Chapters 7, 8 & 9 — Networks, Databases & Modelling — explained with simple definitions, analogies, memory tricks, and past exam tips.
How computers talk to each other — the internet, protocols, TCP/IP, topologies, and cybersecurity.
7The Internet is the global network of connected computers and infrastructure that carries all data (web pages, emails, files, calls). It is hardware.
The World Wide Web (WWW) is a service that runs on top of the internet — it is the collection of web pages you view in a browser. The WWW is software / a service.
The internet is the road network — all the physical roads, motorways, and infrastructure. The WWW is like delivery trucks using those roads. Email, FTP, and video calls are other services using the same road network — not the road itself.
💡 Internet = Infrastructure (hardware). WWW = Web pages (service on that hardware). The web is a subset of what the internet can do. Not all internet use is web browsing!
A common exam trap: students say "internet" when they mean "web". Be precise — if you're talking about web pages, say WWW. If you're talking about the underlying connection, say internet.
Covers a small area — one building, school, or office. Fast, private, cheap to run.
Example: School computer network
Covers large distances — cities, countries, the world. The internet is the biggest WAN.
Example: The internet itself
Very short range around one person — Bluetooth, USB connections.
Example: Phone + earphones via Bluetooth
Local = Little area | Wide = World | Personal = Pocket
🏢 Client-Server
A dedicated server stores all data and manages resources. Client computers send requests; server responds. Users must log in.
✅ Advantages: More secure, central backup, easy to manage, centralised control of access rights.
❌ Disadvantages: Expensive to set up and maintain. If server fails, ALL computers lose access.
👥 Peer-to-Peer (P2P)
All computers are equal peers — each can act as both client and server. Data is shared directly between devices.
✅ Advantages: Cheap, easy to set up. No single point of failure.
❌ Disadvantages: Less secure, hard to manage, data spread across many computers.
Client-Server is like a bank — all your money (data) is stored securely in one place (the server), and you visit (log in) to access it. P2P is like friends lending each other cash directly — quick and simple but risky if someone is untrustworthy.
Examcraft 2024 asked: "With respect to data security, why is client-server a good choice?" → Answer: Data stored centrally = easier to apply security controls, encryption, access controls, backups, and audit logs from one location.
A topology is the physical or logical arrangement of computers in a network.
Formula: n × (n − 1) ÷ 2
DEB 2024 asked to derive this formula and calculate connections for 5 computers (answer = 10). Mesh advantages: highly resilient, no single point of failure, multiple routing paths. Disadvantage: expensive, lots of cabling.
A protocol is a set of standardised rules that computers follow to communicate — governing things like message format, data size, speed, error checking, and timing. Without protocols, computers from different manufacturers couldn't talk to each other.
A protocol is like a shared language combined with manners. When two diplomats meet, they follow a protocol — shake hands, introduce themselves in a set order, speak in an agreed language. Without this, there'd be chaos. TCP/IP is the internet's diplomatic protocol!
HTTP = Homepages | FTP = Files | SMTP = Sending email | VoIP = Voice | TCP = Transport (packets) | IP = Identity (address)
DEB papers frequently ask you to fill in a table of protocol names and functions. Also: "Distinguish between IP and VoIP" — IP routes data packets, VoIP delivers real-time voice using that routing system (2025 paper).
When data travels across the internet, it passes through 4 layers. Think of it like sending a package — each layer adds its own envelope and label.
Imagine posting a letter. Application Layer = you writing the letter. Transport Layer = putting it in an envelope, numbering pages. Internet Layer = writing the address on the envelope. Data Link Layer = the post van physically driving it there.
TCP = "Trustworthy Courier Parcels" — it tracks every parcel (packet), makes sure they all arrive, and asks for replacements if any are lost! Unlike UDP which just fires and forgets.
An IP address is a unique numerical label (e.g. 192.168.1.1) assigned to every device on a network. It identifies WHERE on the internet a device is, so data knows where to go.
