Mobile Internet Protocols and Transport

Introduction and Context

Mobile computing changes the assumptions on which the traditional Internet was designed. A conventional Internet host is usually associated with a relatively stable point of attachment, a persistent path, and modest variation in delay and packet loss. A mobile host may change its point of attachment while communicating, use several interfaces simultaneously, experience radio-induced loss, enter intermittent coverage, or operate through a network with high and variable latency. These conditions affect addressing, routing, session continuity, congestion control, security, and application performance.

This chapter examines the protocols and design principles that support Internet communication in mobile environments. It begins with Mobile IPv4 and IPv6 mobility, then considers routing and addressing, handover and session continuity, transport behaviour over wireless links, TCP variants, QUIC, disconnections, and delay-tolerant networking. The central theme is that mobility is not merely a routing problem. It is a cross-layer problem involving identifiers, locators, signalling, path selection, transport state, security, and application semantics.

Learning Objectives

After completing this chapter, the learner should be able to:

1. Explain the distinction between an IP address as an identifier and as a topological locator.

2. Describe Mobile IPv4, Mobile IPv6, route optimization, and the role of mobility anchors.

3. Analyse handover, binding updates, tunnelling, triangular routing, and session continuity.

4. Explain why wireless loss, delay variation, bandwidth fluctuation, and disconnection affect TCP.

5. Compare traditional TCP, wireless-aware TCP variants, multipath transport, and QUIC.

6. Select suitable mobility and transport mechanisms for practical mobile-network scenarios.

7. Evaluate delay-tolerant networking when an end-to-end path is not continuously available.

8. Identify security, privacy, operational, and regulatory issues in mobile Internet design.

1. Mobility, Naming, and the Internet Architecture

1.1 Location and identity

In ordinary IP networking, an IP address performs two functions: it identifies an interface and indicates where that interface is attached to the routing topology. Mobility separates these functions. A mobile node needs a relatively stable identity for an ongoing communication session while acquiring a new locator at each visited network.

Several identifiers may coexist:

• A link-layer identifier identifies a radio interface within a local technology.
• An IP address identifies a network-layer attachment and is routable within a particular topology.
• A transport endpoint is commonly represented by an address, port number, and protocol.
• An application identity may be a domain name, account, device identity, or service name.

The fundamental mobility problem is therefore: how can packets continue to reach a node when its topological locator changes, without forcing every application to restart?

1.2 Mobility dimensions

A host is nomadic when it moves between networks but does not necessarily preserve active sessions. It is mobile when it changes attachment while maintaining communication. Roaming often refers to operation across administrative domains, such as a subscriber using a foreign cellular network. Mobility may also be classified as:

• Host mobility: a device changes its point of attachment.
• Network mobility: an entire subnet moves, as in an aircraft or vehicle network.
• Session mobility: an application session moves between interfaces or devices.
• Service mobility: a service instance changes its execution location.

These forms may occur together. For example, a vehicle may move across cellular cells while carrying a local Wi-Fi network whose devices retain their private addresses.

2. Mobile IPv4

2.1 Architectural entities

Mobile IPv4 introduces a framework that allows a mobile node to retain a permanent home address while visiting another network. The principal entities are:

• Mobile node (MN): the moving host.
• Home agent (HA): a router on the home network that maintains reachability for the mobile node.
• Foreign agent (FA): a router on the visited network that may provide a care-of address and tunnel delivery to the mobile node.
• Correspondent node (CN): the peer communicating with the mobile node.

When the mobile node is away from its home network, the home agent intercepts packets addressed to the home address and tunnels them to a care-of address (CoA). The care-of address represents the mobile node’s current location.

A care-of address may be foreign-agent based or co-located. In the first case, the foreign agent decapsulates packets and delivers them locally. In the second, the mobile node obtains an address through mechanisms such as DHCP and performs decapsulation itself.

2.2 Registration and tunnelling

The mobile node discovers mobility agents, obtains or learns a care-of address, and registers that address with the home agent. Registration messages are authenticated because an attacker who changed the binding could redirect traffic.

