Practical/DataLink Flashcards
What is the physical layer concerned with?
Concerned with the formation of bits to transmit and receive them between adjacent devices.
What does the physical layer offer in terms of reliability?
Unreliable bit pipe service.
What do the following stand for?
BER
NEXT
Bit Error Rate
Near-End Cross-talk
Explain cross-talk noise.
A signal on one line is picked up by adjacent lines as a small noise signal. Particularly troublesome in NEXT when a strong transmitter output signal interferes with a much weaker incoming receiver signal.
Explain impulse noise.
Caused by external activity or equipment. Electrical impulses on the line which cause large signal distortion for their duration.
Explain Thermal Noise. Give common name.
Caused by the thermal agitation of electrons in the device or transmission line material. Always present.
White noise.
How can Thermal noise be reduced?
Freezing components will reduce.
How does digital transmission combat noise?
All data transmitted as very different values of voltage for either zero or one. Noise would have to be extremely drastic to drop voltage from one to zero. Decision isn’t made if voltage drops to an in-between value.
Give the average probability of bit error for each of the following.
- Fibre optic cables
- Coaxial cables
- Radio
- 10^-9
- 10^-6
- 10^-3 to 10^-6
Distinguish between p and p sub b. What is L in the formula?
p sub b is probability of bit error at physical layer.
p is probability of frame error at datalink layer.
The number of bits in the frame.
Give the formula for the probability of frame error.
Give the approximation formula and the condition.
p = 1 - (1 - p sub b)^L
When p «_space;1%
p = L * p sub b
Explain different type of transmission modes for physical layer.
Simplex: Only one way ever, no return path.
Half-duplex: Can go either way but not at the same time.
Full duplex: Can use both ways at the same time.
What is multiplexing in the physical layer? Name three types.
Sharing a physical link among multiple transmissions.
FDM - frequency division multiplexing
TDM - Time Division Multiplexing
Statistical Multiplexing
Explain FDM.
Each user only utilizes a portion of the bandwidth but can have it for the full time. Signals are multiplexed to different frequencies.
Explain TDM.
Each user gets to utilize the full available bandwidth but can only use it for allocated time slots. NO CROSSTALK as only one transmission at a time.
Explain statistical multiplexing.
Both FDM and TDM divide the link into independent channels (inefficient if traffic is bursty) but in statistical multiplexing the link will never be idle.
T or F. TDM is better for bursty traffic than statistical multiplexing.
False.
What is the disadvantage of statistical multiplexing over TDM and FDM?
- Frames have an additional overhead to indicate which input stream they belong to
- Packet queuing delay is variable.
What does it mean for a device to be “adjacent”?
Means physically connected by a communication channel, assumed to deliver bits in the right order by physical layer.
List three functions of datalink layer.
- Determining how bits in physical layer are framed.
- Dealing with transmission errors
- Controlling the flow of frames so that receiver is not overwhelmed by sender.
What is a codeword?
Is a frame consisting of data bits and redundancy bits (for error detection)
Why is a single parity check bad?
Can’t tell how many errors
Where those bits are
Can’t detect even number of bit errors.
Why is a 2-dimensional parity check bad?
Can correct <= 3 errors
Can only correct any single bit
What is the Hamming Distance?
Is the number of bit positions that differ between two codewords W1 and W2.
What is the minimum hamming distance?
The smallest number of bits to change one valid codeword into another.
General rule for n bit parity check.
Can Detect <= n - 1 errors
Can correct n/2 errors
What operation is used in a 7-4 hamming code?
Exclusive OR
What does CRC stand for?
CRC - Cyclic Redundancy Check
How does data link layer error check?
Adds redundancy bits that check the accuracy of the transmission.
Formula for CRC
Divide Transmitted data (add on n zeros where n is length of CRC) by G a known value. The remainder should be the CRC added on. Output is what is transmitted.
How does CRC known if there is no error?
Remainder is 0.
T or F. IPv6 supports flows unlike IPv4 that does not.
True.
T or F. IPv6 is not compatible with TCP or UDP but is compatible with FTTP.
False.
T or F. Modern Communication Systems use Digital Transmission.
True.
T or F. Subnetting and CIDR are essentially the same as they both use netmasks.
False.
T or F. IPv6 is compatible with IPv4.
False.
T or F. The Physical Layer is concerned with the bit transmission.
True.
T or F. IPv6 uses 128 bit addresses.
