Data Link Control and Media Access

Introduction

The data link layer provides reliable and orderly communication between directly connected network nodes. While the physical layer transmits raw bits, the data link layer organizes those bits into frames, detects or corrects transmission errors, regulates the rate of transmission, and determines which device may use a shared medium. These functions are essential in local area networks, point-to-point links, wireless networks, industrial networks, and many embedded systems.

A data link protocol operates across a single link or broadcast domain. It is therefore different from the network layer, which selects paths across multiple interconnected networks. Common data link technologies include Ethernet, Wi-Fi, Point-to-Point Protocol (PPP), High-Level Data Link Control (HDLC), and various industrial and telecommunications protocols.

The central design problem is that a communication channel is imperfect and often shared. Bits may be corrupted by noise, frames may be lost, and several stations may attempt transmission simultaneously. Data link control supplies mechanisms that make communication efficient, dependable, and fair.

Learning Objectives

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

1. Explain the purpose and methods of framing.
2. Distinguish error detection from error correction.
3. Calculate and interpret parity, checksum, and cyclic redundancy check operations.
4. Explain flow control, stop-and-wait, sliding-window protocols, and ARQ.
5. Describe the structure and functions of HDLC, PPP, and Ethernet frames.
6. Compare random-access, controlled-access, and channel-partitioning methods.
7. Explain the operation and limitations of CSMA/CD and CSMA/CA.
8. Select a suitable access method for a practical network scenario.

1. Framing: Creating Manageable Data Units

The physical layer delivers a stream of bits without inherent boundaries. The receiver must know where one data unit ends and the next begins. Framing is the process of dividing the bit stream into identifiable units called frames.

A typical frame contains:

• A header, carrying addresses, control information, or protocol identifiers.
• A payload, containing data from the network layer.
• A trailer, commonly containing an error-detection field such as a frame check sequence (FCS).
• Sometimes a preamble or start-of-frame delimiter for synchronization.

Framing supports synchronization, addressing, error checking, and multiplexing of network-layer protocols. It also provides a unit to which acknowledgements, retransmissions, and flow-control decisions can refer.

1.1 Character Count

In character-count framing, a length field specifies the number of characters or bytes in the frame. The receiver reads the length and identifies the frame boundary.

This method is simple and efficient when the length field is correct. However, if a bit error changes the length value, the receiver may lose synchronization and interpret subsequent data incorrectly. Recovery may require waiting for a recognizable synchronization pattern.

1.2 Byte-Oriented Framing and Byte Stuffing

A byte-oriented protocol may reserve special bytes to indicate the beginning and end of a frame. If a reserved byte occurs in the payload, the sender inserts an escape byte before it. This operation is called byte stuffing. The receiver removes the escape byte before delivering the payload.

Byte stuffing is appropriate for protocols whose data is naturally organized into bytes. Its overhead depends on the frequency of reserved values in the payload.

1.3 Bit-Oriented Framing and Bit Stuffing

Bit-oriented protocols use a flag pattern, commonly 01111110, to mark frame boundaries. To prevent this pattern from appearing in ordinary data, the sender inserts a 0 after every sequence of five consecutive 1s in the payload. The receiver removes such inserted zeros.

Bit stuffing is independent of character encoding and is therefore suitable for arbitrary binary data. It adds variable overhead, but the overhead is normally small.

1.4 Physical-Layer Coding Violations

Some systems use signal patterns that are not valid data symbols as frame delimiters. A receiver recognizes the unusual physical signal as a boundary. This approach avoids stuffing overhead but requires a physical-layer encoding with unused signal states.

1.5 Frame Synchronization and Design Trade-offs

A good framing method should be easy to implement, resistant to errors, and efficient for expected traffic. Important design considerations include maximum frame size, overhead, recovery after corruption, compatibility with hardware, and support for variable-length payloads.

2. Transmission Errors

A transmitted bit may change from 0 to 1 or from 1 to 0 because of electromagnetic interference, attenuation, crosstalk, faulty equipment, multipath propagation, or synchronization problems. Errors may be classified as:

• Single-bit errors: one bit is corrupted.
• Burst errors: several bits in a sequence are affected, although not necessarily every bit in the sequence is changed.

A burst error of length L is commonly described by the distance from the first corrupted bit to the last corrupted bit, inclusive. Burst errors are especially important on wireless and long-distance links.

