Review Questions

1.        Which drawbacks of CSMA/CD are eliminated by demand priority?

Very long delays (theoretically they could be infinitely  long)  when accessing a shared medium.           

2.        Why did the developers of Fast Ethernet decide to preserve CSMA/CD?

3.        Which topologies are supported by a Fast Ethernet network based on a
shared medium?

 Star and tree

4.        What is the maximum diameter of a Fast Ethernet network?

5.        How many pairs of cable are used for data transmission in 100Base-T4?

Four pairs

6.        What are the differences between Class I and Class II Fast Ethernet repeaters?

Class I repeater supports different types of logical encoding on its ports, translating between 4B/5B (100Base-TX/FX ports)    and 8B/6T (100Base-T4 ports) codes. Class II repeater supports only one type logical encoding on all its ports, either 4B/5B or 8B/6T because such a repeater can not translate codes.  

 

7.        Why is only one Class I repeater allowed in a Fast Ethernet network?

Because class I repeater introduce quite a big delay when transferring bits between its ports so that a presence of two class I repeaters would cause an unstable collision detection.

8.        What is the minimum value of the interpacket gap in Gigabit Ethernet?

96 ns

9.        Because of bandwidth increase, the developers of Gigabit Ethernet had to increase the minimum frame size to 512 bytes. When the data being transmitted cannot fill the data field of the frame, it is complemented to the required length by padding, which does not carry any useful information. What steps have been taken in Gigabit Ethernet to reduce the overhead for short data transmission?

It is allowed to send a sequence of short frames without interpacket gaps provided that their total length is not grater than 8192 bytes.  

 

10.     What steps have been taken in Gigabit Ethernet to ensure data transmission at 1,000 Mbps using twisted pair?

A.      Increased quality of twisted pair cable

B.       Use of four pairs of the cable instead of two - correct

C.       Increased the number of states of the signal code - correct

D.      Implemented Quadrature Amplitude Modulation

11.     Why does Gigabit Ethernet use both multimode and single-mode optical fiber?

A multimode optical fiber is used for distances less than 500 m; for longer distances a single-mode optical fiber should be used as a multimode fiber has too narrow bandwidth for 1000 Mbps data transmission speed signal spectrum.     

 

Problems

1.        Task: Using Tables 13.3 and 13.4, determine what stability reserve a Fast Ethernet configuration with one Class I repeater has.

Tip

When determining the correctness of a Fast Ethernet network, instead of using the one or two hubs rule, it is possible to calculate the PDV, as was done in the "Case Study" in Chapter 12 for the Transmash network.

As for Ethernet 10 Mbps, the Fast Ethernet standard provides source data for calculating the signal PDV. However, the form of this data and the method of calculation have changed. Fast Ethernet provides data on doubled delays introduced by each network segment, without dividing network segments into left, intermediate, and right. In addition, delays introduced by network adapters take into account frame preambles. Therefore, PDV must be compared to 512 bt (i.e., to the time required to transmit a minimum-length frame without a preamble).

For Class I repeaters the RTT can be calculated as follows: Delays introduced during signal transmission using the cable are calculated based on the data provided in Table 13.3, which takes into account that the signal has to pass through the cable twice.

 

Table 13.3. Delays introduced by the cable

Cable type

Doubled delay (bit interval per 1 m)

Doubled delay on the cable of maximum length

UTP Cat 3

1.14 bt

114 bt (100 m)

UTP Cat 4

1.14 bt

114 bt (100 m)

UTP Cat 5

1.112 bt

111.2 bt (100 m)

STP

1.112 bt

111.2 bt (100 m)

Optical fiber

1.0 bt

412 bt (412 m)

Delays introduced by two network adapters (or ports of the switches) interacting through a repeater are shown in Table 13.4.

 

Table 13.4. Delays introduced by network adapters

Network adapter type

Maximum delay for the round trip

Two TX/FX adapters

100 bt

Two T4 adapters

138 bt

One TX/FX adapter and one T4 adapter

127 bt

Taking into account that the doubled delay introduced by a Class I repeater is 140 bt, it is possible to calculate RTT for any network configuration, (accounting for the maximum lengths of cable segments, provided in Table 13.2). If the resulting value is smaller than 512, then according to the collision detection criterion, this network configuration is correct. The 802.3 standard recommends ensuring a reserve of 4 bt for stable network operation. However, it allows choosing a value between 0 and 5 bt.