S5 ICT Practice Paper 4 | Subnetting, TCP/IP, databases, complexity, compression and information security
- 2026-11-22
Senior Secondary ICT
S5 ICT practice paper 4: IPv4 subnetting, TCP/IP layers, normalisation and SQL, algorithm complexity, file sizes and information security, with full answers.
Download Question paper PDF | Answer key PDF
Paper details S5 ICT practice paper 4 with full worked answers on IPv4 subnetting and usable host counts, network and broadcast addresses, the four-layer TCP/IP model, TCP versus UDP, MTU and MSS, database normalisation to 3NF with SQL queries, linear and binary search, sorting complexity and Big-O, bitmap and audio file sizes, run-length encoding, symmetric and asymmetric encryption, phishing and the Personal Data (Privacy) Ordinance.
1. IPv4 subnetting and usable hosts (18 marks)
Usable hosts = 2^n – 2, where n is the number of host bits (network and broadcast addresses are excluded).
| # | Question | Answer and marking notes |
|---|---|---|
| 1 | (a) For the network 192.168.10.0/26, find the number of addresses per subnet, the number of usable hosts, and how many /26 subnets a /24 block yields. | Host bits = 32 – 26 = 6, so each subnet has 2^6 = 64 addresses and 64 – 2 = 62 usable hosts. A /24 block holds 2^(26-24) = 4 /26 subnets. |
| 2 | (b) For the network 10.0.0.0/22, find the subnet mask, the number of addresses and the number of usable hosts. | The mask is 255.255.252.0 (22 ones followed by 10 zeros). Host bits = 10, so there are 2^10 = 1024 addresses and 1024 – 2 = 1022 usable hosts. |
| 3 | (c) For the host address 172.16.5.130/27, find the network address, the broadcast address and the usable host range. | The block size is 2^(32-27) = 32. Since 130 / 32 = 4 remainder 2, the network address is 172.16.5.128 and the broadcast address is 172.16.5.159; usable hosts run from 172.16.5.129 to 172.16.5.158. |
| 4 | (d) A /24 block 192.168.1.0 must be split into at least 5 subnets with at least 25 hosts each. Find the most suitable prefix length and justify it. | Twenty-five hosts need 5 host bits because 2^5 – 2 = 30 is enough, while /28 gives only 2^4 – 2 = 14. The prefix is therefore 32 – 5 = /27 (255.255.255.224), which yields 2^3 = 8 subnets and meets the requirement. |
2. TCP/IP layers and protocols (16 marks)
Name the layer and the protocol in each answer.
| # | Question | Answer and marking notes |
|---|---|---|
| 5 | (a) List the four layers of the TCP/IP model and give one protocol for each. | Application layer (HTTP, DNS, SMTP), transport layer (TCP, UDP), internet layer (IP, ICMP) and network access layer (Ethernet, IEEE 802.11). |
| 6 | (b) Compare TCP and UDP on four points: connection, reliability, overhead and typical use. | TCP is connection-oriented and reliable (acknowledgements, retransmission, in-order delivery) but has a larger header and more overhead, so it suits the web and file transfer. UDP is connectionless, does not guarantee delivery, has a small header and low latency, and suits voice, live streaming and DNS lookups. |
| 7 | (c) A link has an MTU of 1500 bytes, with a 20-byte IP header and a 20-byte TCP header. Find the MSS and the payload as a percentage of the frame. | MSS = 1500 – 20 – 20 = 1460 bytes, so the payload is 1460 / 1500 x 100% = 97.33% of the frame. |
| 8 | (d) Outline the three steps of the TCP three-way handshake and state its purpose. | The client sends SYN with its initial sequence number, the server replies SYN-ACK acknowledging it and sending its own number, and the client returns ACK. This confirms that both sides can send and receive, synchronises the initial sequence numbers and establishes a reliable connection before data flows. |
3. Database normalisation and SQL (18 marks)
Use standard SQL syntax and state the dependency when normalising.
| # | Question | Answer and marking notes |
|---|---|---|
| 9 | (a) State the requirements of first, second and third normal form. | 1NF: every cell holds a single atomic value and there are no repeating groups. 2NF: 1NF plus every non-key column depends on the whole primary key, removing partial dependencies. 3NF: 2NF plus no non-key column depends on another non-key column, removing transitive dependencies. |
| 10 | (b) The table Enrol(student_id, student_name, class_code, class_teacher) has two dependency problems. Identify them and decompose the table into 3NF. | student_name depends only on student_id (a partial dependency if the key is student_id plus class_code), and class_teacher depends only on class_code (a transitive dependency). Decompose into Student(student_id, student_name, class_code), Class(class_code, class_teacher) and Enrol(student_id, class_code). |
| 11 | (c) Write the SQL that lists each class whose average score is above 70, with its average score. | SELECT class_code, AVG(score) AS avg_score FROM Enrol GROUP BY class_code HAVING AVG(score) > 70; With the sample data only A1 qualifies, averaging 80.33. |
| 12 | (d) Write the SQL that lists every student name with the teacher of that student class (a two-table join). | SELECT s.name, c.teacher FROM Student s JOIN Class c ON s.class_code = c.class_code; With the sample data this returns 5 rows. |
4. Algorithm complexity and sorting (16 marks)
Show the expression used for each answer.
