Networking and the Internet - Worksheets, Questions and Revision
13 original exam-style questions - 10 pages of questions with a full mark scheme - free printable PDF.
Networking and the Internet revision guideMore A Level Computer Science practiceMock exams
This topic is chapter 10 of A Level Computer Science Topic Workbook 1 and chapter 8 of A Level Computer Science 10-Minute Tests.
1.10 Networking and the Internet
Networking case studies: college networks, file-sharing and online retail security
Original scenarios written for Revision Library.
Questions 1 and 2 follow Ravensmead Sixth Form College's network technician, Owen Fairweather, as he chooses a topology and identifies the hardware needed for a new ICT suite. Questions 5, 6, 7, 8, 9, 12 and 13 cover the core networking theory behind that design: client-server versus peer-to-peer architectures (via a file-sharing startup, SwiftShare), the TCP/IP protocol stack, IPv4 addressing and subnetting, DNS resolution, packet switching versus circuit switching, and the full sequence of events behind a single web request. Questions 3 and 10 look at how an online retailer, GreenLeaf Retail, defends its website and customer data against network security threats, including encryption and digital certificates. Questions 4 and 11 close with wireless networking standards and security, and the ongoing migration from IPv4 to IPv6.
Question 1
- (a) the layout/arrangement of the devices (nodes) and the connections between them in a network (oe)B1
- (a) Answer: A network topology is the layout or arrangement of the devices (nodes) in a network and the connections between them.
- (b) every device is connected directly to a central device (a switch or hub), rather than to each other (oe)B1
- (b) data sent from one workstation travels to the central switch first, which then forwards it on to the destination workstation (oe)B1
- (b) Answer: In a star topology, every workstation connects directly to a central switch rather than to each other. When one workstation sends data to another, the data travels first to the central switch, which then forwards it on to the intended destination workstation.
- (c) advantage: if one workstation's cable fails, only that workstation is affected; the rest of the network keeps working (oe)B1
- (c) disadvantage: the central switch is a single point of failure; if it fails, the whole network goes down (oe, allow: uses more cabling than a bus)B1
- (c) Answer: Advantage: if one workstation's cable fails, only that workstation loses its connection, and the rest of the ICT suite's network keeps working normally. Disadvantage: the central switch is a single point of failure, so if it fails, every workstation loses its connection at once.
- (d) in a full mesh, every device has a direct connection to every other device, rather than all devices connecting only to one central switch (oe)B1
- (d) because there is no single central device that all traffic must pass through, the failure of any one device or connection does not disconnect the rest of the network (oe)B1
- (d) multiple alternative direct paths exist between devices, so data can still reach its destination via a different connection, unlike a star, where losing the central switch disconnects everyone (oe)B1
- (d) Answer: In a full mesh, every device is connected directly to every other device, rather than all devices connecting only to one central switch as in a star. This means there is no single central device that all traffic depends on, so the failure of any one device or cable does not disconnect the rest of the network: data can still reach its destination via one of the other direct connections. In a star, by contrast, if the central switch fails, every workstation loses its connection at once.
- (e) star topology recommendedB1
- (e) advantage 1: a single cable fault in a star only disconnects the one workstation concerned, whereas a break in a bus's shared backbone cable can bring down the whole network segment (oe)B1
- (e) advantage 2: a star suffers less congestion/fewer collisions as more workstations are added, because each has its own dedicated connection to the switch, whereas a bus shares one single cable between all devices (oe)B1
- (e) Answer: Owen should choose a star topology. First, in a star, a single faulty or damaged cable only disconnects the one workstation it serves, whereas in a bus topology a break anywhere in the shared backbone cable can bring down the whole network. Second, a star suffers less congestion as more workstations are added, because each workstation has its own dedicated connection to the switch, whereas in a bus every device shares the same single cable, increasing the chance of data collisions as traffic grows.
Question 2
- (a) to uniquely identify a network interface/device on a network, e.g. for forwarding frames at the link layer (oe)B1
- (a) Answer: A MAC address uniquely identifies a network interface (device) on a network, so that data frames can be forwarded to the correct device.