The data link layer can detect errors, correct them directly, or request retransmission from the sender. Detection is generally less expensive than correction, so practical networks often combine error detection with Automatic Repeat reQuest (ARQ).

3. Error Detection and Correction

3.1 Detection and Hamming Distance

The Hamming distance between two equal-length bit strings is the number of bit positions in which they differ. A code with minimum Hamming distance d_min can detect up to d_min − 1 errors and can correct up to floor((d_min − 1) ÷ 2) errors, provided the errors fall within the code's designed operating assumptions.

Detection and correction capabilities are properties of the complete set of valid codewords, not merely of an individual frame. In practical communication systems, noise models and coding schemes determine whether these theoretical capabilities are achieved.

3.2 Parity

In even parity, a parity bit is added so that the total number of 1 bits is even. In odd parity, the total number is odd. A single-bit error changes the parity and is detected. However, an even number of bit errors may leave parity unchanged.

Two-dimensional parity applies parity to both rows and columns of a block. It can detect many multiple-bit errors and may locate and correct a single erroneous bit at the intersection of a bad row and bad column.

3.3 Checksum

A checksum divides data into fixed-size words, adds them using one's-complement arithmetic, and complements the result. The receiver performs the same operation, including the checksum, and tests whether the expected result is obtained.

Checksums are computationally simple and remain common in transport and network protocols. Their error-detection strength is generally lower than that of a well-designed CRC for the same number of check bits.

3.4 Cyclic Redundancy Check

A CRC treats a bit string as the coefficients of a polynomial over the binary field GF(2). Addition and subtraction are equivalent to exclusive OR, and division uses no carries or borrows.

Let the data polynomial be D(x), the generator polynomial be G(x), and the degree of G(x) be r. The sender appends r zeros to the data and divides D(x) × x^r by G(x). If the remainder is R(x), the transmitted codeword is:

Because arithmetic is modulo 2, addition is XOR. The remainder R(x) has degree less than r. The receiver divides T(x) by G(x). A zero remainder indicates that no error was detected; a nonzero remainder indicates an error under the CRC's detection rules.

The CRC does not prove that a frame is correct. Some error patterns can produce a zero remainder, although suitable generator polynomials make important classes of errors highly unlikely or mathematically undetectable only under known conditions.

#### Worked Example: CRC

Suppose the data is 1101 and the generator is 1011. The generator has degree 3, so append three zeros:

1101 000

Modulo-2 division of 1101000 by 1011 produces remainder 001. Therefore, the transmitted codeword is:

1101 001

The receiver divides 1101001 by 1011. If the remainder is 000, the frame passes the CRC test. If transmission changes one or more bits and the resulting remainder is nonzero, the frame is rejected.

4. Flow Control

Flow control prevents a fast sender from overwhelming a slower receiver. It is different from congestion control. Flow control concerns the capacity of a particular receiver or link, whereas congestion control concerns excessive traffic within a network.

4.1 Stop-and-Wait Flow Control

The sender transmits one frame and waits for permission or an acknowledgement before sending another. Its main advantages are simplicity and small buffer requirements. Its major disadvantage is poor link utilization when propagation delay is large.

For a link with frame transmission time T_t and one-way propagation time T_p, ignoring processing and acknowledgement transmission time, the approximate stop-and-wait utilization without errors is:

Here U is the fraction of time spent transmitting useful frames. The expression assumes that the sender waits for an acknowledgement and that the acknowledgement itself is negligible in transmission time. If acknowledgement transmission and processing times are significant, they must be added to the cycle time.

The bandwidth-delay product is:

where R is the link rate in bits per second and RTT is the round-trip time in seconds. BDP represents the number of bits that could be in transit during one round trip. A sliding window should be large enough to keep this capacity product substantially utilized.

4.2 Sliding-Window Flow Control

A sliding-window protocol allows several frames to be transmitted before acknowledgements arrive. The sender maintains a sending window, and the receiver maintains a receiving window. As acknowledgements arrive, the window advances, or slides, allowing new frames to be sent.

If sequence numbers use m bits, there are 2^m possible sequence values. The maximum safe window size depends on the protocol. For Go-Back-N, a common condition is a sender window no larger than 2^m − 1. Selective Repeat generally requires the sender and receiver windows to be no larger than 2^(m−1), so that old and new frames cannot be confused after sequence-number wraparound.

