Endsem Examination Bank: Unit 4

Section A: Network Layer Mobility (Mobile IP)

Explain why traditional IPv4 fails when a mobile node moves to a new subnet. Why does dynamically assigning a new IP address via DHCP break existing TCP connections?

Traditional Internet Protocol (IPv4) architecture operates under a fundamental, inflexible assumption: an IP address serves dual, inseparable purposes. First, it acts as a unique endpoint identifier for the device. Second, it acts as a rigid topological locator dictating exactly where that device resides within the global network tree. The network prefix of the IP address must mathematically match the physical subnet where the router is attempting to deliver the packet. If a mobile node physically moves to a new foreign subnet but retains its original Home IP address, the local routers in the foreign network will inspect the packet, realize the network prefix is entirely invalid for their local routing tables, and immediately drop the packet. The node is physically connected, but topologically exiled.

To regain topological validity, the mobile node could broadcast a DHCP request and acquire a new, locally valid IP address from the foreign network. However, this introduces a catastrophic failure at the Transport Layer. A Transmission Control Protocol (TCP) connection is not a vague association; it is a rigid state machine bound by a strict 4-tuple: (Source IP, Source Port, Destination IP, Destination Port).

If the mobile node abruptly changes its Source IP address via DHCP while actively downloading a file, the server at the Correspondent Node receives packets originating from an entirely unknown IP address. The server’s TCP stack cannot map this new IP to the existing socket state. Consequently, the server assumes the original connection has timed out or suffered a fatal error, and the active download is permanently severed. Mobile IP was engineered specifically to solve this paradox: allowing a node to change its topological locator without altering its endpoint identifier, thereby preserving TCP semantics across geographic movements.

Define the roles of the Mobile Node (MN), Home Agent (HA), and Foreign Agent (FA) in Mobile IPv4. Describe the difference between a Home Address (HoA) and a Care-of Address (CoA).

The Mobile IP architecture introduces three distinct entities to manage mobility transparently. The Mobile Node (MN) is the roaming device itself, running a specialized Mobile IP daemon capable of detecting subnet changes and initiating registration requests. The Home Agent (HA) is a specialized router permanently stationed on the MN’s home network. Its primary duty is to act as an anchor point; when the MN is away, the HA intercepts incoming packets destined for the MN and redirects them. The Foreign Agent (FA) is a router located on the visited network. It acts as a local proxy, receiving the redirected packets from the HA, unpacking them, and delivering them across the final local link to the MN.

The protocol relies on two distinct IP addresses to decouple identity from location. The Home Address (HoA) is the permanent IP address assigned to the MN by its home network. It serves purely as the endpoint identifier. To the outside world, and to the upper-layer TCP sockets, the MN always appears to reside at this HoA.

The Care-of Address (CoA) is a temporary IP address that represents the MN’s current topological location within the foreign network. In standard Mobile IPv4, this CoA is typically the IP address of the Foreign Agent itself (a FA-CoA). The FA acts as a funnel, sharing its single CoA among multiple visiting mobile nodes. The HA maintains a dynamic mobility binding table that maps the permanent HoA to the temporary CoA, updating this mapping every time the MN registers a new location.

Trace the step-by-step path of an IP packet traveling from a stationary Correspondent Node (CN) to a roaming Mobile Node. Detail how IP-in-IP encapsulation is used to bypass standard Internet routing tables.

When a stationary Correspondent Node (CN) wishes to communicate with a roaming Mobile Node, it is entirely unaware that the MN has moved. The CN constructs a standard IP packet, setting the Destination IP to the MN’s permanent Home Address (HoA). The global internet infrastructure routes this packet blindly toward the MN’s home network, assuming the device is physically present there.

Upon arriving at the home network, the Home Agent (HA) intercepts the packet. It achieves this by issuing Gratuitous ARP replies, falsely claiming that its own MAC address corresponds to the MN’s HoA, effectively hijacking the traffic intended for the absent MN. Once intercepted, the HA consults its mobility binding table to look up the MN’s current Care-of Address (CoA).

Because standard internet routers would drop a packet destined for the HoA if it were routed toward the foreign network, the HA must bypass the global routing tables using IP-in-IP encapsulation (Tunneling). The HA takes the entire original packet (including the original IP header with the HoA) and encapsulates it inside a brand new, outer IP header. This outer header has a Source IP set to the HA’s address and a Destination IP set to the CoA.