A common packet path is:

CN → HA → FA → MN.

The home agent encapsulates the original IP packet inside another IP packet. This is called IP-in-IP tunnelling; other encapsulation methods may also be used. When the mobile node sends packets directly to the correspondent node, the resulting path may be asymmetric.

2.3 Triangular routing and limitations

Mobile IPv4 commonly produces triangular routing: traffic from the correspondent node travels through the home agent to the mobile node, while return traffic may travel directly to the correspondent node. This can increase delay, consume home-network bandwidth, and create a bottleneck at the home agent.

Important limitations include:

• Dependence on foreign-agent support in some deployments.
• Additional encapsulation overhead.
• Possible ingress filtering problems for packets sent with the home address as source.
• Longer paths and increased signalling during movement.
• Security exposure if registration authentication is weak.
• Difficulty traversing some forms of NAT.

Mobile IPv4 remains conceptually important, but large-scale Internet mobility has increasingly favoured IPv6 mechanisms and application-layer approaches.

3. IPv6 Mobility

3.1 Addressing model

IPv6 mobility uses a permanent home address and a temporary care-of address, but integrates mobility more closely with IPv6 extension headers and neighbour-discovery procedures. A mobile node may configure a care-of address using IPv6 address autoconfiguration and determine whether the address is usable through duplicate-address detection and reachability procedures.

The home agent maintains a binding between the home address and care-of address. A mobile node can send a binding update to its home agent and, subject to security requirements, to a correspondent node. A correspondent node that accepts the binding can send packets directly to the care-of address rather than routing them through the home agent.

3.2 Route optimization

With route optimization, the path is approximately:

CN → visited network → MN.

This reduces delay and eliminates unnecessary tunnelling through the home network. The correspondent node must, however, verify that the mobile node actually owns the claimed home address and care-of address. IPv6 mobility therefore uses a return-routability procedure and authenticated signalling rather than trusting an unauthenticated binding update.

Route optimization improves efficiency but introduces operational complexity. Firewalls, middleboxes, address translators, and policy controls may not handle changing endpoint addresses uniformly. Some networks deliberately prefer a mobility anchor or proxy-based mechanism because it provides policy enforcement, lawful-interception support, accounting, and simpler device behaviour.

3.3 Proxy and network-based mobility

In host-based mobility, the mobile node participates in mobility signalling. In network-based mobility, network entities perform signalling on behalf of the host. Proxy Mobile IPv6 is an example: a local mobility anchor maintains the mobile node’s reachability while access routers signal movement. The device can therefore remain unaware of mobility at the IP layer.

Host-based mechanisms offer greater end-to-end control and may support direct route optimization. Network-based mechanisms can simplify terminals and centralize operational policy. The choice depends on administrative control, device capability, security architecture, and the required handover latency.

4. Routing, Addressing, and Session Continuity

4.1 Handover phases

A handover normally consists of:

1. Movement detection: the device or network determines that the current attachment is weakening or that a new attachment is available.

2. Preparation: authentication, link setup, address configuration, and resource reservation occur on the target network.

3. Binding or path update: mobility entities learn the new locator.

4. Packet forwarding: traffic is redirected to the new path.

5. Release and optimization: the old path is removed and buffered or duplicated packets are reconciled.

A handover may be hard, with a period of disconnection, or soft, with overlapping connectivity. Make-before-break operation generally reduces interruption but may require two active interfaces and additional energy.

4.2 Continuity mechanisms

Session continuity can be supported at several layers:

• Network layer: Mobile IP preserves a stable home address.
• Transport layer: Multipath TCP or QUIC connection migration can preserve a connection while changing paths.
• Session layer: a protocol may use a stable session identifier and re-establish transport connections.
• Application layer: applications checkpoint state and resume transfers after interruption.