| # | Question | Answer and marking notes |
|---|---|---|
| 13 | (a) For a sorted array of 1,048,576 elements, how many comparisons do linear search and binary search need in the worst case? | Linear search needs n = 1048576 comparisons in the worst case; binary search needs ceil(log2 n) = ceil(log2 1048576) = 20. |
| 14 | (b) For 1024 elements, bubble sort makes n(n-1)/2 comparisons in the worst case while merge sort makes about n log2 n. Compute both and compare. | Bubble sort: 1024 x 1023 / 2 = 523776 comparisons. Merge sort: about 1024 x 10 = 10240. That is roughly 51 times fewer, and the gap widens as n grows. |
| 15 | (c) Take the array [42, 17, 68, 9, 55, 23, 71, 34] and run one Lomuto partition with the last element as pivot. Give the resulting array and the pivot index. | The pivot is 34. Sweep left to right, moving every element not greater than 34 to the front, then swap the pivot into the boundary: the array becomes [17, 9, 23, 34, 55, 68, 71, 42] with the pivot at index 3, everything to its left at most 34 and everything to its right greater than 34. |
| 16 | (d) Order O(n^2), O(n log n), O(log n) and O(1) from slowest to fastest growth and give one example of each. | O(1) (array access by index) < O(log n) (binary search) < O(n log n) (merge sort) < O(n^2) (bubble sort). As n grows the highest-order term dominates the running time. |
5. Digital file sizes and compression (16 marks)
1 MiB = 1,048,576 bytes and 1 MB = 1,000,000 bytes.
| # | Question | Answer and marking notes |
|---|---|---|
| 17 | (a) An uncompressed 1920 x 1080 bitmap at 24-bit colour depth: find its size in MiB and in MB. | Each pixel uses 24 bits = 3 bytes, so the total is 1920 x 1080 x 3 = 6220800 bytes = 5.93 MiB, about 6.22 MB. |
| 18 | (b) Find the size of an uncompressed 3-minute WAV file at 44.1 kHz, 16-bit, stereo. | Each second needs 44100 x 2 bytes x 2 channels = 176400 bytes; 3 minutes is 180 s, so the file is 31752000 bytes = 30.28 MiB, about 31.75 MB. |
| 19 | (c) A row of indexed-colour pixels reads WWWWBBBBWWWWWWBBB at 1 byte per pixel. Encode it with run-length encoding using 2 bytes per (colour, count) pair, and find the compressed size and the compression ratio. | The runs are (W,4) (B,4) (W,6) (B,3), so 4 pairs x 2 bytes = 8 bytes against 17 bytes originally; the ratio is 17 / 8 = 2.125. RLE works best when colours come in long runs. |
| 20 | (d) Compare lossy and lossless compression, giving one example of each and when it is used. | Lossless compression restores the file exactly (PNG, ZIP, FLAC) and suits text, source code and medical images. Lossy compression discards some data for a much higher ratio (JPEG, MP3, H.264) and suits photos, music and video, though repeated re-encoding accumulates visible or audible loss. |
6. Information security and legislation (16 marks)
Name the technique and state what it protects.
| # | Question | Answer and marking notes |
|---|---|---|
| 21 | (a) Compare symmetric and asymmetric encryption, giving one algorithm and one use for each. | Symmetric encryption uses one shared key for both directions (AES): fast, so it protects bulk data such as disks and HTTPS session traffic. Asymmetric encryption uses a public key to encrypt and a private key to decrypt (RSA): used for key exchange and digital signatures, but computationally slower. |
| 22 | (b) Explain how a digital signature provides both integrity and non-repudiation. | The sender hashes the message (SHA-256) and signs the hash with a private key. The receiver verifies the signature with the sender public key and recomputes the hash; a match proves the content was not altered (integrity), and only the holder of the private key could have produced the signature (non-repudiation). |
| 23 | (c) Name three common network attacks and one defence for each. | Phishing email (check the sender domain and enable multi-factor authentication); password brute force (long passphrases plus account lockout and MFA); man-in-the-middle (enforce HTTPS and validate certificates). |
| 24 | (d) Of the six data protection principles in Hong Kong Personal Data (Privacy) Ordinance (Cap. 486), explain the two most relevant to collecting and using customer data, with one example each. | Principle 1 (purpose and manner of collection): data must be collected for a lawful purpose directly related to the function, adequate but not excessive, and the purpose must be notified, for example a trial-class form collecting only a name and a phone number. Principle 3 (use): unless the data subject consents or the law allows, data may be used only for the purpose stated at collection, so a contact list may not be passed to a third party for marketing without consent. |
Find a tutor — freeTutor Registration →
A TutorZone practice paper; original questions and passages, not publisher material.



