert (email, log, management system) if a match is found

• Limitations:

  • Requires prior knowledge of an attack to generate a signature — completely blind to new/undocumented attacks

  • Can generate false alarms (signature match ≠ guaranteed attack)

  • Processing overhead from comparing every packet against thousands of signatures may cause missed detections

anomaly based IDS

• Builds a baseline traffic profile during normal operation

• Flags packet streams that are statistically unusual (e.g. sudden spike in ICMP packets, exponential growth in port scans)

• Advantage: can detect new, unknown attacks without prior knowledge of them

• Limitation: extremely difficult to distinguish normal from statistically unusual traffic — high false positive rate

IDS vs IPS

• IDS (Intrusion Detection System) — monitors traffic and generates alerts when potentially malicious activity is detected

• IPS (Intrusion Prevention System) — actively filters out suspicious traffic (blocks rather than just alerts)

• Both perform deep packet inspection (examining actual application data, not just headers)

symmetric key encryption — cons

• One key; secret shared by the sender and receiver

• Sender: Encryption algorithm + key to encrypt plaintext

• Receiver: Decrypts ciphertext using the same key

• E.g. Caesar cipher, monoalphabetic cipher, polyalphabetic encryption

• (-ve) Difficult to securely deliver the key to the receiver; can be sent via a different channel or agreed beforehand

public key encryption (asymmetric) — pros + cons

• Pairs of keys to encrypt and decrypt data; public key + private key

• Sender: Use recipient’s public key to encrypt the data

• Recipient: Uses their private key to decrypt the data

• E.g. Rivest-Shamir-Adleman (RSA)

• (+ve) Eliminates the need to exchange secret keys

• (-ve) Not possible to confirm that the message was sent by the stated recipient (without digital signature)

• (-ve) Not possible to confirm that the message has not been tampered with en-route (without digital signature)

using digital signature to check if the data has been tampered with

a cryptographic technique to indicate the owner or creator of a resource or to signify one’s agreement with a document’s content in a digital world.

• Purpose: To prove that a document signed by an individual was indeed signed by that individual and only that individual could have signed the document.

1. Sender uses one-way hash algorithm to create hash digest

2. Sender uses private key to encrypt the hash to the digital signature

3. Message (encrypted or not) + digital signature are sent to the receiver

4. Receiver uses sender’s public key to decrypt the digital signature back to the sender’s version of hash

5. Receiver uses same hash algorithm to create a new hash digest from the received message.

6. If two hash digests match → not altered and sent by known sender

• Application: Public key certification (e.g. IPsec, SSL)