Lan Ethernet Card Ring Network (2)

I. Token Bus

ARCNET is a local area network developed by datapoint in 1977. It is still widely used,
And uses RG-62 coaxial cable as the NetWare LAN cabling system. ARCNET, which is exactly the same as ibm3270
The cables connecting the terminal and the IBM host are the same, so this network is widely used in the Use base of a large number of IBM hosts.
ARCNET can now use twisted pair wires and optical fiber cables. The new arcnet plus Speed has increased from 2.5 MB/S
To 100 Mb/s (when using optical fiber). The media access method used by this network is the Token Bus.

2) NETWORK STRUCTURE OF ARCNET There are two kinds of ARCNET cabling: one is bus and the other is star bus. The two types are described below.
ARCNET.

1. Bus Structure

The bus structure is similar to that of Ethernet cables. As shown in figure 1, the maximum length of the ARCNET bus is 305 MB and can be connected.
A maximum of 8 devices. The connection between the device and the bus is through the T-connector. The top of the connector is connected to the cable, and the bottom is connected to the network
Card connection. both ends of the cable must end with a resistance of 93 Ω.

The above is the case where coaxial cables are used. If twisted pair wires are used, the above rules will change. In this case
It is constructed by means of a chrysanthemum. That is, the NIC suitable for twisted pair media has two ports: one for connecting to the server and the other for connecting to the server.
The next PC. A maximum of 10 PCs can be connected to such a level chain, and the maximum distance of the twisted pair cannot exceed 122 m (400 feet). First
Both the network adapter and the second network adapter must use the 93 Ω final adapter.

2. Star bus structure

This physical wiring is centered on the hub to form a multi-star interconnected structure, as shown in figure 2.
The bus is called because all workstations share public cables.

This structure can use cables or twisted pair wires. When using cables, the star bus splits signals by using the hub.
The hub can be either active or passive. If the active hub is used, the active cable is changed to one of its ports,
The maximum distance between the workstation and the active hub is 610 m (2000 feet ).
When twisted pair wires are used, the active hub is used. The Hub is connected to the file server, and the workstation is also directly connected to it.
In this way, the distance between the server or workstation and the active hub can be up to 1220 M (4000 feet ).
Must use the 93 Ω final adapter to connect to the port.

4) operations on ARCNET

The station transmission of arcnet lan is broadcast like a bus-type LAN, but the access to the bus is determined by the token.
The operating mechanism of this network is described. Assume that there are four nodes on a bus, and their addresses are 255, 25, and respectively.
When the network is started, the four workstations form a logical ring. Each station tracks two messages:
· Who is the successor;
· Who is the pioneer.
The two types of information are represented by the letter S (successor) and P (Pioneer). The successor of a workstation is defined as logic.
A station with a higher address on the ring; a pioneer is defined as a station with a lower address on the logical ring.
In ARCNET, station address 0 is used for broadcast addresses. Therefore, the minimum station address is 1 and the maximum station address is 255.
According to the regulations in the competition, the workstation address is 1 for the successor Station Address of 255, and the probe Station Address of 1 is 255. Workstation
The addresses of the pioneers and successors are shown in table 3.

　Table 3 Origin Site and origin address of the workstation

Workstation address	Precursor workstation address (P)	Next workstation address (s)
1	255	10
10	1	25
25	10	255
255	25	1

1. ARCNET Frame Structure
Like Ethernet, ARCNET transmission is also called frame. Frame Structure 3. Figure 3 (a) shows the invitation to send
(ITT) the token frame is always passed to its successor workstation.
　　　　
(A) ITT Frame

