Case Study 4
Case Study 4
Case Study 4
QUESTIONS
1. Calculate the TRIB for the 28,800-bps circuits in the NDAS network shown in Figure 4–
16. Assume an average of 1,600 characters per message, with a 1 percent probability of an
erroneous transmission. The modems transmit with a 0.2-second turnaround delay. The
transmissions use an 8-bit ASCII code. The transmissions using 28,800-bps modems use 1
start bit and 1 stop bit, and each message block has 10 control characters, for further error
checking. Assume that the data is sent synchronously over the 28,800-bps circuits, using a
1,600-character block, but that each block has 55 control characters. Consider the
transmission load data for NDAS provided in the case study for Chapter 3. Are these TRIB
ratings adequate?
TRIB = Number of information bits accepted / Total time required to get the bits accepted
• K = 8-bit
• M = 1,600 characters block per message
• R = 28,800 bits/second (28,800 * 2 char/ 10bits (8 bits +1bit start + 1 bit stop) = 5,760
char/sec)
• C = 55 control characters per block
• P = 1%
• T = 0.2 second
TRIB = 8 bits/char * (1,600 – 55) char * (1 – 0.01) / (1,600 chars / 5,760 char/sec) + 0.2 sec
2. What file transfer protocols would you recommend? Be prepared to support your
recommendations.
I would recommend the synchronous transmission which is SDLC. Because it uses bit
stuffing to overcome the transparency problem. It can be used in WAN to connect remote
devices to mainframe computers at the main locations. Then it is also reliable because it transfers
data without errors. This is suitable for NDAS because it can be used in a remote site or local site
and its features contribute to the best network transmissions (transparency, flow control,
multipoint addressing, error detection and recovery).
Kung Thang Case study4 2/22/21
3. Prepare a brief position paper on the types of errors you can expect in the NDAS
network and the steps you believe NDAS can take to prevent, detect, and correct these
errors.
There are some errors such as white noises, impulse noise, crosstalk, echo, attenuation,
and intermodulation noise that can expect in the network. Those errors cannot be removed, but
can be prevented, detected, and corrected. Those errors happen because of different technical
glitches or configuration issues. For example- white noises happens because of a movement of
electrons or a sudden increase in electricity. It can be prevented by increasing the strength of the
electrical signal to suppresses the white noises. Then Crosstalk happens because of improper
multiplexing. It can be prevented by changing the frequencies or size of the guard bands. Echo
happens because of poor connection. It can be prevented by having a good connection.
Attenuation happens because of the loss of power in signal. It can be prevented by using
amplifiers that take the incoming signal, increase its strength, and retransmit it. The parity
checking, checksum, and cyclic redundancy checking methods can be used to detect those errors.
It can also correct those errors with retransmission which is most effective way if it fails to
prevent those errors.