Data Networks Midterm

0.0(0)
studied byStudied by 2 people
0.0(0)
full-widthCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/107

encourage image

There's no tags or description

Looks like no tags are added yet.

Study Analytics
Name
Mastery
Learn
Test
Matching
Spaced

No study sessions yet.

108 Terms

1
New cards

OSI Model Layers

Application, Presentation, Session, Transport, Network, Data Link, Physical.

2
New cards

Connection-Oriented vs Connectionless

Connection-oriented requires setup, data transfer, and teardown; connectionless sends data without setup.

3
New cards

Encapsulation

Each layer adds its own header to data from the upper layer before transmission.

4
New cards

Frames vs Packets

Frames encapsulate packets at the data link layer, including headers and trailers.

5
New cards

Fraction of Bandwidth Used for Headers

Given n layers with header size h and message M, fraction = nh / (nh + M).

6
New cards

Purpose of Modulation

Convert digital data into analog form suitable for transmission over analog media.

7
New cards

Amplitude Modulation

Signal amplitude varies with the carrier, while frequency and phase remain constant.

8
New cards

Frequency Modulation

Carrier frequency varies with the input signal amplitude.

9
New cards

Phase Modulation

Carrier phase varies with the input signal amplitude.

10
New cards

QPSK (Quadrature Phase Shift Keying)

Uses four distinct phase shifts to send two bits per symbol.

11
New cards

QAM (Quadrature Amplitude Modulation)

Combines amplitude and phase variations to send multiple bits per symbol (e.g., 4–6 bits/symbol).

12
New cards

Baud Rate

Number of symbols transmitted per second, not necessarily equal to bit rate.

13
New cards

Bandwidth

Range of frequencies that can pass through a channel without significant attenuation.

14
New cards

Time Division Multiplexing (TDM)

Shares a channel by assigning each signal a fixed time slot in a repeating cycle.

15
New cards

Frequency Division Multiplexing (FDM)

Shares a channel by allocating distinct frequency bands to different signals.

16
New cards

Wavelength Division Multiplexing (WDM)

Optical version of FDM; multiple wavelengths carried on a single fiber.

17
New cards

Delta Modulation

Encodes the direction of change (+1 or -1) between consecutive samples using a single bit.

18
New cards

Predictive Encoding

Transmitter predicts next sample value; only sends difference from prediction.

19
New cards

Message Switching

Entire messages are stored and forwarded, causing delay until message arrival.

20
New cards

Packet Switching

Messages are broken into packets sent independently; improves efficiency and resource use.

21
New cards

Circuit Switching

Establishes a dedicated physical path between sender and receiver for duration of communication.

22
New cards

Stop-and-Wait Protocol

Sender transmits one frame, waits for ACK before sending next; simple but inefficient for long delays.

23
New cards

Piggybacking

ACKs are attached to outgoing data frames to reduce overhead, but delayed data may cause retransmissions.

24
New cards

Bit Stuffing

Inserts a 0 after five consecutive 1s to prevent confusion with the flag pattern (01111110).

25
New cards

Byte Stuffing

Adds escape (ESC) characters before special bytes like FLAG or ESC in data stream.

26
New cards

Flag Pattern in Framing

01111110 is used to mark start and end of frames in bit-oriented protocols like HDLC or PPP.

27
New cards

Purpose of Checksum

Detects transmission errors by comparing recomputed value with the received one.

28
New cards

Hamming Distance

Minimum number of bit changes required to convert one valid codeword to another.

29
New cards

Error Correction Ability

A code with Hamming distance d can detect up to d−1 errors and correct up to (d−1)/2.

30
New cards

Hamming Code Parity Bit Positions

Parity bits are placed at positions 1, 2, 4, 8, 16… corresponding to powers of two.

31
New cards

Hamming Code Example

For 16-bit data, five parity bits are enough to correct single-bit errors (fits within 31 bits).

32
New cards

CRC (Cyclic Redundancy Check)

Uses polynomial division over binary data to generate a remainder that detects burst errors.

33
New cards

CRC Detection Mechanism

Receiver divides received bits by generator polynomial; nonzero remainder indicates error.

34
New cards

Polynomial Used in CRC Example

x³ + 1 corresponds to binary 1001.

