11/14/25, 11:11 PM Study strategy for networks

This structure gives full marks in a 5-mark question.

Illustrate the optimality principle in routing for 3 marks

Here is a perfect 3-mark answer — short, clear, and scoring.

Optimality Principle in Routing (3 Marks)

Definition (1 Mark)
The optimality principle states that:

If a router R lies on the optimal (shortest) path from router A to router B, then the path from R to B must
also be optimal.

In other words, any subpath of a shortest path is itself a shortest path.

Illustration / Diagram (1 Mark)

css Copy code

A ---- R ---- B
\ /
\-- C ---/

If path A → R → B is the shortest,

then R → B must also be the shortest path from R to B.

This property is used to build shortest-path trees in routing algorithms.

Explanation / Use (1 Mark)

Routing algorithms like Dijkstra’s Shortest Path and Bellman-Ford use this principle to compute
minimal cost paths.
It ensures routers construct a shortest-path tree rooted at the destination, where each subpath is
optimal.

✔ Perfect 3-mark coverage:

Definition + diagram + explanation.

Identifu two methods by which multicast routing is handled by network

layer for 3 marks
Source based routing - 1.5 marks
Core based routing - 1.5 marks

Here is the perfect 3-mark answer, matching your marking scheme exactly.

Multicast Routing Methods in the Network Layer (3 Marks)

[Link] 83/116
11/14/25, 11:11 PM Study strategy for networks

1️⃣ Source-Based Tree Routing (1.5 Marks)

Each source builds its own shortest-path multicast tree to all receivers in the group.
Tree is rooted at the source, with branches leading to every destination.
Forwarding uses a (Source, Group) pair to identify entries in multicast routing tables.
Ensures optimal paths for each sender but requires separate trees for each source, increasing overhead.

2️⃣ Core-Based Tree (CBT) / Shared Tree Routing (1.5 Marks)

A single central router (core or rendezvous point) is chosen for a multicast group.
All senders send their multicast packets toward the core, which then distributes them to group
members.
A single shared tree is maintained, identified by (Group) only (not per-source).
Reduces state information and routing overhead, but paths may not always be optimal.

✔ 3 marks covered:
Source-based routing → 1.5 marks
Core-based routing → 1.5 marks

Let me know if you want diagrams for each (useful for 5–6 mark questions).

State Optimality principle. What is a Sink tree? for 3 marks

Optimality principle -1.5
Sink tree Definition – 1.5

Here is a perfect 3-mark answer following your mark split.

1️⃣ Optimality Principle (1.5 Marks)

The optimality principle states:

If a router R lies on the shortest (optimal) path from router A to router B, then the path from R to B is also a
shortest (optimal) path.

In simple terms, every subpath of an optimal path is itself optimal.

This principle is the foundation of shortest-path routing algorithms.

2️⃣ Sink Tree (1.5 Marks)

A sink tree is a shortest-path tree rooted at a particular destination (sink).
It shows the optimal route from every router in the network to that destination based on the optimality
principle.

All routers are leaves

The destination is the root
Every path in the tree is the shortest path to the sink

Sink trees form the basis for routing tables in distance-vector and link-state routing.

✔ 3 marks covered.

[Link] 84/116
11/14/25, 11:11 PM Study strategy for networks

Explain the Count-to-Infinity problem in distance vector routing

Explanation of Count-to-Infinity problem in distance vector routing
with figure -3

Here is a perfect 3-mark answer with a clear explanation + simple figure.

Count-to-Infinity Problem in Distance Vector Routing (3 Marks)

Explanation (2 Marks)
In distance vector routing, routers periodically exchange their distance tables with neighbors.
When a link goes down, routers may incorrectly believe a destination is still reachable through each other
and keep increasing the metric (distance) gradually.

This leads to the count-to-infinity problem:

Wrong routing information propagates slowly through the network.

Routers keep increasing the hop count to a destination until it reaches infinity.
This causes slow convergence, routing loops, and wasted bandwidth.

It happens because routers trust their neighbors’ vectors without checking if the neighbor is using them as
the next hop.

Figure (1 Mark)

css Copy code

A ------- B ------- C
(1) (1)

Initial correct state:

A ↔ B ↔ C, with all links cost = 1
Distances to C:
C=0
B=1
A=2

Now link B–C breaks:

css Copy code

A ------- B X C

B sets its distance to C = ∞

But B receives A’s old info: “A thinks C = 2”
B now believes C = A + 1 = 3
A later sees B’s new value (3) and updates itself to 4

This continues:

makefile Copy code

A: 5 → 6 → 7 → …
B: 4 → 5 → 6 → …

Routers count upward toward infinity, never realizing C is gone.

Summary (One sentence)

[Link] 85/116
11/14/25, 11:11 PM Study strategy for networks

Count-to-infinity happens when routers in a distance-vector network incorrectly reinforce each other’s
outdated routes, causing hop counts to grow slowly toward infinity.

If you want, I can also add how split-horizon and poison-reverse fix it for a 5-mark version.

Distinguish between routing and forwarding for 3 marks

Here is a clean, perfect 3-mark answer — short, sharp, and distinguishable.

Routing vs Forwarding (3 Marks)

Aspect Routing Forwarding

Definition Process of computing the path packets Process of moving a packet from an incoming
should take across the network. interface to the correct outgoing interface.

Function Builds and updates the routing table using Uses the routing table to lookup destination
algorithms (Dijkstra, Bellman-Ford, etc.). and send packet to the next hop.

Scope Global, involves cooperation between Local, happens inside a single router.
routers across the network.

Time-scale Happens infrequently (only when the Happens continuously for every packet.
network topology changes).

Example Router learning new routes via OSPF, BGP. Router forwarding an IP packet to the next
router.

One-line summary:
Routing = deciding the path; Forwarding = sending the packet along that path.

Describe any two techniques for achieving good Quality of Service for
3 marks

Here is a perfect 3-mark answer — two QoS techniques, each crisp and scoring 1.5 marks.

Techniques for Achieving Good Quality of Service (3 Marks)

1️⃣ Traffic Shaping (Leaky Bucket / Token Bucket) – 1.5 Marks
Controls the flow rate of packets entering the network.
Leaky Bucket: sends packets at a fixed, smooth rate to avoid bursts.
Token Bucket: allows bursts but limits long-term average rate using tokens.
Prevents congestion and ensures predictable delay and throughput.

2️⃣ Resource Reservation (RSVP) – 1.5 Marks

End systems and routers reserve bandwidth and buffer space along a path.

[Link] 86/116
11/14/25, 11:11 PM Study strategy for networks

Guarantees minimum delay, jitter, and bandwidth for real-time flows.

Used in multimedia applications that require stable, guaranteed performance.

✔ 3 marks covered:
Two QoS techniques, each clearly explained and directly relevant.