- (b) recognises the address is written as six hexadecimal pairs, and each hexadecimal pair represents one byte (8 bits) (oe)M1
- (b) 48 bits caoA1
- (b) Answer: 48 bits.
- (c) a hub forwards (broadcasts) every incoming frame out of all of its other ports, regardless of which device it is actually intended for (oe)B1
- (c) a switch reads the destination MAC address in the frame and looks it up in its MAC address table (oe)B1
- (c) the switch then forwards the frame only out of the specific port connected to the intended destination device, rather than to every device (oe)B1
- (c) Answer: A hub simply forwards (broadcasts) every incoming frame out of all of its other ports, regardless of which device the frame is actually intended for. A switch instead reads the destination MAC address in each frame, looks it up in its stored MAC address table, and forwards the frame only out of the one port connected to the intended destination device, rather than to every device.
- (d) a switch only forwards data between devices on the same local network, using MAC addresses; it does not route data between different networks using IP addresses (oe)B1
- (d) a router connects and routes traffic between two different networks (the college's internal network and the Internet) using IP addresses, and typically also performs other functions such as NAT, which a switch cannot do (oe)B1
- (d) Answer: A switch only forwards data between devices on the same local network, based on MAC addresses; it has no way of routing data between different networks. A router is needed instead, because it connects two separate networks together (here, the college's internal network and the Internet) and routes traffic between them using IP addresses, typically also handling functions such as Network Address Translation, which a switch cannot do.
Question 3
- (a) phishing correctly identifiedB1
- (a) the attacker sends a fraudulent email that impersonates a trusted organisation (here, GreenLeaf), tricking the recipient into clicking a link and/or revealing sensitive information (e.g. card or login details), often via a fake website designed to look genuine (oe)B1
- (a) Answer: This is a phishing attack. The attacker sends a fraudulent email that impersonates a trusted organisation, here GreenLeaf, tricking the recipient (Tariq) into clicking a link and revealing sensitive information, such as his card details, often via a fake website designed to look like the genuine one.
- (b) the attacker uses a large number of compromised devices (a botnet) to send an overwhelming volume of requests/traffic to the target web server, all at the same time (oe)B1
- (b) this consumes the server's available bandwidth, processing capacity, or connection limits (oe)B1
- (b) as a result, the server cannot process/respond to legitimate requests from genuine customers, becoming very slow or completely unavailable (oe)B1
- (b) Answer: In a DDoS attack, the attacker uses a large number of compromised devices, forming a botnet, to send an overwhelming volume of requests to GreenLeaf's web server all at the same time. This consumes the server's available bandwidth, processing capacity or connection limits. As a result, the server cannot process legitimate requests from genuine customers, and becomes very slow or completely unavailable.
- (c) any valid technical measure, e.g. a firewall/intrusion detection or prevention system configured to detect and filter abnormal traffic, or a traffic-scrubbing/content delivery network/load-balancing service to absorb and spread the excess traffic across multiple servers (oe)B1
- (c) Answer: GreenLeaf could use a firewall or intrusion prevention system configured to detect and filter abnormal traffic patterns, or use a content delivery network / traffic-scrubbing service to absorb and spread the excess traffic across multiple servers.
Question 4
- (a) thick or dense walls/floors (e.g. containing brick, concrete or metal) can block or weaken the signal (oe)B1
- (a) interference from other wireless devices/networks operating on the same or an overlapping channel/frequency (e.g. neighbouring routers, Bluetooth devices, microwaves), or the number of devices sharing the same access point at once (oe)B1
- (a) Answer: First, thick or dense walls or floors, such as those containing brick, concrete or metal, can block or weaken the WiFi signal. Second, interference from other wireless devices or networks operating on the same or an overlapping channel, such as neighbouring routers or Bluetooth devices, can reduce reliability.
- (b) advantage: 5GHz offers a higher maximum data transfer rate, and is typically less congested/subject to interference, since fewer devices use this band (oe)B1
- (b) disadvantage: 5GHz has a shorter range and penetrates solid obstacles, such as walls, less effectively than 2.4GHz, so coverage is reduced (oe)B1
- (b) Answer: Advantage: the 5GHz band offers a higher maximum data transfer rate and is typically less congested, since fewer devices use this band. Disadvantage: the 5GHz band has a shorter range and does not penetrate solid obstacles, such as walls, as effectively as 2.4GHz, so it provides reduced coverage.
