Congestion Control

Definition

Congestion in a network occurs when the number of packets being transmitted through the network approaches the packet handling capacity of the network. Congestion control refers to the mechanisms and techniques used to control or prevent congestion so that network performance does not degrade.

Note: No specific reference notes were found for this topic; the answer below is based on standard networking curriculum (Forouzan / Tanenbaum) as taught in BSc CSIT.

Why Congestion Occurs

When too many packets are present in a part of the network, network performance degrades. This happens because:

• Routers have limited buffer (queue) space
• Links have limited bandwidth
• Processors in routers have limited processing speed

If the load exceeds capacity, packets are dropped, retransmissions increase, and the situation worsens -- leading to congestion collapse.

How Congestion Can Be Handled

Congestion control approaches are broadly classified as:

1. Open Loop Congestion Control (Prevention)

Policies are applied before congestion occurs -- at the design stage.

2. Closed Loop Congestion Control (Reaction)

Policies are applied after congestion is detected.

• Back pressure: A congested node sends a signal to the upstream node to slow down.
• Choke packet: A special packet is sent by the router to the source to reduce its transmission rate.
• Implicit signaling: Source detects congestion from delays or dropped packets (e.g., TCP timeout).
• Explicit signaling: Router explicitly marks packets to inform source/destination of congestion (e.g., ECN -- Explicit Congestion Notification).

Open Loop Congestion Control Policies (Detailed)

A. Acknowledgement Policy

The acknowledgement (ACK) policy at the receiver side can greatly affect congestion.

How it works:

• If the receiver sends an acknowledgement for every packet received, it encourages the sender to keep sending packets rapidly, which can contribute to congestion.
• By delaying acknowledgements, the receiver can slow down the sender and reduce network load.

Techniques:

Effect on Congestion:

• Fewer ACKs mean the sender's window advances more slowly.
• This reduces the rate at which new packets are injected into the network.
• Thus, acknowledgement policy directly helps in preventing congestion at the source.

B. Discarding Policy

The discarding policy is applied at routers/switches. When a router's buffer is full or near full, it must decide which packets to drop.

Approaches:

1. Tail Drop (Default / Simple)

• When the buffer is full, the router simply drops all incoming packets (from the tail of the queue).
• Problem: Can cause global synchronization -- many TCP connections reduce their window simultaneously, then increase together, causing oscillation.

2. Random Early Detection (RED)

• The router monitors its average queue length.
• When the queue length exceeds a minimum threshold (min_th), it starts randomly dropping packets with increasing probability.
• When the queue length exceeds a maximum threshold (max_th), all packets are dropped.

Queue Length:
0 -------- min_th -------- max_th -------- Max Buffer
           |                |
           Start random     Drop all
           dropping         packets

• Advantage: Prevents global synchronization; notifies individual connections early.
• Advantage: Maintains low average queue length and low delay.

3. Priority-Based Discarding

• Packets are assigned priority levels.
• When congestion occurs, lower priority packets are dropped first.
• Higher priority packets (e.g., real-time voice/video) are protected.

4. Weighted RED (WRED)

• Extension of RED where different traffic classes have different drop probabilities.
• Higher priority traffic has a lower drop probability.

Summary Table

Conclusion

Congestion control is essential for maintaining network performance and reliability. Open loop methods like acknowledgement policy and discarding policy prevent congestion proactively, while closed loop methods react to congestion after it is detected. Together, these mechanisms ensure fair and efficient use of network resources.

Practical Implications of OSI Layers and Their Definitions

Introduction

The OSI (Open Systems Interconnection) model is a conceptual framework developed by ISO that standardizes the functions of a communication system into seven distinct layers. It provides a universal language for networking, allowing diverse systems to communicate.