Differentiate between static and dynamic routing for 3 marks

Here is a perfect 3-mark answer — short, crisp, and clearly distinguishes static vs dynamic routing.

Static vs Dynamic Routing (3 Marks)

Aspect Static Routing Dynamic Routing

Route Creation Routes are manually configured by the Routes are automatically learned and updated
administrator. by routing protocols.

Adaptability Does not adapt to network changes; Adapts automatically to failures, congestion, or
requires manual updates. topology changes.

Overhead No routing overhead; very low CPU and Higher overhead due to exchanging routing
bandwidth usage. updates.

Scalability Not suitable for large networks. Highly scalable; used in medium-to-large
networks.

Examples Manually configured routes. OSPF, RIP, BGP, EIGRP.

One-line summary:
Static routing is fixed and manual; dynamic routing is adaptive and automatic.

Describe any two congestion control methods in virtual circuit subnets

Here is a perfect 3-mark style answer (1.5 marks each for two methods).

Congestion Control Methods in Virtual Circuit Subnets (Any Two)

1️⃣ Admission Control (1.5 Marks)
Before establishing a new virtual circuit, the network checks whether it has enough bandwidth, buffers,
and resources.
If the network is already congested, the connection request is rejected.
Prevents overload by not allowing more virtual circuits than the network can handle.
Ensures QoS and stable performance for established circuits.

2️⃣ Traffic Throttling / Flow Control (1.5 Marks)

The network limits the rate at which a sender may inject packets into the virtual circuit.

[Link] 87/116
11/14/25, 11:11 PM Study strategy for networks

Uses mechanisms like:

Leaky Bucket
Token Bucket
Window-based flow control
Prevents the sender from overwhelming intermediate routers, reducing congestion buildup.

✔ Other Possible Methods (If needed)

Route Fluctuation (Dynamic Routing): Reroute circuits around congested areas.
Resource Reservation: Buffers/bandwidth reserved per circuit.

If you want a 5-mark expanded version, I can give detailed explanations and diagrams.

do traffic shaping and resource reservation also come under it?

Short, brutally honest answer:

YES — Traffic Shaping and Resource Reservation can be used in virtual-circuit networks, but they are NOT
the primary, textbook congestion-control techniques specific to VC subnets.

Let me break it down cleanly:

✅ What definitely belongs to congestion control in Virtual-Circuit

Subnets
These are the classic, textbook-correct answers (from Tanenbaum):
1. Admission Control
2. Flow Control / Rate Control (Windowing)
3. Route Recalculation (Dynamic routing to avoid congested areas)
4. Buffer Allocation Techniques

These are the mechanisms explicitly tied to virtual circuits.

⚠ What about Traffic Shaping and Resource Reservation?

1️⃣ Traffic Shaping (Leaky Bucket / Token Bucket)
YES, it can be applied to VC subnets.
But it is considered a general congestion-prevention / QoS mechanism, not a VC-specific mechanism.
You can absolutely write it in an exam as a congestion-control method — it will score because it’s valid
and widely used.

2️⃣ Resource Reservation (RSVP-like ideas)

YES, it also applies conceptually — virtual circuits reserve resources when they’re set up.
This fits perfectly with "preventing congestion by guaranteeing bandwidth per circuit."
Many textbooks treat resource reservation as part of virtual-circuit congestion control.

⭐ Final Verdict (Exam-friendly wording)

Traffic shaping and resource reservation DO apply to congestion control in virtual-circuit subnets, but
the “core” VC congestion methods remain admission control, flow control, and routing adjustments.

[Link] 88/116
11/14/25, 11:11 PM Study strategy for networks

In an exam, you can safely use Traffic Shaping and Resource Reservation as valid congestion-control
methods under virtual circuits — the teacher will accept them.

If you want, I can craft a 5-mark or 8-mark “Congestion control in VC Subnets” full answer including these
methods.

Describe two major differences between the warning bit method and
the
Random Early Detection (RED) method for 3 marks

Here is the perfect 3-mark comparison, crisp and examiner-friendly.

Warning Bit Method vs Random Early Detection (RED) – 3 Marks

Feature Warning Bit Method Random Early Detection (RED)

1. How congestion is Router sets a “warning bit” in packet headers Router drops packets randomly early (before queue
signaled when its queue is almost full. Sender detects this is full). Packet loss itself acts as a congestion signal to
bit and slows down. senders.

2. When action is taken Action occurs late, only when congestion is close Action occurs early, before congestion occurs,
to happening (queue nearly full). preventing queue overflow.

3. Drop behavior No packet is dropped unless the queue overflows; Packets are dropped probabilistically based on
it only gives a warning. average queue size to avoid global synchronization.

Summary (1 sentence):
Warning Bit reacts late using explicit congestion notification; RED reacts early using random packet
drops.

Perfect for 3 marks.

Classify routing algorithms based on the information provided for 3

marks

Here is a perfect 3-mark answer—short, crisp, and exactly what examiners expect.

Classification of Routing Algorithms (3 Marks)

Routing algorithms can be classified based on the information available and how decisions are made:

1️⃣ Static (Non-Adaptive) Routing Algorithms (1 Mark)

Routes are precomputed and do not change based on network conditions.
No real-time feedback from the network.
Examples: Fixed routing tables, shortest path calculated once.

[Link] 89/116
11/14/25, 11:11 PM Study strategy for networks

2️⃣ Dynamic (Adaptive) Routing Algorithms (1 Mark)

Routes change automatically based on current network information such as:
traffic load
topology changes
congestion
Examples: Distance Vector and Link State routing.

3️⃣ Centralized vs Distributed Routing (1 Mark)

Centralized:
One central node gathers global network information and computes all routes.

Distributed:
Each router computes routes using local information and exchanges updates with neighbors.
Used by real protocols (RIP, OSPF).

✔ 3 marks covered:
1. Static routing
2. Dynamic routing
3. Centralized vs distributed routing

Let me know if you want a diagram or examples added.

What is flooding? Write any three applications of flooding for 3 marks

Here is a perfect 3-mark answer—short, clean, and exactly what examiners expect.

What is Flooding? (1.5 Marks)

Flooding is a routing technique in which every incoming packet is sent out on all outgoing links (except the
one it arrived on).

No routing table is needed.

Guarantees that the packet eventually reaches all reachable destinations.
Requires mechanisms like hop count or sequence numbers to prevent infinite looping.

Applications of Flooding (Any Three) (1.5 Marks)

1. Broadcast Routing — Ensures a packet reaches all nodes in the network.
2. Military or Emergency Networks — Very robust; works even when nodes fail.
3. Discovering Routes (e.g., in AODV, DSR) — Route Request (RREQ) packets are flooded.
4. Packet Delivery in Highly Unstable Networks — For networks with rapidly changing topology.
5. Reliable Delivery of Critical Data — Ensures delivery even if some paths fail.