- (c) WEP uses a single, static shared key with the relatively weak RC4 encryption algorithm, which has known flaws and can be cracked (e.g. using freely available tools) relatively quickly (oe)B1
- (c) WPA3 uses a much stronger, modern encryption/authentication method (e.g. Simultaneous Authentication of Equals, SAE), generating a unique session key for each connection rather than reusing one static key (oe)B1
- (c) this makes WPA3 far more resistant to offline dictionary/brute-force attacks and eavesdropping than WEP's outdated, easily broken encryption (oe)B1
- (c) Answer: WEP uses a single, static shared key together with the relatively weak RC4 encryption algorithm, which has known flaws and can be cracked relatively quickly using freely available tools. WPA3 instead uses a much stronger, modern authentication method (Simultaneous Authentication of Equals, SAE), which generates a unique session key for each connection, rather than reusing one static key. This makes WPA3 far more resistant to offline dictionary and brute-force attacks and to eavesdropping than WEP's outdated, easily broken encryption.
Question 5
- (a) in a client-server network, one or more dedicated (central) servers provide services/resources to many client devices, whereas in a peer-to-peer network there is no dedicated server and all nodes have broadly equal status (oe)B1
- (a) Answer: In a client-server network, one or more dedicated central servers provide services or resources to many separate client devices, whereas in a peer-to-peer network there is no dedicated server, and all nodes have broadly equal status.
- (b) each node can act as a client, requesting/downloading a file or resource from another node (oe)B1
- (b) at the same time, each node also acts as a server, by making its own files/resources available and uploading them to other nodes that request them (oe)B1
- (b) Answer: Each node in a peer-to-peer network can act as a client, requesting and downloading a file from another node, while at the same time also acting as a server, by making its own files available and uploading them to other nodes that request them.
- (c) no single central server is required, so there is no single point of failure and no dedicated server hardware/bandwidth needs to be bought and maintained (oe)B1
- (c) as more users share/download a popular file, more peers hold copies of it, so the network's capacity to distribute that file can increase, rather than being limited by one server's bandwidth (oe)B1
- (c) Answer: First, no single central server is required, so there is no single point of failure, and SwiftShare does not need to buy and maintain expensive dedicated server hardware and bandwidth. Second, as more users download a popular file, more peers end up holding a copy of it, so the network's overall capacity to supply that file can increase automatically, rather than being limited by one server's bandwidth.
- (d) L3 A balanced discussion covers reliability/availability, cost/scalability and central control, applied specifically to SwiftShare, and reaches a justified overall recommendation.5-6
- (d) L2 Some accurate points are made on two of the three areas above, with limited balance and/or limited application to SwiftShare.3-4
- (d) L1 A basic, one-sided or generic comment is made on client-server versus P2P, with little application to SwiftShare.1-2
- (d) L0 No creditable content.0
- (d) Indicative content:
- Reliability/availability: in a pure P2P model, a file is only available while at least one peer holding it is online; if the original uploader (and every other holder) goes offline, the file becomes unavailable, whereas a client-server model keeps every file available on a central server as long as that server is running, giving more consistent availability.
- Cost/scalability: a client-server model requires SwiftShare to buy and maintain server storage and bandwidth for every file uploaded, with costs rising directly with the number of users and files; a P2P model instead uses the storage and bandwidth of users' own devices, so it scales more cheaply, and can even become faster for popular files as more peers share them.
- Central control: a client-server model gives SwiftShare central control, making it far easier to monitor content, remove illegal or copyrighted files, apply consistent security updates, and take down harmful material quickly; a P2P model has no central point of control, making illegal content harder to detect and remove and security harder to enforce consistently across every user's device.
- Valid overall view: e.g. a hybrid approach (a central server for indexing, search and moderation, with the actual file transfer handled peer-to-peer) could combine the moderation benefits of client-server with the cost and scalability benefits of P2P; alternatively, a clearly justified single choice of one model for a cost-sensitive early-stage startup.