The global internet routers examine only this outer header. Seeing a valid, topological destination (the CoA), they route the tunneled packet efficiently to the Foreign Agent. Upon receiving the packet, the FA decapsulates it, stripping away the outer IP header. The FA then inspects the inner packet, sees the destination is the HoA, and forwards the raw packet over the local wireless link to the MN. The MN receives a packet that appears exactly as the CN originally sent it.

Explain the Triangular Routing problem. How does Route Optimization solve this issue by utilizing Binding Updates? Why is Route Optimization native to Mobile IPv6 but an optional, complex extension in IPv4?

Triangular Routing is a massive architectural inefficiency inherent in base Mobile IP. Consider a scenario where a user whose Home Agent is located in New York travels to Tokyo. While in Tokyo, the user accesses a local Japanese web server (the CN). The web server sends packets destined for the user’s HoA. Standard routing forces these packets to travel from Tokyo all the way across the Pacific to the HA in New York. The HA intercepts them, encapsulates them, and tunnels them all the way back across the Pacific to the FA in Tokyo. This triangular path introduces hundreds of milliseconds of unnecessary latency, severely degrades throughput, and places an immense, useless load on trans-oceanic fiber optic cables.

Route Optimization solves this by establishing a direct, shortest-path tunnel between the CN and the MN. When the HA receives the first intercepted packet from the CN, it realizes the CN is unaware of the MN’s true location. The HA sends a specialized “Binding Update” control message directly to the CN. This message explicitly informs the CN of the MN’s current Care-of Address. The CN maintains a local binding cache. For all subsequent packets, the CN bypasses the HA entirely. The CN itself performs the IP-in-IP encapsulation, tunneling the packets directly to the CoA in Tokyo, effectively collapsing the triangle into a straight line.

In Mobile IPv4, Route Optimization is a highly complex, optional extension. The fatal flaw of IPv4 is its lack of native cryptographic security. If an HA sends a Binding Update to a CN, the CN has no built-in way to verify that the update is legitimate and not a malicious hijack attempt by a hacker trying to steal the MN’s traffic. Implementing secure Binding Updates in IPv4 requires massive, cumbersome cryptographic patching. In contrast, Mobile IPv6 mandates the inclusion of IPsec at the network layer. Because every IPv6 node possesses native cryptographic authentication headers, Binding Updates can be inherently trusted and verified. Consequently, Route Optimization is a core, default feature of Mobile IPv6 rather than a bolted-on afterthought.

Discuss the major architectural simplifications introduced in Mobile IPv6. Why is the Foreign Agent (FA) entity completely obsolete in IPv6 networks?

Mobile IPv6 fundamentally redesigns and streamlines mobility management by leveraging the vast address space and inherent intelligence of the IPv6 protocol. The most significant architectural change is the complete eradication of the Foreign Agent entity.

In IPv4, the FA was a necessary evil. Because IPv4 addresses are globally scarce, it was highly inefficient to assign a unique, globally routable Care-of Address to every single visiting mobile phone. The FA solved this by providing a single, shared CoA for all visitors, acting as a massive NAT-like funnel.

IPv6, boasting a practically infinite 128-bit address space, faces no such scarcity. Furthermore, IPv6 introduces mandatory Stateless Address Autoconfiguration (SLAAC). When an IPv6 Mobile Node arrives in a foreign network, it simply listens for Router Advertisements containing the local network prefix (the first 64 bits). The MN then takes its own unique MAC address, applies the EUI-64 algorithm to generate the host identifier (the last 64 bits), and mathematically combines them.

The MN instantly and autonomously generates its own, globally routable Co-Located Care-of Address (CCoA). Because the MN acts as its own tunnel endpoint and can decapsulate its own packets, there is no longer any need for a middleman router to unpack the traffic. The FA is entirely obsolete, drastically simplifying the infrastructure required to deploy cellular networks, as operators no longer need to install specialized FA hardware at every access point.

Section B: Micro-Mobility Protocols

Explain the fundamental need for micro-mobility protocols in dense cellular networks. How do they prevent the global Internet from being flooded with local signaling traffic?

Standard Mobile IP was engineered as a “macro-mobility” protocol, designed for infrequent, large-scale movements (e.g., disconnecting a laptop in an office and reconnecting it hours later at a hotel). When applied to a modern, dense urban cellular environment, its performance collapses entirely.

In a dense city, microcells may only cover a few hundred meters. A user moving in a vehicle will cross subnet boundaries every few seconds. If standard Mobile IP were used, every single boundary crossing would force the Mobile Node to acquire a new Care-of Address and transmit a new Binding Update all the way back across the global internet to its Home Agent. This introduces unacceptable latency—by the time the HA receives the update and redirects the tunnel, the user may have already driven into the next cell, resulting in massive packet loss and dropped connections. Furthermore, millions of cars transmitting global Binding Updates every five seconds would effectively launch a DDoS attack against the core internet infrastructure.

