Data Frames and Errors in Networking

The physical layer is responsible for transmitting data bits, but it doesn't understand the meaning of these bits.

A data frame acts as a container for the message, like an envelope for a letter.

Adding a data frame to a message increases the number of bits transmitted because the frame itself contains additional information.

Both the sender and receiver of a data link frame must use the same protocol to ensure they understand each other.

The actual data to be transmitted is called the payload, which is contained within the data frame.

The transport layer header includes a port number, which is like a specific address within a receiver device.

The fields in a header vary depending on the protocol used to create it.

The MAC address, a unique identifier for a device on a network, is typically found in the data link layer header, not the transport layer.

The trailer part of a data link frame often contains error detection information to ensure the data has been transmitted correctly.

Large messages can be split into smaller packets, each with its own route across the network.

White noise, also known as thermal or Gaussian noise, can be reduced but not completely eliminated.

White noise is affected by the temperature of the transmission medium.

Impulse noise, also known as noise spikes, can be difficult to remove from analog signals without affecting the original signal.

Impulse noise in digital signals can often be corrected, allowing the original signal to be recovered.

Both analog and digital signals can be affected by noise.

Telephone crosstalk was a more common problem before fiber-optic cables and shielded wires became widely used.

Echo errors commonly occur at junctions where wires are connected.

Excessive jitter can't be corrected by speeding up transmission rates; instead, it requires addressing the underlying causes of timing irregularities.

Attenuation, the weakening of a signal over distance, can't be eliminated by using repeaters alone; repeaters only amplify existing signals.

Noise during transmission can slow down the transmission rate due to the need for error detection and correction.

Special filters for analog signals and digital signal regeneration can help in preventing white noise.

TCP, the transport layer protocol, performs error detection at the beginning and end of a connection.

Simple parity can only detect an even number of errors, not an odd number.

The CRC error-detection method treats the data packet as a large polynomial.

CRC catches nearly all errors, making it a highly reliable error-detection method.

Sending a message back to the transmitter was an early error-control technique.

NAK represents a negative acknowledgment, indicating that the data received was incorrect.

Sliding window protocols, which allow the transmission of multiple packets before waiting for acknowledgment, have been in use since the 1970s.

Adding a parity bit to ASCII data results in a Hamming distance of 3, not 4.

In a stop-and-wait error control system, a receiver can only acknowledge one packet at a time.

A sliding-window error control system allows a receiver to acknowledge multiple packets with a single ACK.

A packet sniffer is a tool used to capture network packets.

The data contained within a data link package is called the payload.

The physical layer of the OSI model is responsible for transmitting data bits, but it doesn't understand the meaning of those bits.

In the Ethernet header, the destination address is a MAC address, which is a unique identifier for a device on a network.

In the transport layer header, the destination address is a port number, which is a specific address within a receiver device.

White noise is also called thermal noise or Gaussian noise.

Impulse noise is also known as noise spikes.

Impulse noise is typically a burst of analog energy.

High humidity and wet weather can increase electrical crosstalk over a telephone system.

An echo suppressor is a device used to minimize the effect of echo.

Jitter is caused by small timing inconsistencies that become magnified during signal transmission.

Causes of jitter include electromagnetic interference, crosstalk, too many repeaters, and low-quality equipment.

Attenuation is not necessarily an error, but it can indirectly lead to more errors by weakening the transmitted signal.

Fallback negotiation allows the transmitting modem to slow down its transmission speed if data arrives garbled.

Crosstalk noise in twisted pair wires, such as Ethernet cables, can be electrical.

Proper shielding helps prevent crosstalk by isolating signals from each other.

Error detection can be performed in the network access/data link layer of the TCP/IP model.

Parity checks, a basic error-detection technique, are also used in data storage systems.

Even parity adds a bit to ensure an even number of binary 1s in a character.

Isolated single-bit errors occur 50 to 60 percent of the time.

Error bursts, where multiple bits are affected, occur 10 to 20 percent of the time.

Longitudinal parity, also called horizontal parity, is a technique that checks errors across multiple characters.

An arithmetic checksum adds the values of individual characters to be transmitted.

A generator polynomial is an industry-standard bit string used to create the CRC remainder.

Many wide area network protocols use CRC for error detection.

Fiber-optic cable is the medium least prone to errors due to its resistance to interference and attenuation.

Sending a message back to the transmitter is probably the most common form of error control.

Stop-and-wait error control is commonly associated with the Stop-and-wait flow control protocol.

Timeout refers to the amount of time a station waits before retransmitting a packet in Stop-and-wait error control.

A sliding window protocol allows the transmission of multiple packets before receiving acknowledgments.

In a sliding window protocol with a maximum window size of 7, the station can transmit 7 data packets before waiting for acknowledgment.

A sliding window protocol informs the transmitter what piece of data is expected next.

Forward error correction is the process where a receiver can fix errors using redundant information.

If a receiver correctly receives packets 2, 3, and 4, it will respond with an ACK for packet 5, indicating that it's ready for the next packet in the sequence.

If a receiver has just received a packet with bytes numbered 600 to 700, it will respond immediately with an ACK of 701, indicating that it's ready for the next packet starting at byte 701.