- (d) Answer: Extended, levels-marked response: award marks holistically against the levels of response grid, using the indicative content as a guide to creditable points; there is no single correct wording.
Question 6
- (a) correct four layers named: Application, Transport, Internet, Link (allow Network Access/Network Interface for Link) (oe)B1
- (a) correct order given, starting nearest the application: Application, Transport, Internet, Link caoB1
- (a) Answer: Application, Transport, Internet, Link.
- (b) TCP (or UDP) correctly identified as operating at the Transport layerB1
- (b) correct function stated, e.g. TCP establishes a connection and provides reliable, ordered delivery of data between two devices (or UDP provides fast, connectionless delivery with no guarantee of reliability) (oe)B1
- (b) Answer: TCP (Transmission Control Protocol) operates at the Transport layer. Its function is to establish a connection between two devices and provide reliable, ordered delivery of data, using acknowledgements and retransmission of any lost data.
- (c) as the email data passes down through each layer of the sending device's TCP/IP stack, each layer adds its own header (and, at the link layer, typically a trailer) to the data it receives from the layer above (oe)B1
- (c) e.g. the Transport layer adds a header (containing port numbers and sequencing information), forming a segment; the Internet layer then adds a header (containing source and destination IP addresses), forming a packet (oe)B1
- (c) the original email data ends up wrapped inside multiple layers of headers, like nested envelopes, which are then removed in reverse order (de-encapsulated) by the corresponding layer at the receiving device (oe)B1
- (c) Answer: Encapsulation is the process by which each layer of the TCP/IP stack adds its own header (and, at the link layer, a trailer) to the data passed down from the layer above, as the email and its attachment travel down through the sending device's stack. For example, the Transport layer adds a header with port and sequencing information to form a segment, and the Internet layer adds a header with the source and destination IP addresses to form a packet. The result is that the original email data is wrapped inside several nested layers of headers, which are removed one by one, in reverse order, by the matching layer at the receiving device.
- (d) Transport layer: segmentB1
- (d) Link layer: frameB1
- (d) Answer: Transport layer: segment. Link layer: frame.
Question 7
- (a) to translate/resolve human-readable domain names into the numerical IP addresses that computers use to locate each other (oe)B1
- (a) Answer: DNS translates human-readable domain names (such as www.ravensmeadsfc.ac.uk) into the numerical IP addresses that computers use to locate and communicate with each other.
- (b) the browser/OS checks its local cache; since nothing is cached, the request is forwarded to a recursive DNS resolver (e.g. run by the ISP) (oe)B1
- (b) the resolver queries a root (DNS) server, which does not hold the exact record but returns the address of the relevant top-level domain (TLD) server (e.g. for .uk) (oe)B1
- (b) the resolver then queries that TLD server, which returns the address of the authoritative name server responsible for the ravensmeadsfc.ac.uk domain (oe)B1
- (b) the resolver queries the authoritative name server, which holds and returns the actual IP address; this is passed back to the browser, allowing it to connect to the correct web server (oe)B1
- (b) Answer: The browser first checks whether the IP address is already cached; since it is not, the request is sent to a recursive DNS resolver. This resolver queries a root server, which returns the address of the .uk top-level domain (TLD) server. The resolver then queries that TLD server, which returns the address of the authoritative name server for ravensmeadsfc.ac.uk. Finally, the resolver queries that authoritative name server, which holds and returns the actual IP address, and this is passed back to the browser so it can connect to the correct web server.
- (c) caching stores a previous lookup's result locally (for its TTL, time to live) so the same domain does not need to be resolved again via the full root-TLD-authoritative process, reducing lookup time and load on DNS servers (oe)B1
- (c) drawback: if the IP address for a domain changes, the cached (old) address may still be returned/used until it expires, potentially directing users to an incorrect or outdated address in the meantime (oe)B1
- (c) Answer: Caching stores the result of a previous lookup locally for a set period (its time to live), so the same domain does not need to be resolved again via the full process described in part (b), reducing lookup time and the load placed on DNS servers. One drawback is that if a domain's IP address changes, the cached (old) address may still be used until it expires, potentially sending users to an incorrect or outdated address in the meantime.
