GCE A/L Information & Communication Technology cheat sheet

Data is raw unprocessed facts; information is data processed into a meaningful context; knowledge is information combined with experience and judgement.

Digitisation converts analogue content into discrete binary form; digitalisation is the wider reorganisation of processes and services around digital technology.

Moore's law: the number of transistors on an integrated circuit roughly doubles every two years at similar cost; it describes an observed trend, not a physical law.

Evolution milestones in order: mechanical calculators, valve (first-generation) computers, transistors, integrated circuits, microprocessors, personal computers, networks, mobile and cloud computing.

The digital divide is the gap in access to and effective use of ICT between regions, income groups, ages and genders; access alone does not close it, skills and affordability also matter.

ICT impact categories for evaluation: employment (creation, displacement, gig work), education (e-learning, open resources), health (telemedicine), governance (e-government), and environment (e-waste, energy use).

Automation substitutes routine tasks; augmentation supports judgement-intensive tasks. Net employment effect depends on the balance between displacement and new roles.

Green ICT has two directions: greening of ICT (energy-efficient devices, data-centre cooling) and greening by ICT (dematerialisation, smart grids, telecommuting).

Technology adoption models (TAM, diffusion of innovations) explain uptake through perceived usefulness, ease of use, relative advantage, compatibility and observability.

Von Neumann architecture: CPU (ALU, control unit, registers), memory holding both instructions and data, input, output, and a shared bus; the fetch-decode-execute cycle repeats.

CPU performance depends on clock speed, cores, cache size and instruction-level parallelism; cache levels L1 < L2 < L3 in size and speed order.

Memory hierarchy by speed and cost: registers, cache, RAM, SSD/HDD, optical/tape; volatile memory loses contents on power loss, non-volatile retains it.

Number of addressable locations = 2^(address bus width); word length equals the data bus width and determines the largest single-word integer.

Character codes: ASCII uses 7 bits (128 characters); extended ASCII uses 8 bits; Unicode uses variable encoding (UTF-8, UTF-16) to cover all scripts.

Signed integers: sign-and-magnitude and one's complement have two representations of zero; two's complement has one and is used in hardware. Range for n bits: -2^(n-1) to 2^(n-1)-1.

Image size in bits = width x height x colour depth; sound file size = sample rate x bit depth x channels x duration; video adds frame rate.

Compression: lossless (RLE, Huffman, ZIP) preserves exact data; lossy (JPEG, MP3, MPEG) discards perceptually less important data and cannot be reversed exactly.

System software (OS, utilities, drivers) manages hardware and provides services; application software serves end-user tasks; an OS handles process, memory, file, device and security management.

Place values: binary base 2, octal base 8, decimal base 10, hexadecimal base 16. Hex digits 0-9 then A-F; one hex digit equals exactly four binary bits.

Binary to decimal: sum of bit x 2^position from the right, starting at position 0. Decimal to binary: repeated division by 2, reading remainders upwards.

Octal-hex conversion is fastest via binary: group bits in threes for octal and fours for hexadecimal.

Two's complement of an n-bit number: invert all bits and add 1; the leading bit has weight -2^(n-1).

BCD encodes each decimal digit in four bits; it is not the same as binary and is used in displays and financial arithmetic.

Gate truth behaviour: AND outputs 1 only if all inputs are 1; OR outputs 1 if any input is 1; NOT inverts; NAND and NOR are universal gates.

XOR outputs 1 when inputs differ; XNOR outputs 1 when inputs are equal. XOR with a 1 acts as a bit inverter; XOR with 0 leaves a bit unchanged.

Boolean identities: A+A=A, A.A=A, A+A'=1, A.A'=0, A+AB=A (absorption), A+A'B=A+B, and De Morgan: (A.B)'=A'+B', (A+B)'=A'.B'.

Karnaugh maps group adjacent 1s in powers of two (1, 2, 4, 8) including wrap-around edges; each group removes one variable per doubling.

Transmission modes: simplex (one direction only), half-duplex (both directions, not simultaneously), full-duplex (both simultaneously).

Multiplexing shares a link: FDM divides frequency bands, TDM divides time slots, and statistical multiplexing allocates on demand.

OSI seven layers top-down: Application, Presentation, Session, Transport, Network, Data Link, Physical. TCP/IP model: Application, Transport, Internet, Network Access.