A MAC address (Media Access Control) is a permanent, unique hardware address burned into every network card at manufacture. Unlike IP addresses, MAC addresses don't change. Format: 6 pairs of hex digits, e.g. A4:B1:3C:22:F0:9D.
DNS (Domain Name System) is the internet's phone book — it translates human-readable domain names (e.g. www.google.com) into IP addresses (e.g. 142.250.74.46).
IP vs MAC — Key Difference
IP: Changes with location (dynamic). Identifies position on network.
MAC: Never changes. Identifies the physical device.
IP in Cybercrime EXAM
Servers log IP addresses for every connection. Law enforcement can subpoena an ISP to find which customer had a specific IP at a specific time, tracing a suspect's identity and location.
IP address = your home address (can change if you move). MAC address = your fingerprint (always the same, part of you forever). DNS = a sat-nav that translates "McDonald's on O'Connell Street" into GPS coordinates.
DEB 2023 asked: "Why do devices need MAC addresses to connect to the internet?" — MAC addresses identify devices at the hardware level within a local network. Even if IP addresses change, the MAC uniquely identifies the hardware. Also: DEB 2023 asked about DNS as the internet's phone book.
Ransomware is a type of malware that encrypts the victim's files and demands payment (a ransom) to restore access. Famous example: 2021 HSE attack in Ireland.
| Security Measure | What It Does |
|---|---|
| 🔐 Encryption | Data converted to unreadable code — useless without the decryption key |
| 🔑 Strong Passwords + 2FA | Multi-factor authentication requires two forms of proof of identity |
| 🛡️ Firewall | Monitors and blocks unauthorised network access based on rules |
| 💾 Regular Backups | Offline backups mean you can recover files even after ransomware |
| 🔒 VPN | Encrypted tunnel for remote workers — hides data from eavesdroppers |
| 🧑💻 Staff Training | Most breaches start with human error (phishing emails) — training prevents this |
| 🔄 Software Updates | Patches fix security vulnerabilities that hackers exploit |
EFF-BVS = Encryption, Firewall, Frequent backups, Biometrics/2FA, VPN, Staff training
2022 paper asked about the HSE ransomware attack. 2023 DEB asked for 3 IT security measures for remote working. Data breach impacts on individuals: identity theft, financial loss, psychological harm, medical records exposed.
Organising data with tables, keys, relational design, SQL, data cleaning, and analytics pipelines.
8A database is an organised collection of structured data, managed by a DBMS (Database Management System). It allows data to be stored, searched, updated, and retrieved efficiently.
A table organises data in rows and columns. A record (row) is a single entry — one person, one book, one transaction. A field (column) is a single attribute of every record — like "Name" or "Date".
Think of a database like a spreadsheet. The whole spreadsheet file = database. Each sheet tab = a table. Each row = a record (one student). Each column = a field (their name, age, etc.).
| book_id PK | title | year | cost | author | on_loan |
|---|---|---|---|---|---|
| 1 | The Cat in the Hat | 1957 | 6.95 | Dr. Seuss | Yes |
| 2 | Green Eggs and Ham | 1960 | 8.95 | Dr. Seuss | No |
| 3 | Horton Hears a Who! | 1954 | 10.00 | Dr. Seuss | No |
| 4 | How the Grinch Stole Christmas! | 1957 | 9.95 | Dr. Seuss | Yes |
2024 paper used this exact table. You were asked to: (i) define "database" and "record" (ii) choose the primary key (iii) identify 2 data inconsistencies. Classic exam structure — be ready for it!
A field that uniquely identifies every record in a table. It must be: unique (no two records share the same value) and not null (it can never be empty).
Best choice: a dedicated ID number (e.g. book_id, member_id) rather than a name — names can duplicate!
A field in one table that references the primary key of another table. This links the two tables together — it's how relational databases work.
Example: In LOANS, member_id is a FK referencing MEMBERS.member_id.