Question 8
- (a) a single, dedicated path (circuit) between the two devices is set up and reserved before any data/conversation is transmitted (oe)B1
- (a) this exclusive path remains reserved for the entire duration of the call, whether or not data is being sent at that moment, and is only released once the call ends (oe)B1
- (a) Answer: Circuit switching sets up a single, dedicated path (circuit) between the two phones before any conversation is transmitted. This path remains exclusively reserved for the whole call, whether or not either person is actually speaking at that moment, and is only released once the call ends.
- (b) packetisation: the voice data is broken up into smaller units called packets, each labelled with a header containing information such as the source and destination address and a sequence number (oe)B1
- (b) routing: each packet is then transmitted independently across the network, potentially taking a different path/route, decided at each router along the way (oe)B1
- (b) reassembly: on arrival, the packets are put back into the correct order at the destination, using their sequence numbers, to reconstruct the original voice data (oe)B1
- (b) Answer: The VoIP call's voice data is broken up into smaller packets, each with a header containing a source and destination address and a sequence number (packetisation). Each packet is then sent across the network independently, and may take a different route to reach the destination, decided by routers along the way (routing). On arrival, the packets are put back into the correct order using their sequence numbers, to reconstruct the original voice data (reassembly).
- (c) more efficient use of network bandwidth, since capacity is only used while packets are actually being sent, allowing it to be shared dynamically between many connections rather than reserved solely for one call (oe)B1
- (c) more resilient to failure, because if part of the network fails, packets can automatically be rerouted via an alternative path, whereas a circuit-switched connection using that path would simply be lost (oe)B1
- (c) Answer: First, packet switching makes more efficient use of available bandwidth, since network capacity is only used while packets are actually being transmitted, allowing it to be shared dynamically between many different connections, rather than being reserved solely for one call. Second, it is more resilient to failure, because if part of the network breaks down, packets can automatically be rerouted via an alternative path, whereas a circuit-switched connection relying on that path would simply be lost.
- (d) because packets can take different routes and may be delayed, lost, or arrive out of order, the delay between packets can vary (jitter), and late packets may need to be dropped (oe)B1
- (d) for a real-time application like VoIP, this can cause noticeable problems such as choppy audio or gaps in the call, unlike circuit switching, which guarantees a constant amount of dedicated bandwidth for the whole call (oe)B1
- (d) Answer: Because packets can take different routes and may be delayed, lost or arrive out of order, the delay between successive packets can vary (jitter), and packets that arrive too late may have to be dropped. For a real-time application like a VoIP call, this can cause noticeable problems such as choppy audio or gaps in the conversation, unlike circuit switching, which guarantees a constant, dedicated amount of bandwidth for the entire call.
Question 9
- (a) the address used to locate/identify a specific resource (e.g. a webpage) on the Internet, typically specifying the protocol, the domain name (host) and the path to the resource (oe)B1
- (a) Answer: A URL is the address used to locate a specific resource, such as a webpage, on the Internet. It typically specifies the protocol to use (e.g. https), the domain name of the host, and the path to the particular resource.
- (b) the browser checks whether the domain name's IP address is already cached; if not, it sends a request to resolve it to a DNS resolver (oe)B1
- (b) the DNS system resolves the domain name into the web server's IP address (e.g. via the root, TLD and authoritative name servers) and returns it to the browser (oe)B1
- (b) the browser establishes a TCP connection with the web server at that IP address, typically via a three-way handshake (oe)B1
- (b) the browser sends an HTTP (or HTTPS) request over this connection, asking the server for the specific webpage/resource (oe)B1
- (b) the web server processes the request and sends back an HTTP response, containing a status code and the requested HTML document (oe)B1
- (b) the browser receives the HTML, requests any additional linked resources it needs (such as images, CSS or JavaScript files), then parses and renders the complete page for display (oe)B1
- (b) Answer: First, the browser checks whether the domain name's IP address is already cached; if not, it sends a request to a DNS resolver. The DNS system resolves the domain name into the web server's IP address and returns it to the browser. The browser then establishes a TCP connection with that IP address, typically via a three-way handshake. Next, the browser sends an HTTP (or HTTPS) request over this connection, asking for the specific webpage. The web server processes the request and sends back an HTTP response containing the requested HTML. Finally, the browser receives the HTML, requests any additional resources it needs (such as images, CSS or JavaScript), and then parses and renders the complete page on screen.