A stable IP address alone does not guarantee continuity. Transport state includes sequence numbers, congestion windows, retransmission timers, cryptographic keys, and path characteristics. If the transport layer interprets a handover outage as severe congestion, it may reduce its sending rate unnecessarily.

4.3 Packet loss, buffering, and reordering

During handover, packets may be lost, duplicated, delayed, or delivered out of order. A mobility anchor can buffer packets, forward packets from the old access router to the new one, or duplicate traffic temporarily. Buffering reduces loss but increases memory use and may deliver stale data. Forwarding reduces interruption but may create a temporary chain of tunnels. Duplication improves reliability at the cost of bandwidth.

Designers should distinguish loss caused by congestion from loss caused by wireless errors or movement. This distinction is difficult in an end-to-end architecture because packet loss is observed at the receiver without direct knowledge of its cause.

5. Transport Behaviour over Wireless Links

5.1 Why traditional TCP struggles

TCP provides reliable, ordered byte-stream delivery and congestion control. Classical TCP assumes that packet loss is predominantly a signal of congestion. This assumption is often reasonable in wired networks but is imperfect in mobile networks, where loss may result from:

• Radio interference and fading.
• Temporary coverage gaps.
• Handover interruption.
• Route changes.
• Deep device sleep.
• Queue overflow caused by a sudden rate change.

TCP also reacts to variable round-trip time (RTT), bandwidth changes, and packet reordering. A wireless link may have high capacity when conditions are good and very low capacity seconds later. A congestion-control algorithm that is too aggressive creates queues and energy consumption; one that is too conservative wastes available capacity.

5.2 Basic TCP quantities

For a long-lived flow, a rough throughput relationship is often used to understand the impact of loss:

T ≈ C × MSS / (RTT × √p),

where T is estimated steady-state throughput, MSS is maximum segment size, RTT is round-trip time, p is the probability of packet loss, and C is a model-dependent constant. This is not a universal TCP law. It assumes a long-lived congestion-avoidance flow, approximately independent losses, stable RTT, and a Reno-like response. It should not be applied directly to short flows, modern model-based algorithms, wireless-error bursts, or changing paths.

The expression shows why even a modest increase in RTT or loss can reduce throughput. It also illustrates why a mobile path with fluctuating RTT may experience unstable sending rates.

TCP maintains a congestion window, often denoted cwnd, and a receiver-advertised window, rwnd. The usable sending window is bounded by both network congestion and receiver capacity. A simplified statement is:

Flight size ≤ min(cwnd, rwnd).

The flight size is the amount of unacknowledged data. This relation does not itself specify the congestion-control algorithm; it expresses a fundamental sending constraint.

5.3 TCP mechanisms relevant to mobility

Slow start increases the congestion window rapidly at the beginning of a flow or after a severe reduction. It can overrun a wireless bottleneck or radio buffer. Congestion avoidance typically increases the window more cautiously. Fast retransmit responds to duplicate acknowledgements, while fast recovery attempts to avoid returning fully to the initial sending rate after a moderate loss.

Selective Acknowledgement (SACK) allows the receiver to report non-contiguous blocks received successfully. SACK is valuable when multiple segments are lost or reordered, because the sender need not retransmit data already received.

Explicit Congestion Notification (ECN) allows routers to mark packets rather than discard them when congestion is building. ECN can reduce unnecessary retransmissions, but it requires support and correct configuration across the path.

Accurate RTT estimation is essential. A simplified retransmission timeout model is commonly represented as:

RTO = SRTT + K × RTTVAR,

where SRTT is the smoothed RTT, RTTVAR is the smoothed RTT variation, and K is a protocol-defined safety factor. Real TCP specifications include carefully defined update rules, minimum bounds, and timer backoff. The expression should therefore be treated as a conceptual form, not a complete implementation specification.