Pick any 3 of the above.

✔ Total = 3 marks:
Definition of flooding
Three applications

[Link] 90/116
11/14/25, 11:11 PM Study strategy for networks

If you want, I can also give the advantages/disadvantages for a 5-mark question.

Compare the features of link state routing with distance vector routing
for 8 marks
Link State Routing any 4 points with explanation- 4 Marks
Distance Vector Routing any 4 points with explanation - 4 marks

Here is a perfect 8-mark answer, exactly matching your required structure:

4 explained points for Link State (4 marks) and 4 explained points for Distance Vector (4 marks).

1️⃣ Link State Routing (4 Marks – 4 points with explanation)

1. Complete Topology Knowledge
Each router builds a complete map of the network by collecting link-state advertisements (LSAs) from all
routers.
It knows every router, every link, and every cost.

2. Uses Dijkstra’s Algorithm

After building the topology database, each router runs Dijkstra’s shortest path algorithm to compute the
best path to each destination.
Results in fast and accurate route calculation.

3. Fast Convergence
Routers flood LSAs immediately when a link changes, enabling the entire network to update quickly.
This prevents loops and instability.

4. No Count-to-Infinity Problem
Because routers know the full network map, link-state routing does not suffer from routing loops or count-
to-infinity issues like distance vector routing does.

2️⃣ Distance Vector Routing (4 Marks – 4 points with explanation)

1. Partial Knowledge of Network
Routers know only the distance (cost) and next hop to each destination.
They do not know the full network topology, only what neighbors tell them.

2. Uses Bellman–Ford Algorithm

Routers update their tables by exchanging distance vectors with neighbors and applying the Bellman–Ford
equation:

New Distance = min(neighbor distance + cost)

3. Slow Convergence
Updates happen periodically or when tables change.
This leads to slow convergence, especially after failures, and can cause temporary routing loops.

4. Suffers From Count-to-Infinity Problem

Routers may incorrectly increase distances slowly toward infinity when a link fails because they rely on
neighbors’ tables, creating routing loops.

✔ 8-Mark Coverage Summary

[Link] 91/116
11/14/25, 11:11 PM Study strategy for networks

Link State Routing (4 marks):

1. Complete topology knowledge
2. Shortest path via Dijkstra
3. Fast convergence
4. No count-to-infinity

Distance Vector Routing (4 marks):

1. Only distance & next-hop info
2. Bellman–Ford algorithm
3. Slow convergence
4. Count-to-infinity problem

If you want, I can also add a comparison table or a diagram-based version for 10 marks.

Differentiate between the implementation of datagram subnet and

virtual circuit 6
subnet for 6 marks
Datagram subnet - 3 marks
Virtual circuit subnet – 3 marks

Here is a perfect 6-mark answer, exactly matching your required structure:

Datagram Subnet – 3 marks
Virtual Circuit Subnet – 3 marks

1️⃣ Datagram Subnet (3 Marks)

1. Connectionless Communication
No setup phase.
Each packet is treated independently, like postal mail.

2. Routing per Packet

Each packet may take different routes depending on network conditions.
Router uses destination address in every packet for forwarding.

3. No State Information in Routers

Routers do NOT maintain per-connection tables.
Routing tables store only next-hop info, making the design simple.

4. Unreliable but Highly Flexible

Packets may arrive out of order, duplicated, or lost.
Used in IP networks (Internet model).

2️⃣ Virtual Circuit Subnet (3 Marks)

1. Connection-Oriented Communication
Requires a setup phase before data transfer → establishes a “virtual path” in routers.
All packets follow the same pre-established route.

2. Routing Once, Not Per Packet

Path chosen during setup; packets use a VC number, not the full destination address.

[Link] 92/116
11/14/25, 11:11 PM Study strategy for networks

Routers forward packets based on VC tables.

3. Routers Maintain State

Each router keeps per-connection entries (VC number, incoming & outgoing port).
Provides ordered, reliable delivery.

4. More Reliable & Reserved Resources

Easy to guarantee QoS since path is predetermined.
Used in networks like ATM, Frame Relay, and older telecommunication systems.

✔ 6-Mark Summary
Datagram Subnet
Connectionless
Independent packet routing
No state stored in routers
Used in IP networks

Virtual Circuit Subnet

Connection-oriented
Single pre-defined route
Routers store per-connection state
Supports QoS and reliable delivery

If you want, I can also provide a comparison table for revision.

What is count to infinity problem? Describe two techniques to solve the

count to
infinity problem in distance vector routing algorithm for 6 marks
Count to infinity problem- 3 marks
Solution to Count to infinity problem 3 marks

Here is a perfect 6-mark answer with the exact mark split you asked for.

1️⃣ Count-to-Infinity Problem (3 Marks)

In distance vector routing, routers exchange routing tables with neighbors.
Each router believes whatever distances its neighbors advertise.

When a route becomes invalid (e.g., a link fails), routers may incorrectly tell each other that the destination
is still reachable—through each other.
They keep increasing the metric (hop count) one step at a time:

Example:

mathematica Copy code

A — B — C

If C fails, B sets distance to C = ∞.

But A still thinks C = 2 and tells B.
B updates to 3, A updates to 4 … and so on.

The hop count slowly increases toward infinity, causing:

Long convergence time

[Link] 93/116
11/14/25, 11:11 PM Study strategy for networks

Temporary routing loops

Unstable routing tables

This is called the count-to-infinity problem.

2️⃣ Techniques to Solve Count-to-Infinity (3 Marks)

A. Split Horizon (1.5 Marks)
A router does not advertise a route back to the neighbor from which it learned that route.
Prevents loops by ensuring bad information isn’t sent back in the reverse direction.

Example:
If B learned route to C from A,
→ B will NOT tell A that it can reach C.
Thus incorrect routing info cannot circulate endlessly.

B. Poison Reverse (1.5 Marks)

An extension of split horizon.
Router advertises the route back to the neighbor, but with a metric of infinity.
Explicitly tells the neighbor:
“Do NOT use me to reach that destination.”

This aggressively breaks routing loops and ensures faster convergence.

✔ 6-Mark Summary
Count-to-infinity = routers repeatedly increase hop count due to wrong info.
Split Horizon and Poison Reverse are the two main solutions to prevent route looping and limit the
increase to infinity.

If you want, I can add a small diagram for a 7–8 mark version.

Identifu and describe the scheduling techniques to improve the

Quality of
Services (QoS). For 8 marks
Scheduling techniques to improve the Quality of Services(QoS) - 2
marks
1. FIFO queuing,
2. Priority queuing,
3. Weighted fair queuing....
Describing each - 2 marks

Here is a perfect 8-mark answer with clear identification + descriptions of the required scheduling
techniques.