A primary key is like your passport number — unique to you. A foreign key is like writing your passport number on a hotel register — it links YOU (MEMBERS table) to your BOOKING (LOANS table) without copying all your details again.
MEMBERS table
| member_id PK | member_name |
|---|---|
| 1 | Amy |
| 2 | Bill |
| 3 | Chloe |
LOANS table
| member_id FK | book_id FK | date_borrowed |
|---|---|---|
| 3 | 2 | 20/05/2024 |
| 1 | 4 | 01/05/2024 |
| 2 | 1 | 18/05/2024 |
2024 paper: you were given these exact tables and had to fill in 6 empty cells by cross-referencing given information. E.g. "Amy borrowed How the Grinch Stole Christmas! on 1st May" → find Amy's member_id (1) and the book's book_id (4), then enter: 1, 4, 01/05/24.
A relational database stores data in multiple linked tables using keys. This avoids storing the same data in multiple places.
Data redundancy is when the same data is stored unnecessarily in multiple places. This wastes storage and causes inconsistency — if you update one copy but forget another, your data is wrong.
Imagine writing every student's home address in EVERY entry in the school register. If they move house, you'd have to update it in 100 places! A relational database instead stores the address ONCE in a STUDENTS table, and just references the student's ID in other tables. Far more efficient and consistent.
If we have a separate LOANS table, we can tell whether a book is on loan simply by checking if that book_id appears in the LOANS table. If it's there = on loan. If not = available. So there's no need to store a redundant on_loan Boolean field in BOOKS — the LOANS table already tells us this information. This is exactly what relational design achieves: avoid storing derived or duplicate data.
"Don't Repeat Yourself (DRY)" — the golden rule of relational databases. Every piece of data should exist in only ONE place. Use foreign keys to link, not to copy.
| Inconsistency Type | Example from BOOKS table |
|---|---|
| Mixed date formats | 02/03/1904 vs March 2, 1904 — same date, different formats |
| Inconsistent Boolean | Yes / Y / N / No — should all be the same format |
| Typo / misspelling | Dr. Seuss vs Doc Seus — different values for same person |
| Mixed currency format | €10 vs 9.95 — inconsistent number formats |
| Missing values (nulls) | Empty cells where data should exist |
| Duplicates | Same record entered twice |
| Outliers | Value wildly outside expected range (e.g. cost = €9999) |
Data inconsistencies make queries unreliable and reports inaccurate. For example, if you search for "Dr. Seuss", you won't find books by "Doc Seus" — even though it's the same author!
A data analytics project follows these stages — you need to know all 6:
A hypothesis is an educated, testable prediction about your data. It's a statement you want to prove or disprove. Example (from 2024 exam): "Books published before 1960 are more expensive than books published after 1960." You then collect data and analyse it to test this.
Data cleaning removes errors so your analysis is accurate. Types of errors removed:
Continuous data can take any value within a range — temperature (30.1°C, 30.12°C, etc.), height, weight. There are infinite possible values between any two points.
Discrete data can only take specific, separate values — number of students (you can't have 2.7 students!), shoe size, number of goals scored.
Memory trick: Discrete = Distinct separate values. Continuous = flows Continuously like a river.
2024 exam: suggest 2 hypotheses Alex could test with her library database. Good examples: "More expensive books are borrowed more often" or "Books published before 1960 have higher borrowing rates." Keep hypotheses specific and testable!
Creating simplified virtual versions of real-world systems to test, predict, and explore "what if" scenarios safely.
9A model is a simplified representation of a real-world system that captures only the most important variables and relationships — ignoring irrelevant details (abstraction).
A simulation is the process of running a model over time to observe behaviour, test scenarios, and predict outcomes. Simulation = model in action.
A model is like a toy plane — it looks like a real plane but is much simpler (no real engine, no passengers). A simulation is putting that toy in a wind tunnel to test what happens. Together they let engineers test designs safely before building the real thing.