6. TCP Variants and Multipath Transport

6.1 Wireless-aware approaches

Several approaches have been proposed to improve TCP over wireless paths:

• Indirect TCP: splits one end-to-end connection at a base station. The wireless segment can be handled separately, but end-to-end semantics and failure transparency are weakened.
• Snooping TCP: a base station caches packets and locally retransmits segments lost on the wireless hop. Encryption and route changes can limit visibility and effectiveness.
• Explicit-loss notification: intermediate nodes inform the sender that a loss was caused by a wireless error rather than congestion. This can prevent an unnecessary congestion-window reduction, but requires trustworthy signalling.
• Freeze TCP: the receiver or proxy warns the sender before an anticipated disconnection, allowing the sender to suspend timers and transmission. It is useful only when disconnection can be predicted and signalling arrives in time.

These mechanisms illustrate a fundamental trade-off. Cross-layer assistance can improve performance, but it increases dependence on intermediaries and may conflict with end-to-end encryption or changing paths.

6.2 Modern congestion-control algorithms

CUBIC uses a cubic window-growth function intended to improve scalability on high-bandwidth paths. BBR estimates bottleneck bandwidth and propagation delay and attempts to operate near the estimated delivery rate rather than relying solely on loss as a congestion signal. Neither algorithm eliminates mobility challenges: both must adapt to rapidly changing capacity, delayed feedback, radio scheduling, and path changes.

A good evaluation should consider throughput, 95th-percentile latency, loss, fairness, energy consumption, buffer occupancy, and recovery time after handover—not throughput alone.

6.3 Multipath TCP

Multipath TCP (MPTCP) permits one logical connection to use several subflows, such as cellular and Wi-Fi. It can provide resilience, aggregate capacity, or smoother handover. A device may move traffic from one interface to another without presenting a new application connection.

MPTCP must address coupled congestion control, packet scheduling, reordering, authentication of additional subflows, and unequal path quality. Sending data over all interfaces may increase cost and energy use. A policy may instead reserve one interface as backup. Multipath is therefore a resource-management decision as well as a transport feature.

7. QUIC and Transport Evolution

QUIC is a secure transport protocol commonly implemented over UDP. It integrates TLS 1.3 security establishment, provides streams within a connection, and supports connection identifiers that can remain stable when the peer’s network address changes. These features are highly relevant to mobile operation.

7.1 Advantages for mobile systems

QUIC can reduce connection-establishment delay by combining transport and cryptographic handshakes. Stream multiplexing avoids the exact form of cross-stream head-of-line blocking found in a single ordered TCP byte stream: loss on one stream need not prevent delivery of independent streams. Connection migration allows a client to move from Wi-Fi to cellular while retaining cryptographic connection state, provided the protocol’s validation and security procedures succeed.

QUIC still experiences congestion and loss. UDP does not make congestion control optional, and a QUIC implementation must implement congestion control and respect network capacity. Migration can also expose a new RTT, MTU, firewall policy, or path bottleneck. Connection identifiers improve continuity but do not guarantee that middleboxes will permit the new path.

7.2 Operational considerations

QUIC encryption protects payloads and much transport metadata, improving privacy but reducing the visibility available to traditional network monitoring and middleboxes. Network operators increasingly rely on endpoint telemetry, approved observability interfaces, and policy-compliant inspection rather than indiscriminate packet inspection. Deployments must balance confidentiality, troubleshooting, lawful requirements, and data minimization.

8. Disconnections and Delay-Tolerant Networking

8.1 Intermittent connectivity

Some mobile systems cannot maintain a continuous end-to-end path. Examples include remote sensors, rural vehicles, disaster-response teams, maritime networks, and interplanetary communication. Repeatedly applying ordinary TCP in such environments leads to timeouts, repeated handshakes, and loss of application progress.

A disruption-tolerant design treats communication as a sequence of storage, custody, and forwarding operations. Data is carried in complete messages or bundles, stored at intermediate nodes, and forwarded when a contact becomes available. This is the principle of store-carry-forward networking.