Question 10
- (a) symmetric encryption uses a single (the same) key to both encrypt and decrypt the data (oe)B1
- (a) asymmetric encryption uses a mathematically linked pair of keys instead: a public key, which can be shared openly and used to encrypt data, and a private key, kept secret by the recipient, used to decrypt it (oe)B1
- (a) symmetric encryption is faster/less computationally intensive, making it more efficient for encrypting large amounts of data, but requires a secret key to be exchanged safely in advance; asymmetric encryption solves the key-exchange problem but is slower/more computationally intensive (oe)B1
- (a) Answer: Symmetric encryption uses a single key to both encrypt and decrypt the data. Asymmetric encryption instead uses a mathematically linked pair of keys: a public key, which can be shared openly and used to encrypt data, and a private key, kept secret by its owner, used to decrypt it. Symmetric encryption is faster and less computationally intensive, so it is more efficient for encrypting large amounts of data, but it requires a secret key to be exchanged safely in advance; asymmetric encryption solves this key-exchange problem, but is slower and more computationally intensive.
- (b) using asymmetric encryption means no secret key has to be shared/transmitted over the insecure Internet connection in advance, removing the risk of an attacker intercepting it in transit (oe)B1
- (b) only the intended recipient's private key (which is never shared) can decrypt data encrypted with the corresponding public key, so the session key can be sent securely even though the public key itself is sent in the open (oe)B1
- (b) Answer: Using asymmetric encryption for the initial exchange means no secret key ever has to be shared over the insecure Internet connection in advance, removing the risk of an attacker intercepting a shared key in transit. Only the intended recipient's private key, which is never shared with anyone, can decrypt data that has been encrypted with the corresponding public key, so the symmetric session key can be sent securely even though the public key itself is sent openly.
- (c) a digital certificate, issued by a trusted Certificate Authority (CA), binds/confirms that a specific public key belongs to GreenLeaf's genuine web server (oe)B1
- (c) the browser checks the certificate's digital signature against the CA's own trusted public key, to verify that the certificate is authentic and has not been altered/forged (oe)B1
- (c) this allows the browser to trust that it is really communicating with GreenLeaf's genuine server, rather than an imposter carrying out a man-in-the-middle attack, before the session key is exchanged and the connection is encrypted (oe)B1
- (c) Answer: A digital certificate, issued by a trusted Certificate Authority, confirms that a specific public key genuinely belongs to GreenLeaf's web server. The customer's browser checks the certificate's digital signature against the Certificate Authority's own trusted public key, to verify that the certificate is authentic and has not been forged or altered. This lets the browser trust that it is really communicating with GreenLeaf's genuine server, rather than an imposter carrying out a man-in-the-middle attack, before the session key described in part (b) is exchanged and the connection is encrypted.
- (d) L3 A balanced evaluation covers both the value and the limits of technical measures (firewall/encryption) and the value of staff training, applied to GreenLeaf's scenario, reaching a justified overall conclusion.5-6
- (d) L2 Some accurate points are made about technical measures and/or training, but with limited balance or limited application to GreenLeaf's scenario.3-4
- (d) L1 A basic or generic comment is made about security measures, with little reference to GreenLeaf's scenario.1-2
- (d) L0 No creditable content.0
- (d) Indicative content:
- A firewall blocks unauthorised network traffic/access attempts and can filter known malicious patterns, while encryption protects data in transit and at rest from being read if intercepted, so together they defend against many purely technical attacks, such as the DDoS traffic and network eavesdropping described in Question 3.
- Limitation: neither a firewall nor encryption prevents a phishing attack, because phishing exploits human trust and behaviour, an employee or customer being tricked into willingly revealing information or credentials, rather than a technical vulnerability in the network or the data itself; Tariq's suspicious email in Question 3 demonstrates this weakness directly.
- Training staff to recognise phishing emails, e.g. checking sender addresses, spotting urgent or suspicious requests, and not clicking unexpected links, addresses this human factor, reducing the chance that an attacker gains valid credentials or access by deception, which a firewall or encryption cannot prevent once a legitimate password has been given away.