5. Error Control and ARQ

Automatic Repeat reQuest combines error detection, acknowledgements, timers, and retransmission. A receiver acknowledges correctly received frames or reports missing/corrupted frames. If the sender receives a negative acknowledgement or a timer expires, it retransmits.

5.1 Stop-and-Wait ARQ

Stop-and-Wait ARQ sends one frame and waits for an acknowledgement. A sequence number, often alternating between 0 and 1, distinguishes a new frame from a duplicate. This is necessary because an acknowledgement itself may be lost. If the sender retransmits after an acknowledgement was lost, the receiver can discard the duplicate while acknowledging the already accepted frame.

A timer must exceed the expected round-trip time, while remaining short enough to recover promptly from loss. A timeout that is too short causes unnecessary retransmissions; one that is too long reduces responsiveness.

5.2 Go-Back-N ARQ

Go-Back-N permits multiple outstanding frames but normally accepts only frames in order. If frame k is lost, the receiver may discard frame k+1 and later frames. The sender retransmits frame k and all subsequent outstanding frames.

This method is simple and efficient when errors are infrequent. On a noisy link, however, retransmitting correctly received frames wastes bandwidth.

5.3 Selective Repeat ARQ

Selective Repeat acknowledges frames individually and buffers correctly received out-of-order frames. Only missing or damaged frames are retransmitted. It uses more memory and more complex sequence-number management than Go-Back-N, but it is efficient on long-delay or error-prone links.

5.4 Piggybacking

In bidirectional communication, an acknowledgement can be included in an outgoing data frame instead of being sent in a separate frame. This is called piggybacking. It reduces overhead, but the receiver must not delay acknowledgements so long that the sender's timer expires.

6. Link-Layer Protocols and Frame Structures

6.1 HDLC

High-Level Data Link Control is a bit-oriented protocol that supports point-to-point and multipoint links. An HDLC frame commonly contains a flag, address, control field, information field, frame check sequence, and ending flag. The flag is 01111110, and bit stuffing prevents accidental flag patterns in the payload.

HDLC supports information frames for user data, supervisory frames for flow and error control, and unnumbered frames for link management. Its concepts influenced several later link protocols.

6.2 Point-to-Point Protocol

PPP was designed for direct links between two nodes. Its operation includes link establishment and configuration, optional authentication, and network-layer protocol configuration. PPP frames contain flag, address, control, protocol, information, and FCS fields, with delimiters and fields subject to negotiated options.

PPP is important conceptually because it demonstrates that a link protocol may perform framing, error detection, link configuration, authentication, and protocol identification without providing sophisticated error correction.

6.3 Ethernet

Modern Ethernet commonly uses a frame containing destination and source MAC addresses, a type or length field, payload, padding where required, and an FCS. Ethernet switches usually create separate collision domains for their ports, and full-duplex switched Ethernet does not use CSMA/CD during normal operation.

A MAC address identifies a network interface at the data link layer. It is not the same as an IP address: a MAC address is used for local delivery on a link, whereas an IP address supports logical addressing across interconnected networks.

7. Medium Access Control

When several stations share a transmission medium, a medium access control (MAC) method determines who may transmit. MAC methods can be classified as random access, controlled access, or channel partitioning.

7.1 Random Access

In random-access protocols, stations compete for the channel. Collisions may occur, and the protocol responds through detection, avoidance, or retransmission.

#### ALOHA

Pure ALOHA permits transmission whenever a station has data. If a collision occurs, the station waits a random time and retransmits. Its simplicity is attractive, but its maximum theoretical throughput is low because frames can collide throughout a vulnerable period.

Slotted ALOHA restricts transmissions to time slots. This reduces the vulnerable period and improves maximum throughput, but requires time synchronization.

#### CSMA

Carrier Sense Multiple Access requires a station to listen before transmitting. If the channel is sensed idle, the station transmits according to the protocol's persistence rule. If the channel is busy, it waits.

In 1-persistent CSMA, a station transmits immediately when the channel becomes idle. In nonpersistent CSMA, it waits a random interval after finding the channel busy. p-persistent CSMA, commonly associated with slotted channels, transmits with probability p when an idle slot is observed.

7.2 CSMA/CD

Carrier Sense Multiple Access with Collision Detection was used in traditional shared, half-duplex Ethernet. A station:

1. Senses the channel.
2. Transmits if the channel is idle.
3. Continues monitoring while transmitting.
4. Stops immediately if it detects a collision.
5. Sends a jam signal so other stations recognize the collision.
6. Waits for a random backoff interval before retrying.