Scheduling Techniques to Improve QoS (8 Marks)

Scheduling determines which packet is transmitted next when multiple flows compete for the same output
link.
Good scheduling ensures low delay, fairness, and predictable service — essential for Quality of Service (QoS).

1️⃣ FIFO (First-In First-Out) Queuing – 2 Marks

[Link] 94/116
11/14/25, 11:11 PM Study strategy for networks

Concept:
Packets are served in the exact order they arrive.
Simple, no prioritization, one queue per interface.

Characteristics / Impact on QoS:

Easy to implement but provides no QoS guarantees.
Delay becomes high when queue fills.
Suitable for simple best-effort traffic but not ideal for real-time flows.

2️⃣ Priority Queuing – 2 Marks

Concept:
Packets are divided into multiple priority classes (High, Medium, Low).
Router always serves the highest-priority queue first.
Lower-priority traffic is transmitted only when higher-priority queues are empty.

Characteristics / Impact on QoS:

Guarantees low delay for real-time or mission-critical traffic.
Danger of starvation for lower-priority traffic if high-priority load is heavy.
Used in VoIP, video conferencing, and control signals.

3️⃣ Weighted Fair Queuing (WFQ) – 2 Marks

Concept:
Allocates bandwidth fairly among flows but assigns weights so some flows get more bandwidth.
Each flow has its own queue; scheduler serves them in proportion to their weights.

Characteristics / Impact on QoS:

Ensures fair sharing while still giving priority to important flows.
Prevents starvation and controls delay and jitter.
Suitable for mixed traffic: real-time, data, and background flows.

✔ 8-Mark Summary
Technique Purpose QoS Benefit

FIFO Simple queue No QoS guarantee, high delay under load

Priority Queuing Serve high-priority first Low delay for critical traffic, but risk of starvation

WFQ Fair bandwidth distribution with Balanced fairness + controllable QoS for each flow
weights

If you want, I can also add diagrams or expand this into a 10-mark detailed version.

[Link] 95/116
11/14/25, 11:11 PM Study strategy for networks

Consider the following subnet.

Distance vector routing is used, and the following vectors have just
come in to
router C: from B: (5, 0, 8, 12, 6,2); from D: (16, 12, 6, 0, 9, 10); and from
E:
(7,6,3,9,0,4).
The measured delays to B, D, and E, are 6,3, and 5, respectively. What is
C's
new routing table? Give both the outgoing line to use and the
expected delay.

Thought for 24s

Good — compute shortest next-hop by adding C→neighbor delay to that neighbor’s advertised vector, then
pick the minimum for each destination.

Neighbors’ vectors (destinations A,B,C,D,E,F):

B: (5, 0, 8, 12, 6, 2)
D: (16,12, 6, 0, 9,10)
E: (7, 6, 3, 9, 0, 4)

Measured delays from C: C→B = 6, C→D = 3, C→E = 5.

Compute cost via each neighbor and choose min:

To A: via B = 6+5 = 11; via D = 3+16 = 19; via E = 5+7 = 12 → choose B, cost 11.
To B: via B = 6+0 = 6; via D = 3+12 = 15; via E = 5+6 = 11 → choose B, cost 6.
To C: local (self) = 0.
To D: via B = 6+12 = 18; via D = 3+0 = 3; via E = 5+9 = 14 → choose D, cost 3.
To E: via B = 6+6 = 12; via D = 3+9 = 12; via E = 5+0 = 5 → choose E, cost 5.
To F: via B = 6+2 = 8; via D = 3+10 = 13; via E = 5+4 = 9 → choose B, cost 8.

Final routing table at C

Destination Outgoing line (next hop) Expected delay (cost)

A B 11

B B 6

C — (local) 0

D D 3

E E 5

F B 8

That’s the new routing table for C.

Illustrate the packet routing process of mobile hosts for 6 marks

Explanation of the mobile host’s packet routing process consisting of
four steps -4
Figure - 2

Below is the perfect 6-mark answer with 4-step explanation (4 marks) and a clean diagram (2 marks)—
exactly what examiners expect for “packet routing for mobile hosts.”

[Link] 96/116
11/14/25, 11:11 PM Study strategy for networks

Packet Routing Process for Mobile Hosts (6 Marks)

Mobile IP handles hosts that move from one network to another without changing their IP address. Packet
delivery involves four major steps.

1️⃣ Mobile Host Registers with Foreign Agent (FA) – 1 Mark

When a mobile host moves to a new network, it detects the Foreign Agent’s advertisement.
It sends a registration request to its Home Agent (HA) through the FA.
HA updates its mobility binding (mobile host ↔ care-of address).

2️⃣ Correspondent Sends Packet to Home Address (1 Mark)

A sender (correspondent host) sends packets to the mobile host’s permanent home IP address.
These packets reach the Home Agent, because the home address still points to the home network.

3️⃣ Home Agent Tunnels Packet to Care-of Address (1 Mark)

The Home Agent intercepts packets destined for the mobile host.
It encapsulates (tunnels) them inside a new IP packet and forwards them to the mobile host’s care-of
address (Foreign Agent’s address).

4️⃣ Foreign Agent Delivers Packet to Mobile Host (1 Mark)

The Foreign Agent decapsulates the tunneled packet.
It delivers the original packet to the mobile host over the foreign network.
Reply packets often go directly to the correspondent (route optimization), but basic Mobile IP routes them
back via HA.

Diagram of Packet Routing for Mobile Host (2 Marks)

pgsql Copy code

+-----------------------------+
| Correspondent Host |
+---------------+-------------+
|
| Sends packets to home IP
v
+---------+---------+
| Home Agent |
| (in home network) |
+----+--------+-----+
| ^
| |
Tunnels to | | Registration
care-of addr | |
v |
+-------+--------+------+
| Foreign Agent |
| (in visited network) |
+----------+------------+
|
| Delivers packet
v
+-------+-------+

[Link] 97/116
11/14/25, 11:11 PM Study strategy for networks
| Mobile Host |
+---------------+

✔ 6-Mark Answer Breakdown

Four steps explained clearly → 4 marks
Labeled diagram → 2 marks

If you want, I can also prepare the reverse path routing explanation or the triangle routing problem for 5–6
mark questions.

What is meant by congestion in networks? How does it affect network

performance? Describe any five congestion control techniques for
datagram
networks for 8 marks
Congestion Definition – 2
Congestion’s effect network performance: packet drop and reduced
throughput – 1
Any five congestion control techniques for datagram networks - 5

Here is the perfect 8-mark answer with the exact structure you asked for.

1️⃣ What is Congestion? (2 Marks)

Congestion occurs when the demand for network resources exceeds the available capacity, causing routers
and links to become overloaded.
Too many packets arrive faster than routers can process, store, or forward.