Micro-mobility protocols solve this by creating localized, autonomous routing domains. They establish a hierarchical boundary between local radio movement and global internet routing. When a user crosses cells within a specific city or region, the micro-mobility protocol handles the handoff internally, seamlessly altering local routing tables to track the user. Crucially, the global Care-of Address remains completely static. Because the global CoA doesn’t change, the Home Agent is entirely unaware that the user is moving locally, and zero signaling traffic escapes onto the global internet. The global internet is effectively shielded from the chaotic, high-frequency handoffs occurring at the radio edge.

Describe the operation of Cellular IP. Explain the distinct functions of the Routing Cache and the Paging Cache. How does Cellular IP manage a semi-soft handoff?

Cellular IP operates on a radical premise: within the localized micro-mobility domain, it discards standard IP routing entirely. Instead of routing based on IP prefixes, the domain acts like a massive, intelligent Layer 2 switch, tracking mobile nodes based purely on their MAC addresses.

The protocol manages mobility using two highly specialized databases. The Routing Cache is utilized exclusively for active transmission. When a mobile node transmits an uplink data packet, every intermediate Cellular IP node inspects the source MAC address and records the interface on which the packet arrived. This creates a temporary, reverse-path breadcrumb trail. When downlink packets arrive from the internet, the nodes simply follow this cached breadcrumb trail backward, ensuring the packet perfectly tracks the mobile node’s physical path.

The Paging Cache is utilized for idle nodes. Because active routing caches expire rapidly, idle nodes send periodic, lightweight “paging update” packets. These updates build a generalized, less granular path through the network. If a downlink packet arrives for an idle node, the network utilizes the paging cache to broadcast a wake-up signal only to the specific geographic zone where the node was last seen, conserving immense battery power compared to network-wide broadcasting.

Cellular IP excels in executing Semi-Soft Handoffs. When an MN approaches a cell boundary, it begins sending route updates to the new Base Station before breaking the physical radio link with the old Base Station. The Cellular IP domain dynamically bifurcates the downlink traffic at the crossover point, sending identical streams of data to both the old and new Base Stations simultaneously. During the critical milliseconds while the radio physically retunes, data is buffered at the new BS. This guarantees zero packet loss during the radio switch, providing a flawless transition for real-time traffic.

Contrast the HAWAII protocol with Cellular IP. How does HAWAII manipulate standard IP routing tables dynamically at the Crossover Router to inject host-specific (/32) routes?

While Cellular IP relies on proprietary MAC-layer tracking, fundamentally altering the nature of the network, HAWAII (Handoff-Aware Wireless Access Internet Infrastructure) takes a more standardized approach. HAWAII utilizes standard, unmodified IP routing protocols within the domain.

When a Mobile Node moves from its old Base Station to a new Base Station, it sends a path setup message. This message travels upward through the network hierarchy until it intersects with a router that sits on the path to the old Base Station. This critical intersection point is called the Crossover Router.

Instead of tunneling, HAWAII relies on dynamic routing injection. The setup message forces the Crossover Router to inject a highly specific, host-specific route (/32 subnet mask) into its standard IP routing table. Because standard IP routing always prefers the most specific match, any packets arriving at the Crossover Router destined for the MN’s specific IP address will bypass the general subnet rules. They will hit the newly injected /32 route and be instantly deflected down the new physical path toward the new Base Station. This approach achieves micro-mobility seamlessly without requiring complex tunneling encapsulation or abandoning IP routing principles.

Hierarchical Mobile IPv6 (HMIPv6) introduces a sophisticated proxy tier into the standard IPv6 mobility framework to drastically reduce signaling overhead. The cornerstone of this architecture is the Mobility Anchor Point (MAP). The MAP is a specialized router positioned at the boundary between a localized regional network (like a city) and the global internet.

The architecture relies on the MN maintaining two simultaneous, distinct IP addresses. The Regional Care-of Address (RCoA) is derived directly from the MAP’s network prefix. The MN registers this RCoA globally with its Home Agent and all active Correspondent Nodes. To the outside world, the MN appears to be permanently parked at the MAP. The MAP intercepts all traffic destined for the RCoA.

The On-Link Care-of Address (LCoA) is the true, highly localized IP address derived from the specific Access Router (cell tower) the MN is currently communicating with. The MN registers this LCoA only with the local MAP.