Binary exponential backoff selects a number of slot intervals from an expanding range after repeated collisions. If the collision attempt number is k, the contention range is bounded according to the Ethernet standard, rather than growing without limit.

Collision detection requires that a station still be transmitting when a collision signal can return from the farthest part of the collision domain. This requirement influenced the minimum Ethernet frame size and maximum network diameter. Switched full-duplex Ethernet eliminates collisions on a point-to-point switch link, so CSMA/CD is not used in that normal operating mode.

7.3 CSMA/CA

Wireless networks generally cannot rely on collision detection. A wireless station may be unable to hear a collision while transmitting, and hidden terminals may be outside one another's sensing range. Wi-Fi therefore uses Carrier Sense Multiple Access with Collision Avoidance.

A station senses the channel, waits for an interframe space, and selects a random backoff counter. The counter decreases during idle time and freezes while the channel is busy. Transmission begins when the counter reaches zero. The receiver normally sends an acknowledgement, allowing the sender to infer successful delivery.

The optional RTS/CTS exchange can reduce hidden-terminal effects. A sender first transmits Request to Send; the receiver responds with Clear to Send. Other stations that hear these frames defer for the announced reservation period. RTS/CTS adds overhead and is most useful when collision costs are high, such as with large frames or hidden terminals.

Wireless performance is also affected by interference, fading, signal strength, mobility, exposed terminals, and changing data rates. Collision avoidance reduces collisions but cannot guarantee that they never occur.

7.4 Controlled Access

Controlled-access methods assign or grant transmission opportunities rather than relying entirely on contention.

• Reservation: stations reserve future slots, often using a reservation mini-frame.
• Polling: a central controller invites each station to transmit. Polling is orderly but creates controller overhead and a possible single point of failure.
• Token passing: possession of a token grants permission to transmit. It offers predictable access time, but token loss, duplication, or network partition must be handled.

Controlled access is useful where fairness, bounded delay, or deterministic service is more important than minimal control overhead.

7.5 Channel Partitioning

Channel-partitioning methods divide a resource among users.

• FDMA divides frequency bands.
• TDMA divides time into slots.
• CDMA assigns spreading codes and allows overlapping transmissions separated by code correlation.
• OFDMA assigns groups of orthogonal subcarriers to users and is widely used in contemporary broadband wireless systems.

Partitioning can provide predictable access but may waste capacity when assigned users have no data. Dynamic scheduling can improve utilization by allocating resources according to demand and channel conditions.

8. Comparing Access Methods

The correct method depends on propagation conditions, number of stations, traffic burstiness, required delay bound, fairness requirements, mobility, and whether collision detection is physically possible.

9. Applied Case: Selecting Access Control for a Campus Network

Background

A college is modernizing three network areas: a wired computer laboratory, a Wi-Fi classroom building, and an industrial automation laboratory. The network team must select appropriate link-layer behaviour and reduce complaints about slow or inconsistent access.

Problem

The computer laboratory uses modern switches, the classroom has many mobile devices, and the automation laboratory requires predictable delivery times. Applying one access strategy to all three environments would produce poor results.

Analysis

The switched laboratory has point-to-point full-duplex links. Each switch port forms an independent collision domain, so CSMA/CD is not required. Performance should instead be improved through appropriate switch capacity, segmentation, and adequate uplinks.

The classroom is a shared wireless environment. Stations cannot reliably detect collisions while transmitting, and some devices may be hidden from others. CSMA/CA with acknowledgements is appropriate. Access points should be positioned to improve signal quality, channels should be planned to reduce interference, and RTS/CTS may be tested where hidden terminals are observed. Excessive broadcast traffic and obsolete low-rate clients should be investigated because they consume airtime.

The automation laboratory requires bounded access delay. A controlled schedule, industrial Ethernet mechanism, or carefully engineered time-sensitive networking solution is more suitable than unrestricted contention. Traffic should be prioritized, redundant paths considered, and timing requirements measured rather than assumed.

Approach and Outcome

The team retains switched full-duplex Ethernet in the computer laboratory, redesigns wireless coverage and channel allocation in the classroom, and introduces scheduled or prioritized industrial communication in the automation area. Monitoring confirms fewer wireless retransmissions, stable wired throughput, and more predictable automation latency.