Practical Implications of the OSI Model

The OSI model has several important practical implications:

1. Standardization: Allows hardware and software from different vendors to work together (interoperability).
2. Troubleshooting: Network engineers can isolate problems layer by layer (e.g., a physical cable fault vs. a routing issue).
3. Modular Design: Each layer can be developed, updated, or replaced independently without affecting other layers.
4. Protocol Development: Guides the design of networking protocols (e.g., TCP/IP maps to OSI layers).
5. Security Implementation: Security measures can be applied at specific layers (e.g., firewalls at Network layer, SSL/TLS at Session/Presentation layer).
6. Teaching and Learning: Provides a structured way to understand complex networking concepts.
7. Vendor Independence: Promotes open standards so no single vendor controls the entire communication stack.

The Seven Layers of the OSI Model

Layer 7: Application Layer

Example: When you open a browser and type a URL, the Application layer handles the HTTP request.

Layer 6: Presentation Layer

Example: Converting an image file to JPEG format or encrypting data using SSL before transmission.

Layer 5: Session Layer

Example: When you log into a remote server, the Session layer manages the login session and ensures it stays active.

Layer 4: Transport Layer

Example: TCP breaks a large file into segments, numbers them, and reassembles them at the destination, requesting retransmission if any segment is lost.

Layer 3: Network Layer

Example: A router uses IP addresses to forward packets from your home network to a web server on the internet.

Layer 2: Data Link Layer

Example: An Ethernet switch uses MAC addresses to forward frames only to the correct port on a LAN.

Layer 1: Physical Layer

Example: A fiber optic cable transmitting light pulses representing binary data operates at the Physical layer.

Summary Table

Conclusion

The practical value of the OSI model is that it turns one large problem, moving data between two machines, into seven smaller problems with clearly defined boundaries. A fault can then be localised layer by layer, a cable or a NIC at layer 1 and 2, an IP or routing problem at layer 3, a port or firewall problem at layer 4 and an application or DNS problem above it, which is exactly how network troubleshooting is taught and practised. The same boundaries let vendors build interoperable equipment, let a protocol at one layer be replaced without touching the others, and give engineers a common vocabulary in which a "layer 3 switch" or a "layer 7 firewall" means something precise.

Circuit Switching

Definition

Circuit switching is a method of communication in which a dedicated physical path (circuit) is established between the sender and receiver before data transmission begins, and this path remains reserved for the entire duration of the communication session. The classic example is the traditional telephone network (PSTN).

The communication takes place in three phases:

1. Circuit Establishment - A dedicated path is set up from source to destination through intermediate switches.
2. Data Transfer - Data is transmitted over the reserved circuit.
3. Circuit Termination - The circuit is released/disconnected after communication ends.

Advantages of Circuit Switching

Disadvantages of Circuit Switching

Summary

Circuit switching trades resource efficiency for guaranteed, consistent performance, making it well-suited for voice telephony but less ideal for modern data networks where packet switching is preferred.

Short Notes on Reliable Protocol and Satellite Network

1. Reliable Protocol

A reliable protocol is a communication protocol that guarantees the delivery of data from sender to receiver correctly, completely, and in order.

Key Characteristics:

• Error Detection and Correction: Uses checksums, CRC, or parity bits to detect errors in transmitted data.
• Acknowledgement (ACK): The receiver sends an acknowledgement to the sender after successfully receiving data.
• Retransmission: If ACK is not received within a timeout period, the sender retransmits the lost or corrupted data.
• Sequencing: Data packets are numbered so the receiver can reorder them if they arrive out of sequence.
• Flow Control: Prevents the sender from overwhelming the receiver (e.g., sliding window protocol).
• Connection-Oriented: Typically establishes a connection before data transfer (e.g., TCP).

Example:

TCP (Transmission Control Protocol) is the most common reliable protocol. It ensures all segments are delivered, acknowledged, and reassembled in correct order.

2. Satellite Network

A satellite network is a type of wireless network that uses artificial satellites orbiting the Earth as relay stations to transmit data, voice, and video signals between ground stations.

Types of Satellite Orbits:

Key Characteristics:

• Wide Coverage: A single satellite can cover a large geographic area.
• High Latency (GEO): Due to long signal travel distance, GEO satellites have significant propagation delay.
• Broadcast Capability: Signals can be sent to multiple receivers simultaneously.
• Used in Remote Areas: Provides connectivity where terrestrial networks are unavailable.
• Bandwidth Limitations: Satellite bandwidth is shared among many users.