- Balanced conclusion: firewalls and encryption remain essential technical defences and should not be dropped, but they only address technical routes of attack; a robust strategy for GreenLeaf needs both technical measures (firewall, encryption, digital certificates) and organisational/human measures (staff training, awareness), since many real breaches begin with a successfully phished employee or customer rather than a direct technical breach.
- (d) Answer: Extended, levels-marked response: award marks holistically against the levels of response grid, using the indicative content as a guide to creditable points; there is no single correct wording.
Question 11
- (a) leading zeros correctly removed within each group and/or the longest run of consecutive all-zero groups correctly identified (the three groups in positions 3 to 5)M1
- (a) 2001:db8::ff00:42:8329 caoA1
- (a) Answer: 2001:db8::ff00:42:8329
- (b) IPv4's 32-bit address space (around 4.3 billion addresses) has effectively run out/been exhausted, given the huge growth in the number of Internet-connected devices, so a much larger address space was needed (oe)B1
- (b) Answer: IPv4's 32-bit address space of around 4.3 billion addresses has effectively run out, given the huge growth in the number of Internet-connected devices, so IPv6's much larger 128-bit address space was introduced to provide enough unique addresses for the future.
- (c) L3 A clear discussion of at least three distinct, accurate challenges (e.g. compatibility/transition mechanisms, cost of upgrading hardware, and the slow/gradual pace of adoption), reaching a supported conclusion.5-6
- (c) L2 Some accurate challenges are identified, with limited depth or development.3-4
- (c) L1 A basic or generic comment is made about IPv4 to IPv6 migration, with little development.1-2
- (c) L0 No creditable content.0
- (c) Indicative content:
- Compatibility: IPv4-only and IPv6-only devices/networks cannot communicate directly, requiring transition mechanisms such as dual-stack (a device/router runs both protocols at once) or tunnelling (encapsulating IPv6 packets inside IPv4 packets to cross IPv4-only parts of the network), which add complexity and cost.
- Cost/hardware: much existing networking hardware and software was designed only for IPv4 and may need replacing or updating to support IPv6, a significant expense for large organisations/ISPs that provides little immediate visible benefit to them.
- Gradual/slow adoption: because no single central authority can force the whole Internet to switch at once, migration must happen gradually, with IPv4 and IPv6 coexisting for an extended period; this is complicated further because NAT (Question 12) has let IPv4's limited address space last far longer than expected, reducing the urgency to migrate.
- Skills/administration: network staff need to learn new addressing and configuration conventions for IPv6's much larger 128-bit address space, and reconfigure existing security tools, firewalls and routing tables that were built around IPv4.
- Valid conclusion: e.g. migration is technically necessary in the long run given IPv4 address exhaustion, but is being slowed by cost, compatibility issues, and the continued effectiveness of interim workarounds such as NAT.
- (c) Answer: Extended, levels-marked response: award marks holistically against the levels of response grid, using the indicative content as a guide to creditable points; there is no single correct wording.
Question 12
- (a) three of the four octets converted correctlyM1
- (a) 10101100.00010000.00000101.11001000 cao (allow without full stops)A1
- (a) Answer: 10101100.00010000.00000101.11001000
- (b) 8 host bits correctly identifiedM1
- (b) 28 = 256 total addresses calculatedM1
- (b) 254 usable host addresses (256 minus 2 for the network and broadcast addresses) caoA1
- (b) Answer: 8 host bits; 254 usable host addresses.
- (c) a public IP address is globally unique and directly routable/reachable over the Internet, assigned by an ISP or regional Internet registry (oe)B1
- (c) a private IP address (e.g. from 192.168.x.x, 10.x.x.x or 172.16.x.x-172.31.x.x) is only unique within its own local network and is not routable directly over the Internet; separate private networks can reuse the same private ranges (oe)B1
- (c) Answer: A public IP address is globally unique and can be reached directly over the Internet; it is assigned by an ISP or a regional Internet registry. A private IP address (from a reserved range such as 192.168.x.x, 10.x.x.x or 172.16.x.x to 172.31.x.x) is only unique within its own local network and cannot be routed directly over the Internet; separate private networks, such as different homes, can reuse the same private address ranges without conflict.