Devices by layer: hub (physical, broadcasts to all ports), switch (data link, MAC-based forwarding), router (network, IP-based), gateway (protocol translation).

Switching: circuit switching reserves a path for the whole session; packet switching routes independent packets and is more resilient and efficient for bursty data.

IP addressing: IPv4 is 32-bit dotted decimal; IPv6 is 128-bit hexadecimal. Private ranges: 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16.

Subnet mask separates network and host bits; usable hosts = 2^host bits - 2 (network and broadcast addresses excluded).

Topologies: star (central failure risk, easy fault isolation), bus (single backbone, collision-prone), ring (token passing), mesh (redundant, costly).

Error detection: parity detects odd numbers of bit errors; checksum sums data; CRC uses polynomial division and is stronger. ARQ (stop-and-wait, sliding window) adds retransmission.

The Internet is a global network of networks using TCP/IP; the World Wide Web is one application on it, accessed via HTTP/HTTPS and identified by URLs.

URL structure: scheme://host[:port]/path?query#fragment. DNS resolves domain names to IP addresses through a hierarchical distributed database.

HTTP is stateless; cookies and sessions maintain state. HTTPS adds TLS encryption and server authentication via digital certificates.

HTTP methods: GET retrieves (idempotent, data in URL), POST submits (data in body, not idempotent), PUT replaces, DELETE removes.

Client-side scripting (JavaScript) runs in the browser for interactivity and validation; server-side scripting (PHP, Python, Java) runs on the server and can access databases.

Markup and styling: HTML defines structure and semantics, CSS controls presentation and responsive layout, JavaScript controls behaviour, XML carries self-describing structured data.

E-services: e-commerce (B2B, B2C, C2C, C2B), e-banking, e-learning, e-government, e-health. Payment options include cards, digital wallets and mobile money.

Search engines crawl, index and rank pages; ranking uses relevance signals and link analysis. SEO improves visibility through structure, keywords and page speed.

Web accessibility guidelines (WCAG principles: perceivable, operable, understandable, robust) support users with disabilities through alt text, contrast and keyboard navigation.

DBMS functions: data definition, data manipulation, concurrency control, backup and recovery, integrity enforcement, security and access control.

Relational model: relations (tables), tuples (rows), attributes (columns); primary key uniquely identifies a row, foreign key references another relation's primary key.

Normalisation: 1NF removes repeating groups (atomic values); 2NF removes partial dependencies on part of a composite key; 3NF removes transitive dependencies on non-key attributes.

SQL categories: DDL (CREATE, ALTER, DROP), DML (SELECT, INSERT, UPDATE, DELETE), DCL (GRANT, REVOKE), TCL (COMMIT, ROLLBACK).

SELECT order of clauses: SELECT ... FROM ... WHERE ... GROUP BY ... HAVING ... ORDER BY. WHERE filters rows before grouping; HAVING filters groups after aggregation.

Cardinality: one-to-one, one-to-many, many-to-many. A many-to-many relationship is resolved into two one-to-many links via a junction (link) table.

ACID properties: Atomicity (all or nothing), Consistency (valid state to valid state), Isolation (concurrent transactions do not interfere), Durability (committed changes survive failure).

Data warehouse stores integrated historical data for analysis (OLAP); operational database supports day-to-day transactions (OLTP).

Information system components: people, procedures, data, hardware, software and networks; a transaction processing system records routine events, a management information system summarises them.

Algorithm properties: finiteness, definiteness, input, output, effectiveness. Represent with pseudocode or flowcharts; symbols: oval terminal, parallelogram input/output, rectangle process, diamond decision.

Complexity classes in growth order: O(1) < O(log n) < O(n) < O(n log n) < O(n^2) < O(2^n). Big-O describes the upper bound as input grows.

Searching: linear search is O(n) and works on unsorted data; binary search is O(log n) but requires sorted data.

Sorting: bubble, insertion and selection are O(n^2); merge sort and quicksort average O(n log n); merge sort needs extra space, quicksort worst case is O(n^2).

Data structures: array (fixed size, index access O(1)), linked list (dynamic, sequential access), stack (LIFO: push/pop), queue (FIFO: enqueue/dequeue), tree, graph, hash table.

Traversal: pre-order (root, left, right), in-order (left, root, right) gives sorted output for a binary search tree, post-order (left, right, root).