35
New cards

Purpose of Sequence Numbers

Distinguish new frames from retransmissions and ensure proper ordering.

36
New cards

Positive Acknowledgment with Retransmission (PAR)

Sender retransmits frame if ACK not received within timeout; ensures reliability over noisy channels.

37
New cards

Timer Expiration

When ACK doesn’t arrive before timeout, sender retransmits the unacknowledged frame.

38
New cards

Simplex Stop-and-Wait (Protocol 2)

Adds flow control to prevent receiver buffer overflow; uses dummy ACK frames.

39
New cards

Simplex for a Noisy Channel (Protocol 3)

Adds sequence numbers and timers for error + flow control; 1-bit sequence prevents duplication.

40
New cards

Sliding Window Protocol

Allows multiple outstanding frames before ACKs; improves efficiency over long-delay channels.

41
New cards

Sender Window

Set of sequence numbers sender can send before blocking; equals max unacknowledged frames.

42
New cards

Receiver Window

Set of frames receiver is willing to accept; outside this range are discarded.

43
New cards

Go-Back-N Protocol

On error, sender retransmits from the damaged frame onward; receiver only accepts next expected frame.

44
New cards

Selective Repeat Protocol

Only the specific damaged frame is retransmitted; receiver buffers out-of-order frames.

45
New cards

Window Size in Selective Repeat

Must be ≤ (sequence number space / 2) to avoid ambiguity between new and old frames.

46
New cards

Stop-and-Wait Efficiency Formula

η = 1 / (1 + 2a), where a = propagation delay / transmission time.

47
New cards

Frame Transmission Delay

Frame size (bits) / channel bandwidth (bps).

48
New cards

Propagation Delay

Distance / propagation speed; for 800 km fiber at 2×10⁵ km/s, delay = 4 ms.

49
New cards

Queuing Delay Formula

(nL + (L − x)) / R, where n = packets waiting, L = length, R = link rate, x = bits already sent.

50
New cards

Channel Utilization Example (Satellite)

With 1 Mbps link, 1000-bit frames, and 270 ms one-way delay, efficiency ≈ 0.18%.

51
New cards

Collisions in Broadcast Subnet

Fraction wasted = 1 − (np(1−p)ⁿ⁻¹ + (1−p)ⁿ), where n = hosts, p = transmit probability.

52
New cards

ATM Fixed-Length Cells

Small, fixed cells (53 bytes) simplify fast hardware switching and reduce delay variation.

53
New cards

Connection Setup Phases

Connection establishment → data transfer → connection release.

54
New cards

Negotiation in Protocols

Exchange of control messages to agree on parameters like max size or QoS before data transfer.

55
New cards

Flow Control Purpose

Prevents sender from overwhelming receiver with data faster than it can process.

56
New cards

Error Control Purpose

Ensures corrupted or lost frames are detected and retransmitted to maintain reliability.

57
New cards

Data Link Layer Services

Framing, error control, flow control, and reliable frame delivery to the network layer.

58
New cards

Piggybacking Problem

If receiver has no data to send back soon, delayed ACKs can cause sender timeouts.

59
New cards

Abnormal Case in 1-Bit Sliding Window

Occurs when sender and receiver start transmitting simultaneously, causing duplication.

60
New cards

One-Bit Sliding Window Equivalence

Stop-and-Wait; window size = 1.

61
New cards

Difference Between Go-Back-N and Selective Repeat

Go-Back-N retransmits all frames after error; Selective Repeat retransmits only the lost one.

62
New cards

Acknowledgment Types

Individual ACK for each frame or cumulative ACK acknowledging all up to a given frame.

63
New cards

When to Use Selective Repeat

Preferred when propagation delay is high and bandwidth is large; minimizes retransmission overhead.

64
New cards

Parity Check in 2D Scheme

Detects all single and double errors; may miss some triple errors if aligned along same row/column.

65
New cards

Role of Data Link Layer

Provides reliable communication over a single link between adjacent nodes.

66
New cards

Why Use Small Fixed-Length Cells in ATM

Simplifies switch design and reduces delay for real-time traffic.