8.2 Delay-Tolerant Networking architecture

The Bundle Protocol provides an overlay suitable for challenged networks. It supports persistent identifiers, message lifetime, administrative records, fragmentation, and optional custody-related functions. Routing may use scheduled contacts, predicted mobility, social relationships, delivery probability, or epidemic replication.

Routing strategies include:

• Direct delivery: a source waits until it meets the destination.
• First contact: a message is forwarded to the first suitable relay.
• Epidemic routing: copies are widely replicated, increasing delivery probability but consuming storage and energy.
• Spray and wait: only a controlled number of copies is created.
• Contact-graph routing: known future contacts are used to calculate a route.

DTN protocols require expiration policies, storage quotas, duplicate suppression, priority classes, and abuse controls. The objective is not necessarily low instantaneous latency; it may be eventual delivery under intermittent opportunities.

8.3 When to use DTN

DTN is appropriate when disruption is expected, contact opportunities can be exploited, and applications can tolerate delay or asynchronous delivery. It is inappropriate for interactive voice or safety control requiring bounded latency unless a separate continuously connected path exists.

9. Worked Application: Handover in an Emergency-Response Vehicle

Consider an illustrative emergency-response vehicle carrying video cameras, telemetry sensors, and a command tablet. It uses a high-capacity but coverage-limited Wi-Fi link near an incident site and a cellular link with broad coverage. The video application should continue during movement, while sensor data must eventually reach the command centre even if both links fail temporarily.

Problem. A naïve design binds the video session to the Wi-Fi address. When the vehicle leaves Wi-Fi coverage, the TCP connection stalls and the video restarts over cellular. Sensor packets sent during a tunnel are lost.

Analysis. The system has three different requirements. Video needs low delay and can tolerate some loss. Command messages need reliable, authenticated delivery. Sensor records can tolerate delay but should not be lost. A single transport and one addressing method are therefore unsuitable.

Approach. The vehicle uses a mobility-capable connectivity manager and maintains cellular readiness before Wi-Fi quality becomes unacceptable. For interactive video, QUIC connection migration or an application session identifier allows the flow to move to cellular. Congestion control is allowed to reacquire the new path capacity rather than assuming the old Wi-Fi rate remains valid. Sensor records are placed into persistent bundles with timestamps, sequence numbers, expiration, and integrity protection. The command service uses authenticated messages with acknowledgements and idempotent transaction identifiers, so a retransmitted command does not cause duplicate action.

During planned movement, packets may be duplicated briefly across interfaces for critical control traffic, while bulk video uses the less expensive or less congested path. The system monitors handover interruption, packet loss, RTT, video stall time, battery consumption, and cellular data usage.

Outcome. Video interruption is reduced because the cellular path is prepared before Wi-Fi release. Sensor data survives a temporary outage through store-and-forward operation. The design does not attempt to force delay-tolerant telemetry into a real-time transport connection.

Learning points. Session continuity is an application requirement, not merely an IP feature. Different data classes require different reliability and delay models. Mobility decisions should include cost, energy, security, and operational policy, not just signal strength.

10. Security, Privacy, and Governance

Mobility signalling can redirect traffic, reveal a user’s location, or enable denial of service. Essential protections include authenticated binding updates, replay protection, secure neighbour discovery, access-control enforcement, and validation of new paths. A mobility anchor must be hardened because compromising it can affect many subscribers.

Location privacy is a major concern. Frequent care-of-address changes, binding updates, radio measurements, and network logs can reveal movement patterns. Systems should minimize collection, restrict access, define retention periods, and provide transparent governance. Pseudonymous identifiers can reduce casual tracking but do not guarantee anonymity when timing and traffic patterns remain observable.

Operators must also address lawful access, emergency-service obligations, cross-border data transfer, spectrum regulation, and service-level commitments. QUIC and end-to-end encryption strengthen confidentiality but may require revised observability and incident-response procedures. Security controls should not be added only after deployment; they should be incorporated into mobility, routing, and transport design.