In simple terms:
Too much traffic → router buffers fill → delays and losses increase.

2️⃣ How Congestion Affects Network Performance? (1 Mark)

Packet loss: Routers drop packets when buffers overflow.
Reduced throughput: More retransmissions + long queues = lower effective data transfer rate.
High delay & jitter: Queues grow, causing unpredictable delays.

These effects make the network unstable and degrade application performance.

3️⃣ Congestion Control Techniques for Datagram Networks (Any Five –

5 Marks)
(Datagram networks = connectionless networks like IP, no guaranteed path. control is done dynamically)

1. Packet Dropping / Random Early Detection (RED)

Routers monitor queue length and start dropping packets early before the buffer is full.
This signals senders to slow down and prevents sudden collapse of throughput.

2. Source-Based Congestion Control

Routers notify the source about congestion using techniques like:

[Link] 98/116
11/14/25, 11:11 PM Study strategy for networks

Choke packets
Warning bits

The sender slows down transmission rate to avoid overwhelming routers.

3. Hop-by-Hop Backpressure
A congested router sends a signal to its upstream router, forcing it to reduce its sending rate.
This propagates backward until the source slows down.
Effective in tightly controlled networks with reserved capacity.

4. Traffic Shaping (Leaky Bucket / Token Bucket)

Controls how fast packets enter the network:

Leaky Bucket: smooths traffic by outputting at a fixed rate.

Token Bucket: allows bursts but limits long-term average rate.

Prevents sudden heavy bursts from causing congestion.

5. Load Shedding
Under severe congestion, routers drop low-priority packets first (e.g., using priority or drop policies).
Ensures important traffic gets through even when the network is overloaded.

6. Admission Control (optional extra point)

Before accepting new flows, the network checks whether resources are available.
If not, it rejects the new flow to prevent congestion.

7. Queue Management (Fair Queuing, WFQ) (optional)

Packets are placed in per-flow queues to ensure fairness and prevent any one flow from dominating the
router.

✔ Marks Breakdown (8 Marks)

Congestion definition → 2 marks
Effects on performance → 1 mark
Any five congestion control techniques → 5 marks

If you want, I can also give you a diagram-based version for a 10-mark question.

Explain the steps involved in Multicast routing for 6 marks

Multicast routing steps – group management, spanning tree
construction,
pruning - 3*2 =6

Here is the exact 6-mark answer with the three required steps, each worth 2 marks.

[Link] 99/116
11/14/25, 11:11 PM Study strategy for networks

Multicast Routing – Steps Involved (6 Marks)

Multicast routing delivers a packet from one sender to multiple receivers belonging to a multicast group.
The process involves three major steps:

1️⃣ Group Management (2 Marks)

Hosts join or leave multicast groups using protocols like IGMP (Internet Group Management Protocol).
Routers maintain membership lists for each multicast group on their interfaces.
When a host wants to receive multicast traffic, it sends an IGMP Join message.
When it leaves, it sends an IGMP Leave message.
Routers track which interfaces have active members and forward multicast packets only where needed.

Purpose:
To know which receivers belong to the multicast group and on which interfaces they are located.

2️⃣ Spanning Tree Construction (2 Marks)

Routers construct a multicast distribution tree to deliver packets efficiently without loops.
Two approaches are used:
Source-Based Tree: Each sender builds its own shortest-path tree to all group members.
Core-Based Tree (CBT): A single shared tree rooted at a core router serves all senders.
The tree ensures one copy of the packet travels over each link, preventing duplicates.

Purpose:
To establish an efficient, loop-free path for delivering multicast packets.

3️⃣ Pruning (2 Marks)

Routers remove branches of the multicast tree where no group members exist.
If a router receives a multicast packet but has no downstream members, it sends a prune message
upstream.
This prevents unnecessary forwarding and saves bandwidth.
In dense-mode multicast (e.g., DVMRP), prune messages are temporary and expire after some time.

Purpose:
To eliminate unused branches, improving efficiency and reducing traffic.

✔ 6-Mark Summary
1. Group Management → Routers learn group membership via IGMP.
2. Spanning Tree Construction → Build efficient, loop-free multicast delivery paths.
3. Pruning → Remove branches with no receivers to reduce waste.

This is the exact structure needed to score full marks.

Explain distance vector routing algorithm with an example. for 8 marks

Below is a perfect 8-mark answer: clear explanation + full working example exactly how examiners expect.

Distance Vector Routing Algorithm (8 Marks)

[Link] 100/116
11/14/25, 11:11 PM Study strategy for networks

Distance Vector Routing (DVR) is a dynamic routing algorithm where each router maintains a table (vector)
containing the distance (cost) to every other router and the next hop.
Routers periodically exchange vectors with neighbors and update their tables using the Bellman–Ford
equation:

Dx (y) =
​ min ​ (c(x, v) + Dv (y))
​

v∈neighbors

Where:
Dx (y) = cost from router x to destination y
​

c(x, v) = cost of link from x to neighbor v

Routers gradually learn shortest paths as they continuously share updated distances.

Working Steps (Explanation – 4 Marks)

1. Initialization
Each router knows:
Distance to itself = 0
Distance to direct neighbors = link cost
Distance to all others = ∞

2. Exchange of Distance Vectors

Periodically, each router sends its distance vector to all neighbors.
Neighbors use this information to update their own routing tables.

3. Update Using Bellman–Ford Rule

For each destination, a router checks whether a better path exists through a neighbor.
If yes, update the distance and next hop.

4. Convergence
Routers continue exchanging vectors until no further changes occur (the network converges).
All routers then have shortest-path routes.

Example (4 Marks)
Consider the following network:

css Copy code

A
/ \
1 4
/ \
B---2---C

Link costs:
A–B = 1, A–C = 4, B–C = 2.

Step 1: Initial Distance Vectors

Router A:
Dest Cost Next Hop

A 0 -

B 1 B

[Link] 101/116
11/14/25, 11:11 PM Study strategy for networks

Dest Cost Next Hop

C 4 C

Router B:
Dest Cost Next Hop

B 0 -

A 1 A

C 2 C

Router C:
Dest Cost Next Hop

C 0 -

B 2 B

A 4 A

Step 2: Routers Exchange Vectors

A receives B’s distance vector:
B → C cost = 2
So A checks if going to C via B is cheaper:

A→B →C =1+2=3<4

A updates its entry:

Dest Old Cost New Cost Next Hop

C 4 3 B

Step 3: B Receives A’s Update

B checks if going to C via A is cheaper:

B →A→C =1+3=4>2

No change.