Learning Points

1. CSMA/CD is not a general solution for all Ethernet networks; it belongs mainly to shared half-duplex operation.
2. Wireless access requires collision avoidance because collision detection is difficult during transmission.
3. Deterministic applications may require controlled access or scheduled transmission.
4. Link-layer design must reflect physical conditions, traffic patterns, and service requirements.

10. Practical Design and Troubleshooting Guidance

When diagnosing a link problem, begin by identifying whether the issue is framing, physical transmission, error control, flow control, or media access. Useful evidence includes FCS errors, runt or oversized frames, retransmission counts, duplex mismatches, collision statistics, wireless retry rates, signal-to-noise measurements, and queue occupancy.

Best practices include:

• Use full-duplex switched Ethernet wherever practical.
• Configure matching speed and duplex settings and verify auto-negotiation behaviour.
• Choose MTU and frame sizes consistently across a path.
• Use CRC or an equivalent strong error detector for frame integrity.
• Set ARQ timers from measured round-trip behaviour, not arbitrary guesses.
• Size sliding windows according to bandwidth-delay product and sequence-number limits.
• Avoid excessive retransmission caused by weak wireless signals or interference.
• Use authentication and encryption at appropriate layers; FCS detects accidental corruption but does not provide confidentiality or authenticity.
• Monitor control-plane and broadcast traffic because access performance can be degraded even when link capacity appears sufficient.

11. Limitations, Security, and Contemporary Developments

Error detection cannot distinguish all possible corruptions, and retransmission may be impossible when a delay bound is strict. Flow control can prevent receiver overflow but cannot solve network-wide congestion. Random access is simple and flexible but offers weaker delay guarantees under heavy load. Controlled access improves predictability but requires coordination.

MAC addresses and link-layer frames can be observed, forged, or manipulated by an attacker on an insecure local network. Frame checks do not authenticate the sender. Secure protocols, access control, network segmentation, encryption, and monitoring are therefore necessary. Wireless networks additionally require robust authentication and protection against rogue access points, deauthentication attacks, and unauthorized association.

Modern networks increasingly combine high-speed switching, programmable hardware, virtualization, software-defined networking, scheduled traffic, and wireless resource allocation. Wi-Fi generations use more sophisticated scheduling, multiple antennas, improved coding, and multi-user transmission. Industrial and carrier networks use time synchronization and traffic shaping where predictable latency is essential. Artificial intelligence may assist with anomaly detection and channel optimization, but it does not replace protocol correctness, security controls, or engineering measurement.

Chapter Summary

The data link layer transforms an unstructured bit stream into frames and provides local delivery functions. Framing methods include character count, byte stuffing, bit stuffing, and physical-layer delimiters. Errors may be detected with parity, checksums, or CRCs; correction may use coding or retransmission. CRCs use modulo-2 polynomial division and are particularly effective for detecting burst errors.

Flow control prevents a sender from overwhelming a receiver. Stop-and-wait is simple but may waste capacity on long-delay links, while sliding windows keep multiple frames in transit. ARQ protocols recover from errors using acknowledgements, timers, sequence numbers, and retransmissions. Go-Back-N retransmits from a damaged frame onward, whereas Selective Repeat retransmits only missing or damaged frames.

HDLC, PPP, and Ethernet illustrate different link-layer designs. Medium access methods include random access, controlled access, and channel partitioning. CSMA/CD detects collisions in traditional shared Ethernet, while CSMA/CA reduces collision likelihood in wireless networks. Polling, token passing, reservation, FDMA, TDMA, CDMA, and OFDMA provide alternatives when fairness, predictability, or simultaneous sharing is required. Effective network design selects the method that matches the medium, traffic, delay, reliability, and security requirements.

Key Terms

Framing: Dividing a bit stream into identifiable data link frames.

Frame: A data unit handled by the data link layer.

Payload: The portion of a frame carrying upper-layer data.

FCS: Frame Check Sequence used to detect transmission errors.

Parity: An error-detection method based on the count of 1 bits.

Checksum: A computed value used to detect changes in transmitted data.

CRC: Cyclic Redundancy Check based on modulo-2 polynomial division.

Hamming distance: Number of differing bit positions between two codewords.

Flow control: Regulation of transmission to protect a receiver from overload.

Sliding window: A protocol mechanism permitting multiple unacknowledged frames.

ARQ: Automatic Repeat reQuest using acknowledgement and retransmission.

Go-Back-N: ARQ method retransmitting a damaged frame and later outstanding frames.