- (d) NAT is performed by the router at the boundary between the private local network and the Internet (oe)B1
- (d) when outgoing data leaves the local network, the router translates each device's private IP address (and port number) into its own single public IP address (and a different port number), recording the mapping in a translation table (oe)B1
- (d) when a reply arrives back at the public IP address, the router uses the translation table to determine which internal device/port the reply should be forwarded to (oe)B1
- (d) Answer: NAT is carried out by the router at the boundary between the local network and the Internet. When a device on the local network sends data out to the Internet, the router translates that device's private IP address (and port number) into its own single public IP address (with a different port number), and records this mapping in a translation table. When a reply comes back addressed to the public IP address, the router looks up the translation table to work out which internal device (and port) the reply belongs to, and forwards it there.
Question 13
- (a) TCP is connection-oriented, establishing a connection (e.g. via a three-way handshake) before sending data, whereas UDP is connectionless and sends data without first establishing a connection (oe)B1
- (a) TCP provides reliable, ordered delivery using acknowledgements and retransmission of lost packets, whereas UDP gives no such guarantee but has lower overhead/latency (oe)B1
- (a) Answer: TCP is connection-oriented, establishing a connection (e.g. via a three-way handshake) before sending any data, whereas UDP is connectionless and simply sends data without first setting up a connection. TCP also guarantees reliable, ordered delivery, using acknowledgements and retransmitting any lost data, whereas UDP gives no such guarantee, but has lower overhead and therefore lower latency.
- (b) HTTP: 80B1
- (b) HTTPS: 443B1
- (b) FTP (control): 21B1
- (b) SMTP: 25B1
- (b) Answer: HTTP: 80. HTTPS: 443. FTP (control connection): 21. SMTP: 25.
- (c) live streaming needs data delivered with minimal delay, and can usually tolerate the occasional lost or corrupted packet (e.g. a brief glitch) without significantly affecting the viewer's experience (oe)B1
- (c) UDP has lower overhead than TCP, because it does not perform a connection handshake, acknowledgements or retransmission, so data can be sent and displayed with less delay (oe)B1
- (c) if TCP were used, a lost packet would need to be retransmitted and the stream would have to wait/buffer for it, introducing a delay more disruptive to a live broadcast than simply dropping that one packet (oe)B1
- (c) Answer: Live video streaming needs data to arrive with minimal delay, and can usually tolerate the occasional lost or corrupted packet, such as a brief visual glitch, without significantly affecting the viewer's experience. UDP has lower overhead than TCP, since it performs no connection handshake, acknowledgements or retransmission, so packets can be sent and displayed with less delay. If TCP were used instead, a lost packet would have to be retransmitted, forcing the stream to wait or buffer for it, which introduces a delay far more disruptive to a live broadcast than simply dropping that one packet would be.
- (d) a webpage's HTML and resources must arrive complete and in the correct order for the page to display and function correctly, so TCP's reliable delivery and correct sequencing are important (oe)B1
- (d) a missing or corrupted packet in a webpage would produce a broken or incomplete page, which is generally far more disruptive/noticeable than the brief quality dip UDP-based streaming can tolerate (oe)B1
- (d) loading a webpage does not usually require the strict real-time delivery that streaming/VoIP need, so the small extra delay caused by TCP's handshake and acknowledgements is an acceptable trade-off for guaranteed accuracy (oe)B1
- (d) Answer: A webpage's HTML and other resources must arrive complete, uncorrupted and in the correct order for the page to display and work correctly, so TCP's guarantees of reliable, ordered delivery (using acknowledgements and retransmission) are important. A missing or corrupted packet would produce a broken or incomplete page, which is far more disruptive to the user than the brief quality dip that UDP-based streaming can tolerate. Loading a webpage also does not need the strict real-time delivery that live streaming or VoIP require, so the small extra delay from TCP's handshake and acknowledgements is an acceptable trade-off for guaranteeing the page arrives correctly.
Mark your answers
Check answers without an account. Some answers can be checked automatically; use the mark scheme to mark the rest.