True
T or F. There is always noise on any transmission medium.
True.
T or F. Statistical Multiplexing requires additional overhead to indicate the stream.
True.
T or F. Crosstalk is caused by lightning strikes.
False.
T or F. Many frames will have the same redundant data sequence used for error detection and correction.
True.
T or F. Frequency Division multiplexing and time division multiplexing are similar in that they split the link into independent parts.
True.
T or F. A two dimensional parity scheme will always detect the error even though sometimes it can not correct it.
False.
T or F. A two dimensional parity checking scheme will detect all even number of errors.
False.
T or F. Sending multiple copies of the same frame, and using them to decide if the frame is correct, will reduce the overall probability of frame error p.
True.
T or F. If I cut the frame size by half then the resulting probability of error of the frame is also halved.
False.
T or F. Half Duplex can send data in both directions.
True.
T or F. Thermal noise is reduced by heating up the receiver.
False.
T or F. The probability of frame error is always bigger than the probability of bit error (assuming more than 1 bit in a frame)
True.
T or F. Time Division Multiplexing (TDM) gives the whole of the resource for some of the time.
True.
T or F. Impulse noise is caused by a strong signal being close to a weak signal.
False.
T or F. Frequency Division Multiplexing (FDM) gives the whole of the resource for some of the time.
False.
T or F. For the same probability of bit error, if I increase the size of the frame, then the probability of frame error goes up.
True.
T or F. The physical layer delivers a reliable bit pipe service.
False.
T or F. The six bit generator polynomial has a CRC that is six bits long.
False.
T or F. Larger Frames have a higher probability of error p than smaller frames with the same probability of bit error (BER) pb.
True.
T or F. When I divide the incoming bit sequence by the Generator Polynomial and there is no remainder, then I assume the Frame has not been corrupted.
True.
T or F. The sixteen bit Generator Polynomial has a CRC that is fifteen bits long.
True.
T or F. The Datalink Layer is only concerned with the framing of bits and transmission errors.
False.
T or F. A two dimensional parity bit scheme can detect up to three bits in error.
True.
T or F. If there are five 1’s in a row followed by a 0 there is no need to use bit stuffing.
False.
T or F. It is not possible to have an error in a Frame when the remainder after dividing in the Generator Polynomial into the received data is zero.
False.
Should parity bits be included when saying what was the data that was sent?
No.
T or F. A Stop-And-Wait with the value of “a” of 3 can achieve an efficiency of 50% if the probability of bit error is small enough.
False
T or F. The only way to reduce thermal noise is to cool the receiver.
True.
T or F. A domain name can be resolved to many MAC addresses by the use of DNS and ARP.
True.
T or F. An autonomous system with a subnet mask of 255.255.240.0 has more hosts in each of the subnets than one with a subnet mask of 255.255.248.0
True.
T or F. IP address 137.84.211.43/21 and 137.84.219.143/21 are in the same subnet.
False.
T or F. The peak of aggregates can be a lot smaller than the aggregate of peaks.
True.
T or F. Datagram packet switching can lead to packets having to be re-ordered at the receiver.
True.
T or F. IPv6 is compatible with IPv4.
Fale.
T or F. There can be queueing at an intermediate node even if the following packet is a lot larger than the first packet being sent.
True
T or F. In TCP you cannot change the Receiver’s Advertised Window after a timeout.
False
T or F and explain. A 16 bit CRC will use a 16 bit known Generator that is used as both the sender and receiver.
False.
Generator must be 17. (n + 1)
T or F. To prevent the flags appearing in the data using bit stuffing, it is not necessary to place a 0 after five 1’s if there was a 0 next.
False.
T or F. Time division multiplexing allows each user to get all of the bandwidth some of the time
True.
T or F. A Hamming code can detect a single bit error but can not correct it.
False.
T or F. In TCP congestion control, the threshold has to be less than the Receiver’s Advertised Window at all times.
False.
T or F. A TCP port and IP address can only be used for one connection at a time.
False.
T or F. When a Network Layer PCI is added to the Transport Layer PDU this results in a Network Layer PDU.
True.
T or F. Statistically multiplexing uses all the slots if any of the inputs want to send data.
True.
T or F. Frequency Division Multiplexing gives you all of the bandwidth some of the time.
False.
T or F. When the remainder is zero at the receiver after the data has been divided by the Generator, then there has been no error in the data that was sent.
True.