Applications:

• Internet access in rural/remote areas
• Military communications
• GPS and navigation
• Weather forecasting
• Television broadcasting

Note: Since no specific curriculum notes were provided, this answer is based on standard networking concepts as covered in typical BSc CSIT Data Communication and Networking courses.

CRC Encoding: Transmitted Frame Calculation

Given Data

• Message sequence $M = 10010001$
• Generator polynomial $G(X) = x^2 + x + 1$

Step 1: Generator Bit Pattern

$G(X) = x^2 + x + 1 \Rightarrow \mathbf{111}$

Degree $r = 2$, so append 2 zeros.

Step 2: Augmented Dividend

$$1001000100$$

Step 3: Modulo-2 Division by 111

Let me perform the division carefully, tracking each bit.

Dividend: 1 0 0 1 0 0 0 1 0 0
Divisor:  1 1 1

I'll process left to right. Take leading 3 bits, XOR when leading bit is 1.

1001000100 : 111

100 → leading 1, XOR 111 = 011
bring down 1 → 0111 → work on 111
111 → leading 1, XOR 111 = 000
bring down 0 → 0000 → 000
000 → leading 0, no XOR
bring down 0 → 0000 → 000
000 → leading 0, no XOR
bring down 1 → 0001 → 001
001 → leading 0, no XOR
bring down 0 → 0010 → 010
010 → leading 0, no XOR
bring down 0 → 0100 → 100
100 → leading 1, XOR 111 = 011
(no more bits)

Let me align this against the full 10-bit dividend to verify by tracking the running remainder bit by bit.

Running remainder method (shift in each dividend bit, XOR divisor 111 when MSB of the 3-bit window is 1):

To avoid confusion, use standard long-division layout:

        1 0 1 1 0 0 1
      -----------------
1 1 1 ) 1 0 0 1 0 0 0 1 0 0
        1 1 1
        -----
        0 1 1 0
          1 1 1        (align at next 1)

Let me do it strictly with the standard algorithm:

   1 0 0 1 0 0 0 1 0 0
 XOR 1 1 1                (pos 1-3): 100^111 = 011
 → 0 1 1 1 0 0 0 1 0 0
     XOR 1 1 1            (pos 2-4): 111^111 = 000
 → 0 0 0 0 0 0 0 1 0 0
                         (pos 3-5 leading 0, skip)
                         (pos 4-6 leading 0, skip)
                         (pos 5-7 leading 0, skip)
                         (pos 6-8 leading 0, skip; bits are 0,0,1)
             XOR 1 1 1    (pos 7-9): bits 0 1 0 leading 0 skip

After the second XOR the string is $0000000100$. The remaining bits are $...100$ at positions 8,9,10.

Reading positions 8-10: $1,0,0$. Leading bit position 8 is 1:

Position 8-10: $100 \oplus 111 = 011$.

Final string tail: last 2 bits = 11.

The remainder (last $r=2$ bits after full reduction) = $\mathbf{11}$.

Verification of Remainder

Reconstruct: string became $0000000100$, then XOR 111 at positions 8-10: $$100 \oplus 111 = 011$$ Final 2 bits = $\mathbf{11}$.

CRC = 11

Careful reduction gives remainder 11; a hand division that drops or misaligns bits in Step 3 would give 01 instead.

Step 4: Transmitted Frame

$$\text{Frame} = \underbrace{10010001}{M} ; \underbrace{11}{CRC} = \mathbf{1001000111}$$

Verification (divide frame by 111, expect remainder 0)

$1001000111$:

• $100 \oplus 111 = 011$ → $0111000111$
• $111 \oplus 111 = 000$ → $0000000111$
• tail $111 \oplus 111 = 000$ → remainder $\mathbf{00}$ ✓

Remainder is zero, confirming correctness.

Result

Transmitted Encoded Frame = 1001000111 (CRC = 11)