When the MN moves physically from one Access Router to another within the city, its LCoA changes. The MN sends a rapid, localized Binding Update to the MAP, instructing the MAP to tunnel the RCoA traffic to the new LCoA. Crucially, because the global RCoA remains entirely unchanged, the Home Agent and Correspondent Nodes are completely oblivious to the movement. Zero signaling traffic escapes the regional domain, achieving highly scalable micro-mobility while leveraging native IPv6 security architectures.

Section C: Ad-Hoc Networks and Routing

A Mobile Ad-Hoc Network (MANET) is a radical departure from traditional networking. Its defining characteristics include complete decentralization and a total lack of fixed infrastructure (no base stations, no centralized routers). The topology is hyper-dynamic, as nodes move randomly, causing radio links to break and form continuously. The network relies entirely on multi-hop routing, meaning every individual consumer device must simultaneously act as a router, forwarding traffic on behalf of other devices to bridge gaps in coverage. Furthermore, these networks are heavily constrained by battery power, limited computational resources, and erratic wireless bandwidth.

Deploying traditional Link-State protocols, such as OSPF (Open Shortest Path First), in a MANET results in rapid, catastrophic failure. Link-state protocols operate on the principle that every single node must possess an identical, complete map of the entire network topology to calculate shortest paths via Dijkstra’s algorithm.

To maintain this identical map, any time a link breaks or forms, the affected nodes must broadcast a Link State Advertisement (LSA) that floods across the entire network. In a static enterprise network, this happens rarely. In a highly mobile MANET, where nodes are constantly moving in and out of range, links are breaking thousands of times per minute. The network would respond by generating a continuous, massive tsunami of LSA flood packets. This signaling storm would instantly consume 100% of the available wireless bandwidth and drain the batteries of every node just to process routing updates, leaving absolutely zero capacity for actual user data to be transmitted.

Compare Vehicular Ad-Hoc Networks (VANETs) with MANETs across mobility patterns, power constraints, and computational capabilities. Why did IEEE 802.11p remove the lengthy BSS authentication phase for vehicular environments?

While VANETs are a subclass of MANETs, their physical characteristics differ wildly, demanding entirely different protocol designs.

Regarding mobility patterns, MANET nodes (like humans carrying phones) move randomly in a two-dimensional plane at low velocities. VANET nodes (vehicles) possess highly predictable, one-dimensional mobility—they are strictly confined to the physical layout of the road network. However, their relative velocities are extreme, often exceeding 100 km/h on highways.

Regarding constraints, MANET nodes are severely limited by battery life and processing power. VANET nodes possess virtually unlimited power (supplied by the vehicle’s alternator) and can house massive computational processors and high-gain antenna arrays without concern for weight or energy consumption.

The IEEE 802.11p standard (WAVE - Wireless Access in Vehicular Environments) was specifically engineered to address the extreme velocity of VANETs. In a standard Wi-Fi (802.11) network, before data can be transmitted, a device must scan for beacons, perform a multi-step association request, and execute a lengthy cryptographic authentication handshake. If two vehicles approaching each other at a combined relative speed of 200 km/h attempt this, they will remain in effective radio range for less than three seconds. The standard Wi-Fi authentication phase would consume the entire contact window, and the vehicles would pass out of range before a single byte of data was exchanged. 802.11p solves this by completely excising the BSS authentication phase. It operates in “Outside the Context of a BSS” (OCB) mode, allowing vehicles to instantly broadcast critical, life-saving collision avoidance telemetry the millisecond they detect another radio, prioritizing speed over cryptographic overhead.

Compare the philosophies of Proactive (Table-Driven) and Reactive (On-Demand) routing protocols.

Proactive (Table-Driven) routing protocols operate on the philosophy of continuous readiness. Every node continuously shares routing updates in the background, aggressively attempting to maintain a comprehensive, up-to-date routing table for every possible destination in the network, regardless of whether traffic is actually flowing to that destination. The primary benefit is zero latency; when an application needs to send a packet, the route is already calculated and instantly available. The severe drawback is the massive background signaling overhead. In highly dynamic networks, the protocol expends massive energy constantly recalculating routes to nodes that may never be communicated with.

Reactive (On-Demand) routing protocols operate on the philosophy of extreme conservation. Nodes maintain absolutely no background routing tables. A route is only calculated the exact moment an application explicitly generates a data packet intended for an unknown destination. The node pauses the data, initiates a network-wide Route Discovery flood, waits for the reply, and then caches the specific path. The primary benefit is phenomenal efficiency; if no data is being sent, the routing overhead is literally zero. The drawback is high initial latency; the very first packet in a session suffers a massive delay while the route discovery process executes, which can cause real-time applications to stutter.