Programming constructs: sequence, selection (if/case), iteration (for, while, do-while). A while loop tests before the body; do-while executes at least once.

Parameter passing: by value copies the argument; by reference passes the address so the procedure can modify the caller's variable.

SDLC models: waterfall (sequential, rigid), iterative/incremental, spiral (risk-driven), agile (short sprints, continuous feedback). Testing levels: unit, integration, system, acceptance.

SDLC phases: requirements analysis, feasibility study, design, development, testing, implementation, maintenance. Maintenance typically consumes the largest share of lifetime cost.

Feasibility types: technical, economic (cost-benefit), legal, operational, schedule. A feasibility study precedes full development commitment.

Requirement-gathering techniques: interviews, questionnaires, observation, document review, prototyping, joint application design (JAD).

Data flow diagram symbols: external entity (rectangle), process (rounded rectangle or circle), data store (open rectangle), data flow (arrow). Context diagram shows the system as one process.

Modelling tools: entity relationship diagram for data, use case diagram for actor-system interaction, class diagram for structure, flowchart for process logic, Gantt chart for schedule.

Implementation (changeover) strategies: direct (cheap, risky), parallel (safe, costly), phased (by module), pilot (one site first).

Project management: work breakdown structure, critical path method (longest dependent path sets minimum duration), PERT for uncertainty, Gantt for timeline view.

Testing types: black-box (functionality without internals), white-box (code paths), alpha (in-house), beta (real users), regression (after change).

Documentation: user documentation (guides, help) and technical documentation (design, code, test records); both are needed for maintainability.

CIA triad: Confidentiality (only authorised access), Integrity (data unaltered and accurate), Availability (accessible when needed). Authentication, authorisation and non-repudiation support them.

Threats: malware (virus, worm, Trojan, ransomware, spyware), phishing and social engineering, denial of service, man-in-the-middle, SQL injection, brute force.

Controls: encryption (symmetric for speed, asymmetric for key exchange and signatures), hashing for integrity, digital signatures for authenticity, firewalls, IDS, access control, backups, security policies.

Encryption versus hashing: encryption is reversible with a key; a cryptographic hash is one-way and used for password storage with salt.

Computer Misuse Act-type offences: unauthorised access, unauthorised access with intent to commit further offences, unauthorised modification. Intellectual property is protected by copyright, patents and trademarks.

Data protection principles: lawful and fair processing, purpose limitation, data minimisation, accuracy, storage limitation, integrity and confidentiality, accountability; data subjects have access and correction rights.

Ethical frameworks for ICT decisions: professional codes of conduct, stakeholder analysis, informed consent, privacy by design, avoidance of harm.

Cyberbullying, misinformation, filter bubbles and digital addiction are social issues; e-waste and energy consumption are environmental issues.

Backup strategy: 3-2-1 rule (three copies, two media, one offsite); disaster recovery plan defines recovery point objective and recovery time objective.

Cloud service models: IaaS (virtual machines, storage), PaaS (runtime and development platform), SaaS (ready applications). Deployment: public, private, community, hybrid.

Cloud characteristics (NIST): on-demand self-service, broad network access, resource pooling, rapid elasticity, measured service.

Virtualisation abstracts physical resources so multiple virtual machines share one host; containers package an application with its dependencies and share the host OS kernel.

Mobile computing constraints: limited battery and processing, variable connectivity, small screens, security of personal data; mobile OS permission models control app access.

IoT architecture layers: perception (sensors/actuators), network (connectivity), processing/cloud (analytics), application (user services). Protocols include MQTT, CoAP, Zigbee, LoRaWAN.

Big data characteristics (the Vs): volume, velocity, variety, veracity, value. Processing tools: Hadoop MapReduce for batch, Spark for in-memory processing, NoSQL for schema-flexible stores.

AI subfields: machine learning (supervised, unsupervised, reinforcement), natural language processing, computer vision, robotics, expert systems. Deep learning uses multi-layer neural networks.

Supervised learning trains on labelled data for classification or regression; unsupervised learning finds structure such as clustering in unlabelled data.

Emerging trends: edge and fog computing, 5G, blockchain as a distributed immutable ledger, digital twins, AR/VR, and generative AI with associated bias, privacy and accountability concerns.