11. Contemporary Developments and Best Practices

Current mobile systems increasingly combine multiple access networks, edge computing, programmable data planes, and application-layer resilience. Edge services reduce RTT by placing computation near users, but mobility requires state replication or rapid service relocation. Software-defined networking can centralize policy and automate path changes, while network slicing can provide differentiated service classes. These technologies improve control but increase orchestration complexity and create larger failure domains.

AI and machine learning can predict radio quality, handover need, traffic demand, or contact opportunities. Such predictions should be treated as uncertain inputs, not unquestionable decisions. Training data may encode geographic or demographic bias, and erroneous predictions can trigger costly handovers or deny service. Human override, auditability, fallback policies, and privacy-preserving data practices are important.

Recommended engineering practices are:

1. Separate stable application identity from changing network locators.

2. Measure handover interruption, tail latency, loss bursts, reordering, energy, and cost.

3. Use make-before-break when the device and policy permit it.

4. Preserve application progress through resumable transfers and idempotent operations.

5. Use SACK, ECN, suitable congestion control, and accurate timer management.

6. Select QUIC, MPTCP, Mobile IP, or application-layer mechanisms according to the actual continuity requirement.

7. Use DTN techniques when an end-to-end path cannot be assumed.

8. Authenticate mobility signalling and protect location information.

9. Test with realistic traces: coverage gaps, asymmetric paths, link changes, loss bursts, and overloaded cells.

10. Avoid optimizing average throughput at the expense of reliability, fairness, battery life, or safety.

Chapter Summary

Mobile Internet communication must reconcile a stable communication identity with changing topological location. Mobile IPv4 uses home agents, foreign agents, care-of addresses, registration, and tunnelling, but may suffer triangular routing and additional overhead. IPv6 mobility improves autoconfiguration and supports route optimization, while proxy-based approaches move mobility intelligence into the network.

Handover involves detection, preparation, binding updates, forwarding, and cleanup. Packet loss and delay during this process can disrupt transport state, so continuity may be supported at the network, transport, session, or application layer. Traditional TCP interprets loss primarily as congestion, an assumption weakened by wireless errors, intermittent coverage, and mobility. SACK, ECN, modern congestion control, wireless-aware methods, and MPTCP address particular aspects of the problem but involve trade-offs.

QUIC combines secure transport establishment, streams, and connection migration, making it well suited to many mobile applications, although it remains subject to congestion and path changes. When continuous connectivity is impossible, Delay-Tolerant Networking uses persistent bundles and store-carry-forward routing. Practical designs should classify traffic by delay, reliability, and security requirements; evaluate performance beyond throughput; and incorporate privacy, authentication, governance, energy, and operational constraints.

Key Terms

Mobile node: A device that changes its network attachment while communicating.

Home address: A stable address associated with a mobile node’s home network.

Care-of address: A temporary locator representing a mobile node’s current attachment.

Home agent: A mobility entity that maintains reachability for a mobile node at its home network.

Foreign agent: A visited-network router that can support Mobile IPv4 delivery.

Correspondent node: The peer communicating with a mobile node.

Binding update: Signalling that associates a stable address with a current locator.

Route optimization: Directing traffic toward a mobile node’s current location rather than through a home agent.

Triangular routing: A path in which traffic travels through a home agent before reaching a mobile node.

Handover: The process of transferring connectivity from one access point or network to another.

Session continuity: Preservation or resumability of an ongoing communication session during movement.

Congestion window: Sender-controlled limit on outstanding data based on perceived network congestion.

Selective Acknowledgement: A TCP capability that reports received non-contiguous data blocks.

Explicit Congestion Notification: A mechanism that marks packets to signal congestion without dropping them.

Multipath TCP: A transport extension that uses multiple subflows for one logical connection.

QUIC: A secure, multiplexed transport protocol commonly implemented over UDP.

Connection migration: Movement of a transport connection to a new network path.

Delay-Tolerant Networking: Networking designed for intermittent connectivity and long or variable delays.

Store-carry-forward: Holding data at a node until a forwarding contact becomes available.