Explain the Destination-Sequenced Distance-Vector (DSDV) protocol. How do sequence numbers generated by the destination node definitively solve the “Counting to Infinity” problem inherent in classic Bellman-Ford?

DSDV is a proactive routing protocol designed specifically to fix the fatal flaws of the classic Bellman-Ford distance-vector algorithm when applied to highly dynamic ad-hoc networks.

The classic Bellman-Ford algorithm suffers catastrophically from the “Counting to Infinity” routing loop. If a link between Node A and Node B physically breaks, but Node C still advertises a path to B, Node A might mistakenly route packets to C, which routes them back to A. The nodes slowly increment their hop-count metrics back and forth in an infinite loop, congesting the network with trapped packets until the metric reaches an arbitrary “infinity” limit.

DSDV eradicates this loop mathematically by forcing a strict temporal ordering on all routing updates. It requires the destination node itself to generate a monotonically increasing Sequence Number, which is permanently attached to every route advertisement.

When any node receives a new routing update, it obeys a rigid law: it must select the route possessing the newest (highest) Sequence Number, entirely disregarding the hop count. A route with a higher hop count but a newer sequence number is always preferred over a short route with a stale sequence number. If a link physically breaks, the detecting node instantly advertises the broken route with an “infinity” metric and deliberately increments the Sequence Number to an odd value. Because this odd sequence number is mathematically newer than any existing cached route in the network, it instantly overrides and invalidates the stale paths globally. This instantaneous invalidation prevents nodes from latching onto phantom routes, definitively solving the Counting to Infinity problem.

Contrast Ad-hoc On-Demand Distance Vector (AODV) with Dynamic Source Routing (DSR). Why is AODV generally preferred over DSR as the network scales in size?

Both AODV and DSR are reactive protocols, but they differ fundamentally in how they handle path data during the data transmission phase.

DSR relies heavily on Source Routing. During the route discovery phase, the Route Request packet accumulates a massive list of every single IP address it passes through. When the Route Reply returns to the sender, the sender learns the entire, explicit topological path. Crucially, the sender must append this entire list of node IDs to the IP header of every single data packet it transmits. The intermediate routers simply look at the list in the packet header to determine where to forward it next.

AODV discards Source Routing in favor of hop-by-hop tracking. During the route discovery phase, as the Route Request propagates, intermediate nodes dynamically build temporary “forward” and “reverse” pointers in their local RAM. When data transmission begins, the data packet contains only the final Destination IP. The intermediate nodes consult their local RAM pointers to forward the packet to the next hop.

As a MANET scales in physical size and node density, path lengths increase. If a destination is 25 hops away, DSR must append 25 distinct IP addresses to the header of every single data packet. This creates massive, compounding byte overhead, severely degrading the actual data throughput on constrained wireless links. AODV, conversely, maintains a constant, tiny IP header size regardless of how many hops the packet must traverse, making it vastly more efficient and scalable in large networks.

Explain the operation of the On-Demand Multicast Routing Protocol (ODMRP). Why is building a “Forwarding Mesh” vastly superior to building a traditional Multicast Tree in a highly mobile environment?

ODMRP is a highly specialized reactive protocol engineered for multicast communications (sending one data stream to multiple specific receivers simultaneously) in chaotic environments.

Instead of relying on complex tree-building algorithms, ODMRP utilizes a simpler, more robust mechanism. When a source wishes to transmit, it periodically floods a “Join Query” across the network. Nodes that wish to receive the multicast respond with a “Join Reply”. As the Join Replies travel back to the source, every intermediate node that forwards the reply mathematically flags itself as a member of the “Forwarding Group.”

This Forwarding Group forms an overlapping, highly redundant Forwarding Mesh, rather than a rigid, singular Multicast Tree.

The superiority of the mesh approach becomes glaringly obvious in high-mobility MANETs. A traditional Multicast Tree is exceedingly fragile. It relies on a single, critical path to deliver data to a downstream branch. If a single branch-node moves out of radio range, the entire downstream sub-tree is violently disconnected. Data is lost until the protocol detects the break and executes a complex tree-rebuilding algorithm.

A Forwarding Mesh, however, provides massive path redundancy. Because multiple nodes in the forwarding group are broadcasting the same multicast packets simultaneously, there are multiple viable paths from the source to any given receiver. If a node physically moves or a link breaks due to fading, the data stream simply propagates through the remaining, overlapping paths in the mesh without requiring any explicit path recalculation. This inherent redundancy ensures extreme reliability and high packet delivery ratios in environments where link failure is the norm rather than the exception.