Selective Repeat: ARQ method retransmitting only frames that require recovery.

MAC address: Link-layer address identifying a network interface locally.

CSMA/CD: Carrier Sense Multiple Access with Collision Detection.

CSMA/CA: Carrier Sense Multiple Access with Collision Avoidance.

Polling: Controlled access in which a controller grants transmission turns.

Token passing: Controlled access based on possession of a permission token.

TDMA: Channel partitioning by time slots.

RTS/CTS: Wireless reservation exchange using Request to Send and Clear to Send.

Review Questions

Short Answer Questions

1. What is the purpose of framing at the data link layer?
2. Distinguish error detection from error correction.
3. Why are sequence numbers needed in Stop-and-Wait ARQ?
4. State one difference between Go-Back-N and Selective Repeat.
5. Why is CSMA/CD generally unnecessary on modern full-duplex switched Ethernet?

Long Answer Questions

1. Explain character count, byte stuffing, bit stuffing, and physical-layer coding violations as framing methods.
2. Describe CRC operation, including the roles of the data polynomial, generator polynomial, remainder, and receiver test.
3. Compare stop-and-wait flow control with sliding-window flow control and discuss bandwidth-delay effects.
4. Explain the operation of CSMA/CA and the hidden-terminal problem.
5. Compare random access, controlled access, and channel-partitioning methods, including suitable applications for each.

Analytical / Application Questions

1. A link has a transmission time of 2 ms and a one-way propagation time of 20 ms. Using the stated stop-and-wait approximation and ignoring acknowledgement transmission time, calculate utilization. Explain why a sliding window would improve performance.
2. A wireless classroom experiences high retransmission rates only when two distant groups of devices communicate with the same access point. Analyze the likely cause and recommend two engineering responses.
3. An industrial controller requires bounded access delay, while a separate office network has bursty ordinary data traffic. Recommend a suitable access strategy for each environment and justify the choices using reliability, fairness, and delay considerations.

Multiple Choice Questions with Answers

Question 1

What is the primary purpose of framing?

A. To assign IP addresses to hosts

B. To identify boundaries in a bit stream

C. To encrypt all application data

D. To select an internet route

Correct Answer: B. To identify boundaries in a bit stream

Question 2

In even parity, what condition is enforced?

A. The total number of 1 bits is odd

B. The total number of 0 bits is zero

C. The total number of 1 bits is even

D. Every byte contains exactly four 1 bits

Correct Answer: C. The total number of 1 bits is even

Question 3

If a generator polynomial has degree r, how many zeros are appended before CRC division?

A. r

B. r − 1

C. 2r

D. r + 1

Correct Answer: A. r

Question 4

Which ARQ method retransmits a damaged frame and all later outstanding frames?

A. Selective Repeat

B. Go-Back-N

C. Pure ALOHA

D. Polling

Correct Answer: B. Go-Back-N

Question 5

What is a principal weakness of Stop-and-Wait flow control on a long-delay link?

A. It cannot use acknowledgements

B. It always causes duplicate addresses

C. The sender may remain idle while awaiting a response

D. It prevents error detection

Correct Answer: C. The sender may remain idle while awaiting a response

Question 6

Which technique is associated with traditional shared half-duplex Ethernet?

A. CSMA/CD

B. CSMA/CA

C. FDMA only

D. Token encryption

Correct Answer: A. CSMA/CD

Question 7

Why does Wi-Fi use collision avoidance rather than ordinary collision detection?

A. Wireless frames have no headers

B. A station generally cannot reliably detect a collision while transmitting

C. Wireless networks never experience interference

D. Acknowledgements are prohibited in Wi-Fi

Correct Answer: B. A station generally cannot reliably detect a collision while transmitting

Question 8

What does token possession normally grant in a token-passing network?

A. Permission to transmit

B. A new IP address

C. A larger CRC field

D. Exclusive access to all routers

Correct Answer: A. Permission to transmit

Question 9

A link has T_t = 2 ms and T_p = 20 ms. Using U ≈ T_t ÷ (T_t + 2T_p), what is the approximate stop-and-wait utilization?

A. 4.8%

B. 9.1%

C. 20%

D. 50%

Correct Answer: A. 4.8%

Question 10

Which field commonly provides frame-level error detection in Ethernet?

A. Destination IP address

B. Time-to-live field

C. Frame Check Sequence

D. Application port number

Correct Answer: C. Frame Check Sequence