Contact-graph routing: Routing based on known or predicted future connectivity opportunities.

Review Questions

Short Answer Questions

1. Distinguish a home address from a care-of address.

2. What is triangular routing, and why can it be inefficient?

3. Why can wireless loss cause TCP to reduce its sending rate unnecessarily?

4. State one advantage and one limitation of QUIC connection migration.

5. Define store-carry-forward networking.

Long Answer Questions

1. Explain the operation of Mobile IPv4, including registration, tunnelling, and the roles of the home and foreign agents.

2. Compare host-based IPv6 mobility with network-based mobility.

3. Discuss the effects of handover-induced loss, delay, and packet reordering on TCP.

4. Compare MPTCP and QUIC as mechanisms for mobile session continuity.

5. Explain the architecture, routing choices, and application suitability of Delay-Tolerant Networking.

Analytical / Application Questions

1. Design a mobility architecture for a smartphone that moves between Wi-Fi and cellular networks. Justify the roles of IP mobility, transport migration, and application-layer recovery.

2. An autonomous vehicle periodically loses cellular coverage for several minutes. Determine which data should use ordinary interactive transport and which should use DTN techniques. Explain your decision using delay, reliability, and safety requirements.

3. Evaluate a proposed system that selects handovers only by received signal strength. Identify at least four missing performance or governance factors and propose a better decision framework.

Multiple Choice Questions with Answers

Question 1

In Mobile IPv4, which entity maintains reachability for a mobile node when it is away from its home network?

A. Foreign agent

B. Home agent

C. Correspondent node

D. DHCP server

Correct Answer: B. Home agent

Question 2

What is the principal purpose of a care-of address?

A. To identify the application process

B. To represent the mobile node’s current network location

C. To encrypt the home address

D. To identify the foreign agent’s hardware

Correct Answer: B. To represent the mobile node’s current network location

Question 3

Which routing pattern commonly results from basic Mobile IPv4 operation?

A. Flooding

B. Triangular routing

C. Source routing only

D. Anycast routing

Correct Answer: B. Triangular routing

Question 4

Why can traditional TCP perform poorly on a mobile wireless path?

A. It cannot use port numbers

B. It treats many non-congestion losses as congestion indications

C. It requires IPv6 exclusively

D. It never retransmits lost data

Correct Answer: B. It treats many non-congestion losses as congestion indications

Question 5

What does TCP Selective Acknowledgement enable?

A. Reporting received non-contiguous data blocks

B. Changing the IP address without signalling

C. Eliminating congestion control

D. Encrypting all application data

Correct Answer: A. Reporting received non-contiguous data blocks

Question 6

Which feature of QUIC is particularly relevant to movement between network interfaces?

A. Connection migration using connection identifiers

B. Removal of congestion control

C. Dependence on foreign agents

D. Use of only circuit-switched links

Correct Answer: A. Connection migration using connection identifiers

Question 7

What is a defining characteristic of Multipath TCP?

A. It uses multiple subflows for one logical connection

B. It disables retransmission on wireless links

C. It replaces IP addressing with MAC addressing

D. It operates only in delay-tolerant networks

Correct Answer: A. It uses multiple subflows for one logical connection

Question 8

Which technique is most appropriate when an end-to-end path may be unavailable for several minutes?

A. Ordinary slow start only

B. Store-carry-forward networking

C. Immediate route deletion

D. Unmodified real-time streaming

Correct Answer: B. Store-carry-forward networking

Question 9

In the approximate relation T ≈ C × MSS/(RTT × √p), increasing RTT while other variables remain fixed generally causes estimated throughput to:

A. Increase

B. Remain exactly unchanged

C. Decrease

D. Become independent of packet loss

Correct Answer: C. Decrease

Question 10

Which is a major security concern in mobility signalling?

A. Binding-update redirection by an attacker

B. Excessive screen brightness

C. Loss of application syntax

D. Inability to use port numbers

Correct Answer: A. Binding-update redirection by an attacker