Step 4: C Receives Updates

C checks if going to A via B is cheaper:

C →B →A=2+1=3<4

C updates its table:

Dest Old Cost New Cost Next Hop

A 4 3 B

Final Routing Tables After Convergence

Router A

[Link] 102/116
11/14/25, 11:11 PM Study strategy for networks

Dest Cost Next Hop

A 0 -

B 1 B

C 3 B

Router B
Dest Cost Next Hop

B 0 -

A 1 A

C 2 C

Router C
Dest Cost Next Hop

C 0 -

B 2 B

A 3 B

✔ 8-Mark Answer Summary

Explanation of DVR mechanism → 4 marks
Example with table updates and final convergence → 4 marks

This matches exactly what examiners want for an 8-mark question.

Explain any three closed loop congestion control techniques for 6

marks

Here is a perfect 6-mark answer — three closed-loop congestion control techniques, each explained for 2
marks.

Closed-Loop Congestion Control Techniques (6 Marks)

Closed-loop techniques monitor the network, detect congestion, and take corrective action based on
feedback.

Below are three correct, standard closed-loop techniques, each worth 2 marks.

1️⃣ Choke Packets (2 Marks)

When a router becomes congested, it sends a choke packet back to the sender.
The sender, upon receiving the choke packet, reduces its transmission rate.
This forms a feedback loop:
router detects overload → sends warning → sender slows down → congestion reduces.
Used in datagram networks where congestion appears suddenly.

Key idea: The network explicitly tells the source to slow down.

[Link] 103/116
11/14/25, 11:11 PM Study strategy for networks

2️⃣ Load Shedding (2 Marks)

When congestion becomes severe and buffers are nearly full, routers drop packets intentionally.
Packets may be dropped based on:
Priority
Random selection
Fair queuing
Low-priority or “least important” packets are discarded first to protect critical flows.

Key idea: Router removes packets to relieve pressure and stabilize the network.

3️⃣ Random Early Detection (RED) (2 Marks)

Router monitors average queue length.
Instead of waiting until the buffer is full, it drops packets early, with a probability that increases as
queue grows.
Early drops signal sources to slow down (TCP reacts by reducing window).
RED prevents global synchronization and keeps queues small.

Key idea: Avoid congestion by proactively dropping packets before buffers overflow.

✔ 6-Mark Summary
Technique Closed-loop mechanism

Choke Packets Router detects congestion → sends feedback to source to slow down

Load Shedding Router drops selected packets to reduce overload

RED Router randomly drops packets early to prevent queue buildup

This answer is the ideal structure for scoring full 6 marks.

Explain how routing is performed using link state algorithm. Illustrate

with an
example for 10 marks

Below is a perfect 10-mark answer, structured exactly how a university examiner expects:
Explanation of link-state routing (6 marks) + Worked example with Dijkstra (4 marks).

Link State Routing – Explanation (6 Marks)

Link State Routing is a dynamic routing algorithm where each router builds a complete map of the network
topology and computes shortest paths using Dijkstra’s algorithm.
It operates in the following steps:

1️⃣ Discovering Neighbors (1 Mark)

Each router identifies all directly connected neighbors by sending HELLO packets.
It learns the link cost to every neighbor (delay, bandwidth, or hop count).

[Link] 104/116
11/14/25, 11:11 PM Study strategy for networks

2️⃣ Measuring Link Costs (1 Mark)

The router measures the cost of each link using techniques like:
Round-trip delay
Bandwidth measurement
Static weights

This cost is used in later shortest-path calculations.

3️⃣ Building Link State Packets (LSPs) (1 Mark)

Each router creates an LSP containing:
Its ID (router name)
List of neighbors
Cost to each neighbor
Sequence number
TTL (Time To Live)

This packet describes the router’s local view of the network.

4️⃣ Flooding LSPs to All Routers (1 Mark)

Each router floods its LSP to every other router in the network.
Flooding guarantees all routers receive the same link-state information.
Old or duplicate LSPs are discarded using sequence numbers.

5️⃣ Building the Complete Network Topology (1 Mark)

Using received LSPs, every router constructs an identical directed graph of the network.
This graph contains:
All routers
All links
All link costs

Thus, every router has a global, synchronized topology database.

6️⃣ Running Dijkstra’s Algorithm (Shortest Path First) (1 Mark)

Using the topology graph, each router runs Dijkstra’s algorithm to compute the shortest path to every
destination.
The output is a shortest-path tree (SPT) rooted at the router.
Routing table entries are derived from this SPT.

Worked Example (4 Marks)

Consider the following network:

lua Copy code

(1) (4)
A -------- B -------- C
\ |
\ (2) (3) |
\--------- D --------

Link costs:

A–B = 1
A–D = 2

[Link] 105/116
11/14/25, 11:11 PM Study strategy for networks

B–C = 4
D–C = 3

Let router A compute its routing table.

Step 1: Initialize Dijkstra

Node Cost from A Next Hop

A 0 -

B 1 B

D 2 D

C ∞ -

Visited set = {A}

Step 2: Visit closest node → B

From B, update C:
A→B→C=1+4=5
So: C = 5 via B

Table:

Node Cost Next Hop

A 0 -

B 1 B

D 2 D

C 5 B

Visited set = {A, B}

Step 3: Visit next closest → D

From D, update C:
A→D→C=2+3=5

Same cost as previous path → keep earlier entry.

Visited set = {A, B, D}

Step 4: Visit last node → C

All nodes now visited; Dijkstra is complete.

Final Routing Table at A

Destination Shortest Cost Next Hop

A 0 -

B 1 B

C 5 B

[Link] 106/116
11/14/25, 11:11 PM Study strategy for networks

Destination Shortest Cost Next Hop

D 2 D

✔ 10-Mark Answer Summary

6 marks → Detailed explanation of link-state routing:
1. Neighbor discovery
2. Link cost measurement
3. LSP generation
4. LSP flooding
5. Building topology
6. Dijkstra (SPF algorithm)

4 marks → Worked numerical example with full Dijkstra calculation.

If you want, I can also draw the shortest-path tree or provide a comparison between DVR and LS for exam
revision.

Write notes on load shedding. for 4 marks

Here is a perfect 4-mark answer—clear, crisp, and exactly what examiners expect.

Load Shedding (4 Marks)

Load shedding is a congestion control technique used when a network becomes heavily overloaded and
routers can no longer process or buffer incoming packets. When congestion reaches a critical level, the
router intentionally drops packets to reduce the load and keep the network from collapsing.

1️⃣ Why Load Shedding Is Needed (1 Mark)

When input traffic exceeds router capacity, buffers fill up, delays explode, and throughput collapses.
To prevent complete failure, routers must discard some packets so they can continue forwarding others
efficiently.

2️⃣ How Load Shedding Works (1 Mark)

Routers monitor buffer occupancy.
When the queue crosses a certain threshold, the router begins selectively dropping packets.
If congestion worsens, the router drops packets more aggressively.
It continues until the queue length returns to safe levels.