✅ Benefits of Modelling
❌ Limitations of Modelling
2025 paper asked: "State one limitation and one benefit of modelling." Also asked for 2 areas where modelling improves outcomes (Education, Mental wellbeing, Environment, Traffic management). Know at least 2 areas with a brief description!
Abstraction is the process of removing unnecessary details from a model so it only includes variables that are relevant to the outcome being studied. This makes the model simpler, faster, and more useful.
A London Underground map is a brilliant example of abstraction. It shows stations and connections but ignores actual geography, distance between stations, curves of the tunnels, and what's above ground. It removes all the "noise" and shows only what a passenger needs: which line to take and where to change.
| Model Type | Details Kept | Details Abstracted Away |
|---|---|---|
| Traffic flow model | Speed, direction, road layout | Car colour, driver name, radio station playing |
| Weather model | Temperature, pressure, humidity | Individual leaf positions on trees |
| Disease spread model | Infection rate, proximity, immunity | Patient's favourite food, eye colour |
| Driving test model (2025 paper) | Age, transmission type, instructor experience | Student's nationality, hair colour |
Abstraction = "What actually matters?" Only include variables that could actually influence the outcome you're modelling. Everything else = irrelevant detail = remove it.
An output variable is the predicted result or outcome — the thing you're trying to predict. E.g. "Test outcome (PASS/FAIL)"
Input variables are the features that influence the output. E.g. Age, Transmission type, Instructor experience.
| ID | Gender | Age | Transmission | Instructor Exp (yrs) | Test Outcome |
|---|---|---|---|---|---|
| 1 | Male | 18 | Manual | 8 | FAIL |
| 2 | Female | 22 | Automatic | 5 | PASS |
| 3 | Male | 20 | Automatic | 8 | PASS |
| 4 | Male | 19 | Manual | 3 | FAIL |
| 5 | Female | 30 | Automatic | 12 | PASS |
| 6 | Female | 25 | Manual | 2 | FAIL |
| 7 | Female | 23 | Automatic | 10 | PASS |
| 8 | Male | 17 | Manual | 6 | FAIL |
| 9 | Male | 19 | Manual | 5 | FAIL |
| 10 | Female | 21 | Manual | 15 | PASS |
Output variable: Test outcome (PASS/FAIL)
Input variables: Age, Gender, Transmission type, Instructor experience (years)
Pattern 1: All students who drove Automatic transmission PASSED — none who drove Automatic failed.
Pattern 2: Students with higher instructor experience tended to pass — the one Manual driver who passed (ID 10) had 15 years' instructor experience (the highest).
Additional variables to add: Number of lessons taken, number of previous test attempts, weather on test day, time of day, test centre location.
Agent-based modelling (ABM) is a type of simulation where each individual entity (an agent) is programmed with simple rules. The simulation runs all agents simultaneously and observes how their interactions produce complex system-level outcomes.
Emergent behaviour is complex, system-level patterns that arise from many simple individual interactions — these patterns are not directly programmed, they just emerge naturally from the interactions of agents following their rules.
Think of a murmuration of starlings — thousands of birds swirling in amazing coordinated patterns. Nobody is giving orders. Each bird follows 3 simple rules: don't collide, stay close to neighbours, match their direction. The beautiful swirling patterns emerge from those simple individual rules. That's emergent behaviour!
DEB 2024 asked: "Name a scenario for ABM and explain benefits." DEB 2023 used traffic. 2025 paper used disease spread. Know: (i) define ABM (ii) define emergent behaviour (iii) describe how ABM works for a specific scenario (iv) name one strategy to test and how to measure success.
2025 paper asked you to pick 2 areas and describe how modelling helps. Here are all 4 with descriptions:
E-M-E-T: "Every Model Eliminates Trouble" = Education, Mental wellbeing, Environment, Traffic management
Click a card to flip it. Use the filters to focus on one chapter.
🃏