67
New cards

Why Multiplexing is Cost-Effective

Combines multiple low-bandwidth channels into one high-bandwidth trunk, reducing infrastructure cost.

68
New cards

Bandwidth vs Bit Rate

Bandwidth is frequency range (Hz); bit rate is data transmitted per second (bps).

69
New cards

Propagation vs Transmission Delay

Propagation = time signal takes to travel; transmission = time to push bits into medium.

70
New cards

Difference Between Packet, Message, and Circuit Switching

Circuit = dedicated path; packet = split data units; message = store-and-forward entire message.

71
New cards

Go-Back-N Behavior on Error

Receiver discards erroneous and subsequent frames; sender retransmits from error frame onward.

72
New cards

Selective Repeat Receiver Behavior

Buffers correct frames received after an error; waits for missing frame to complete sequence.

73
New cards

Purpose of ACK Number

Specifies the next frame expected; confirms receipt of all prior frames.

74
New cards

Bit Errors and Checksums

Longer checksum reduces probability of undetected error, but never to zero.

75
New cards

Importance of Window Size

Determines pipeline length and channel utilization efficiency.

76
New cards

Advantages of Layered Design

Simplifies debugging, allows interoperability, and isolates changes between protocol layers.

77
New cards

OSI vs TCP/IP Layers

Both have network, transport, and application layers; OSI has more layers and supports both service types.

78
New cards

Reasons for Using Standards

Enable interoperability and economies of scale, though they can become outdated or politically compromised.

79
New cards

Connection-Oriented Example

TCP uses connection-oriented service for reliable stream delivery.

80
New cards

Connectionless Example

UDP sends datagrams without establishing a connection.

81
New cards

Common Physical Media

Twisted pair, coaxial cable, optical fiber, and wireless radio.

82
New cards

Nyquist Sampling for PCM

Requires at least twice the signal bandwidth (4 kHz voice → 8 kHz sampling → 125 µs/sample).

83
New cards

FDM Guard Band Purpose

Prevents channel overlap and interference between adjacent frequency bands.

84
New cards

Header Overhead Calculation Example

For 7 layers adding 20 bytes each to a 1000-byte message → 140 / 1140 = 12.3% overhead.

85
New cards

Reason for Using Hamming Codes

Allows single-bit error correction and double-bit detection at the data link layer.

86
New cards

CRC Remainder Example

Dividing x⁷ + x⁵ + 1 by x³ + 1 yields remainder x² + x + 1.

87
New cards

Transmission Efficiency Improvement

Achieved by pipelining frames via Go-Back-N or Selective Repeat protocols.

88
New cards

Problem Solved by Sliding Window Protocols

Overcome inefficiency of Stop-and-Wait in long-delay networks.

89
New cards

Purpose of Dummy Frames in Protocol 2

Receiver uses dummy frames to signal readiness and regulate data flow.

90
New cards

When ACK is Lost

Sender retransmits frame; receiver discards duplicate and sends ACK again.

91
New cards

Window Size = 1 Implication

Equivalent to Stop-and-Wait operation, fully sequential transmission.

92
New cards

Ack Timeout Too Short

Causes unnecessary retransmissions even when frames were correctly received.

93
New cards

Ack Timeout Too Long

Decreases efficiency due to idle waiting before retransmission.

94
New cards

Checksum vs CRC

Checksum sums bits modulo arithmetic; CRC uses polynomial division for stronger error detection.

95
New cards

Advantages of Packet Switching

Efficient bandwidth use, resilience to failure, and dynamic routing capability.

96
New cards

Purpose of Error Control in DLL

Ensures frames are delivered correctly to network layer through retransmission schemes.

97
New cards

Reason for Piggybacking Limitation

Waiting for data to send can delay ACK, triggering sender’s timer and retransmission.

98
New cards

Propagation Delay Dominant Channels

In long-distance satellite or fiber links, pipelined transmission significantly increases utilization.

99
New cards

Fraction of Slots Wasted in Collisions

1 − [n p (1−p)ⁿ⁻¹ + (1−p)ⁿ] represents wasted slots due to multiple simultaneous transmissions.

100
New cards

Benefit of Predictive Encoding

Reduces redundancy by sending only differences from predicted signal, improving compression.