3️⃣ Packet Selection Strategies (1 Mark)

Routers can drop packets using different policies:

Random drop: randomly discard packets.

Priority-based drop: drop low-priority or background traffic first.
Fair dropping: drop packets proportionally across flows.
Drop-tail: drop only when buffer is full (least effective).

[Link] 107/116
11/14/25, 11:11 PM Study strategy for networks

These strategies ensure critical or real-time packets survive during congestion.

4️⃣ Advantages / Importance (1 Mark)

Prevents total network collapse under heavy load.
Ensures high-priority packets still get through.
Stabilizes queues and reduces delay.
Helps keep throughput at acceptable levels even during overload.

✔ Final 4-Mark Summary

Load shedding = intentional packet dropping during congestion to maintain network stability.
It reduces router overload and protects important traffic using smart discard policies.

If you want, I can also provide examples (e.g., RED, weighted random drop) or diagrams for a 5–6 mark
version.

Explain how shortest path routing is performed. Illustrtatw with an

example for 8 marks

Here is the perfect 8-mark answer with a clear explanation (4 marks) and a full worked example (4 marks).

Shortest Path Routing – Explanation (4 Marks)

Shortest path routing finds the minimum-cost route between a source and every other node in a network.
The most common algorithm used is Dijkstra’s Shortest Path First (SPF).

Steps in Shortest Path Routing:

1️⃣ Represent the network as a weighted graph

Routers → nodes
Links → edges
Link costs → weights (distance, delay, bandwidth cost, etc.)

2️⃣ Initialize distances

Distance to self = 0
Distance to all other nodes = ∞
Mark all nodes as unvisited

3️⃣ Iteratively pick the unvisited node with smallest temporary distance
Add it to the set of permanently visited nodes
Update distances to its neighbors using the rule:

New distance = current distance + edge weight

If this is smaller than the existing recorded distance, update it.

4️⃣ Repeat until all nodes are visited

This produces a shortest path tree (SPT) from the source to every node.
Routing tables are built using the SPT:
destination → next hop → cost.

[Link] 108/116
11/14/25, 11:11 PM Study strategy for networks

Shortest Path Routing Example (4 Marks)

Consider the network:

mathematica Copy code

2
A ------ B
| \ \
5| \1 \3
| \ \
C---- D ---- E
2 4

Link costs:
A–B = 2
A–C = 5
A–D = 1
B–E = 3
D–C = 2
D–E = 4

We compute shortest paths from A.

Step 1: Initialization
Node Distance from A Next hop

A 0 —

B ∞ —

C ∞ —

D ∞ —

E ∞ —

Visited set = { }

Step 2: Visit A (distance 0)

Update neighbors:
B via A = 2
C via A = 5
D via A = 1

Table:

Node Distance Next hop

A 0 —

B 2 B

C 5 C

D 1 D

E ∞ —

Visited set = {A}

[Link] 109/116
11/14/25, 11:11 PM Study strategy for networks

Step 3: Visit D (distance 1)

Update neighbors:
C via D: 1 + 2 = 3 (better than old 5 → update)
E via D: 1 + 4 = 5

Table:

Node Distance Next hop

A 0 —

B 2 B

C 3 D

D 1 D

E 5 D

Visited set = {A, D}

Step 4: Visit B (distance 2)

Update neighbor E:
E via B: 2 + 3 = 5 (same as existing → no change)

Visited set = {A, D, B}

Step 5: Visit C (distance 3)

No better path found.

Visited set = {A, D, B, C}

Step 6: Visit E (distance 5)

Done.

Final Shortest Paths from A

Destination Cost Next Hop

A 0 —

B 2 B

C 3 D

D 1 D

E 5 D

✔ 8-Mark Answer Summary

Explanation (4 marks):
Represent network as graph
Initialize distances
Pick smallest unvisited node
Update neighbors

[Link] 110/116
11/14/25, 11:11 PM Study strategy for networks

Build routing table from shortest-path tree

Example (4 marks):
Step-by-step Dijkstra execution
Final shortest path table

If you want, I can also convert this into a neat diagram-based version for a 10-mark question.

* E via B: 2 + 3 = **5** (same as existing → no change) Visited set = {A, D, B}

why is C via B not taken

You’re asking why the algorithm didn’t update C’s route to go through B
— in other words:

"Why isn't C via B taken instead of C via D?"

Here’s the brutally clear answer:

✅ Because the path A → B → C is not cheaper than A → D → C.

Let’s compute both paths.

Option 1: C via D
A→D→C

Cost =

A→D = 1
D→C = 2

Total = 1 + 2 = 3

This is the path selected.

Option 2: C via B
A→B→C

But: There is no direct B–C link in the topology you gave.

Look at the network again:

mathematica Copy code

2
A ------ B
| \ \
5| \1 \3
| \ \
C---- D ---- E
2 4

The C link goes only to A and D, not to B.

So if you try to reach C via B, you’d need:

[Link] 111/116
11/14/25, 11:11 PM Study strategy for networks
A→B→A→D→C

Compute this insane backward route:

A→B = 2
B→A = 2
A→D = 1
D→C = 2

Total = 7

That is way worse than 3.

🔥 Key Point — Link B–C does not exist

So the algorithm never even considers a B→C neighbor update, because:

B’s neighbors = {A, E}

C is not directly reachable from B

Dijkstra only updates distances through DIRECT neighbors, so:

A can update B, C, D because they are neighbors.

B can update A, E only.
C can update A, D only.

✔ Final Answer (Short):

C via B is not taken because B is not a direct neighbor of C.
The only valid paths to C from A are:

A → D → C = cost 3 (chosen)
A → C = cost 5 (not chosen)

Any path involving B first is longer or loops back, so it is discarded.

If you want, I can draw a shortest-path tree to make it even clearer.

Explain the techniques for achieving good QoS for 6 marks

Here is a perfect 6-mark answer with three QoS techniques, each worth 2 marks as typically expected.

Techniques for Achieving Good Quality of Service (QoS) – 6 Marks

Quality of Service (QoS) ensures predictable performance for applications such as voice, video, and real-time
data.
Common QoS techniques include traffic shaping, scheduling, and resource reservation.

1️⃣ Traffic Shaping (Leaky Bucket / Token Bucket) – 2 Marks

Traffic shaping controls the rate at which packets enter the network, preventing sudden bursts that cause
congestion.

Leaky Bucket
Packets enter a bucket and leak out at a fixed, steady rate.
Smoothens bursty traffic.

[Link] 112/116
11/14/25, 11:11 PM Study strategy for networks

Prevents congestion buildup.

Token Bucket
Tokens accumulate at a fixed rate.
A packet can be sent only if a token is available.
Allows controlled bursts while keeping long-term rate stable.

Effect: Reduces delay and jitter; protects network from overload.

2️⃣ Scheduling Algorithms (FIFO, Priority Queuing, WFQ) – 2 Marks

Scheduling decides which packet is sent next when multiple flows compete for a link.

FIFO (First-In-First-Out)
Simple; no QoS guarantees.
Packets served in arrival order.

Priority Queuing
Traffic categorized into priority classes.
High-priority packets (voice/video) transmitted first.
Low-priority packets may wait longer.

Weighted Fair Queuing (WFQ)

Each flow gets a fair share of bandwidth.
Weights ensure more important flows get proportionally more bandwidth.
Reduces delay and jitter for real-time traffic.

Effect: Ensures fairness, prevents starvation, and supports real-time data.

3️⃣ Resource Reservation (RSVP / Admission Control) – 2 Marks

Ensures that applications get guaranteed bandwidth and delay levels.

RSVP (Resource Reservation Protocol)

Sender and receiver reserve bandwidth, buffers, and router resources along a path.
Provides guaranteed QoS for voice/video streams.

Admission Control
Before accepting new flows, routers check whether necessary resources exist.
If not, the new flow is blocked to protect existing traffic.

Effect: Prevents congestion and maintains predictable performance for ongoing sessions.

✔ 6-Mark Summary
Technique Purpose QoS Benefit

Traffic Shaping Control input rate Smooth traffic, reduce jitter & congestion

Scheduling Decide packet transmission order Fairness, priority handling, real-time support

Resource Reservation Reserve bandwidth/resources Guaranteed delay & throughput

If you want, I can also provide an 8-mark expanded version or diagram-based explanation.

[Link] 113/116
11/14/25, 11:11 PM Study strategy for networks

Explain the characteristics of Routing Information Protocol (RIP) for 6

marks

Here is a perfect 6-mark answer — six strong, well-explained characteristics of RIP.

Characteristics of Routing Information Protocol (RIP) – 6 Marks

1️⃣ Distance Vector Routing Protocol
RIP is based on the distance vector algorithm.
Each router periodically shares its entire routing table with its immediate neighbors.
Routes are selected based on minimum hop count.

2️⃣ Hop Count as Metric (Maximum = 15)

RIP uses hop count as the only metric to measure route cost.

1 hop = one router away

Max valid hop count = 15
Anything ≥ 16 = unreachable
This limits RIP to small networks.

3️⃣ Periodic Updates (Every 30 Seconds)

Routers send their routing tables to neighbors every 30 seconds.
These are full-table updates, increasing overhead but keeping routes refreshed.

4️⃣ Slow Convergence & Count-to-Infinity Problem

RIP converges slowly when the network topology changes.
It suffers from classic distance-vector issues such as routing loops and the count-to-infinity problem,
requiring additional safeguards.

5️⃣ Loop Avoidance Techniques

RIP includes mechanisms to reduce routing loops:

Split Horizon
Split Horizon with Poison Reverse
Hold-down timers
These stabilize the network and prevent incorrect routing updates.

6️⃣ RIP Versions: IPv4 and IPv6 Support

RIP v1: classful, no subnet information
RIP v2: classless, supports VLSM & authentication
RIPng: supports IPv6
This makes RIP usable on both IPv4 and IPv6 networks.

✔ Summary (6 Marks Covered)

1. Distance-vector protocol
2. Hop count metric (max 15)

[Link] 114/116
11/14/25, 11:11 PM Study strategy for networks

3. 30-second periodic updates

4. Slow convergence / count-to-infinity
5. Loop-prevention mechanisms
6. Versions: RIP v1, v2, RIPng

If you want, I can also give a diagram-based explanation or comparison between RIP and OSPF.

A computer on a 6-Mbps network is regulated by a token bucket. The

token
bucket is filled at a rate of 1 Mbps. It is initially filled to capacity with 8
megabits. How long can the computer transmit at the full 6 Mbps? for
8 marks

Below is the full 8-mark solution, step-by-step, exactly how examiners expect it.

Given:
Network transmission rate = 6 Mbps
Token bucket fill rate = 1 Mbps
Bucket capacity (initial tokens) = 8 megabits

The computer wants to transmit at full 6 Mbps.

Concept (2 Marks)
In a token bucket:
To send 1 bit, you must spend 1 token.
Tokens arrive at 1 Mbps.
If you transmit at 6 Mbps, you consume tokens faster than they arrive.

Token consumption rate = 6 Mbps – 1 Mbps = 5 Mbps

This is the net drain rate from the bucket.

Token Bucket Depletion Calculation (4 Marks)

Bucket has 8 megabits stored initially.

Net drain = 5 Mbps

Time until bucket empties:

8 Mb
Time = = 1.6 seconds
5 Mb/s
​

Final Answer (2 Marks)

✔ The computer can transmit at the full 6 Mbps for 1.6 seconds.
After 1.6 seconds, the bucket is empty and the source must drop to the token fill rate (1 Mbps).

If you want, I can also solve a graph-based explanation or what happens after bucket empties.

[Link] 115/116
11/14/25, 11:11 PM Study strategy for networks

[Link] 116/116
11/14/25, 11:12 PM Study strategy for networks

Determine the solutions for the duplicate packets produced during

flooding for 6 marks

Here is a perfect 6-mark answer with clear, complete explanations of all solutions used to handle duplicate
packets during flooding.

Solutions for Duplicate Packets Produced During Flooding (6 Marks)

Flooding sends each incoming packet out on all outgoing links, which naturally creates duplicate packets.
To control this, networks use several mechanisms.

Below are the main solutions (any 3–4 expected for full 6 marks).

1️⃣ Sequence Numbers (1.5 Marks)

Each source attaches a unique sequence number to every packet.
Routers maintain a table of (Source Address, Sequence Number) pairs.
If a packet with the same pair arrives again → it is discarded immediately.
Only the first copy is forwarded.

Effect: Prevents processing and forwarding of duplicates.

2️⃣ Hop Count / Time To Live (TTL) Field (1.5 Marks)

Each packet is assigned a maximum hop count or TTL.

Every router decrements the hop count by 1.

When hop count reaches zero, the packet is discarded.
Effect: Prevents packets from circulating forever and reduces duplicate propagation.

3️⃣ Packet Lifetime / Timestamp (1.5 Marks)

Packet carries a timestamp or expiration time.
If the packet is too old, routers discard it.
Ensures copies do not survive long in the network.

Effect: Eliminates outdated duplicates quickly.

4️⃣ Visited Node List / "Don't Send Back" Rule (1.5 Marks)
Routers can store a record of recently seen packets or avoid sending packets back on the link they arrived
from.

Prevents immediate backflow duplicates.

Reduces unnecessary reverse flooding.
Effect: Limits redundant propagation of copies.

[Link] 1/1