Alert

EOT

Did

Did

　　　　
　　　　
(B) FBE Frame

Alert

Enq

Did

Did

Alert

ACK

Alert

Nak

Alert

Soh

Sid

Did

Did

CP

Data

CRC

CRC

No matter which frame the ARCNET frame is guided by alert, it is similar to the forward code used in Ethernet (see this article this year
Article 3). Alert is composed of a 6-bit interval signal (1). It consists of a forward pulse followed by a forward pulse
A dual-pulse representation composed of a negative pulse. The empty number (0) is expressed by a non-pulse representation. The EOT is the transmission end controller (04hex) in the ASCII code ).
The second two bytes are did (endpoint identifier), that is, the information of the subsequent workstation. The purpose of reusing did is to increase
Reliability.
Figure 3 (B) is the idle buffer query (FBE) frame. Enq is the query character (05hex) in the ASCII character set.
Two bytes of did are wks that want to know the idle buffer status through inquiry.
Find the reliability of the end workstation.
Ack (confirmation) frames are composed of alert and ACK. Alert is described earlier. Ack is true in the ASCII character set.
Recognize character (06hex). When Ack is sent in response to the FBE frame, it indicates that the receiving workstation has available buffer space.
Ack frames do not have the did field because such frames are sent as broadcast frames. ACK frame 3 (c. Nak
(Denied) frame 3 (d). The Nak is a denial character (15hex) in the ASCII character set. When the FBE frame is returned, the Nak is sent
The receiving workstation does not have the available buffer space. The nak frame does not have the did field, and the reason is similar to the ACK frame.
Same.
Figure 3 (e) shows the data frame. In a frame, Soh (Title start) is the title start character (01hex) in the ASCII character set ).
Sid (Source Vertex ID) and (end vertex ID) indicate the address of the Source Vertex and the end workstation. CP (continuous pointer) field indicates that the workstation is stored in
The start point of Data Transmission found in the storage device. The data field data has a variable length, which is between 1 byte and 508 byte.
This field is added by the sender to carry the user data. 2 bytes CRC field, which is used to protect the data field.
2. Operation Process
At startup, the logic sequence must be formed first, that is, the logical ring. Each station keeps track of and maintains its precursor workstation and
The station ID of the subsequent workstation. The Regulations on the frontend and successor workstation are shown in table 3. Each workstation has its own successor
(NID) is set to the station address (ID) plus 1, and set the timeout value according to the following formula
(Timeoutt ):
Timeout = 146x (255-ID) μs
A workstation with the maximum address value first times out, so it creates an ITT frame and sends the token frame to its successor.
If no response is received after 74us, the workstation with the maximum address value considers that the station with the next NID address does not exist.
Increase the NID value by 1 and send the did again as the new value of ITT. This process is repeated until the workstation with the maximum address value finds
The subsequent workstation found repeat this process like the previous workstation.
Once all the active workstations are found, the normal token transfer operation can start. Configure the time range from 24 to 61us, take
Depends on the number of active sites and the workstation address value. To make the initial timeout value 0 and minimize the configuration time, at least
Set a workstation address in ARCNET to 255.
A workstation with ITT frame can send up to one frame before passing the token frame to the successor station.
Before clicking a node, you must check whether there is sufficient buffer space to receive frames. FBE frame is used to perform this query.
If a buffer is available on the site, the ACK frame is sent back; otherwise, the Nak frame is sent back.
Once an ACK frame is received after the FBE frame is sent, the data frame PAC can be sent.
If the token is correctly transmitted because of a fault, the network must be reconfigured. Generate another type of Reconfiguration
The problem is to add a workstation or remove the workstation on the token transmission ring. Therefore, reconfiguration is hard to avoid.
If an active workstation does not receive the ITT frame after 840ms, the recon pattern consists of eight signal transmission intervals,
Sent 765 times after an empty number. The recon pattern continues for 2754us to ensure that any token frame in the transmission is damaged, and
If the token frame is lost and 78us is not active, all workstation will realize that reconfiguration is happening.
Set its own successor to its own address (ID) plus 1, and set the timeout value. Later, the process will be the same as that at startup.
In ARCNET technology, deleting a workstation is a simple process and you do not need to call all reconfiguration mechanisms. For example:
If the workstation with the address 10 has been removed from the ring, and does not respond to the ITT frame sent from its precursor workstation 1
Workstation 10 no longer exists when workstation 1 exceeds 74us. workstation 1 increases its NID value by 1 (new value is 11) and
Send ITT to workstation 11. If there is still no response after 74us, repeat the above process. The next station address is 25.
Station 1 requires (25-10) x74us = 1.1ms to discover that its successor workstation is 25.
If workstation 10 wants to re-enter the ring, it must wait for the token for 840ms. If it has not passed through the ITT frame to be
Send the invitation. It must call the all Reconfiguration Mechanism.

5) ARCNET plus

ARCNET plus is the second generation product of ARCNET, which significantly improves the performance of ARCNET.
Up to 8 times of the original 2.5 Mbit/s, that is, 20 Mbit/s, and it is also backward compatible with ARCNET.
To understand the mechanism for improving performance, we must be clear about the reasons for the low delivery rate of ARCNET. Use of ARCNET Workstation
The logical "1". The logical "0" is composed of two static intervals. One
The interval is 1/5 MHz = NS. It can be seen that sending 1-bit (0 or 1) information requires two such intervals.
ARCNET sends data in multiple integer bytes. Each byte uses a 3-bit pattern (110) to enable the receiver and the sender
This indicates that each 8 data bits has an additional overhead of 3 bits. Therefore, the data rate of ARCNET is 8/11.
* 25 Mb/s = 1.82 Mb/s. Obviously, there is a large waste of bandwidth. The static period is a waste, and more accurate additional overhead occupies the total bandwidth.
The width is 27%.
ARCNET plus increases bandwidth utilization efficiency through the following two channels:
· Elimination of static periods;
· Change the bitwise of each byte to the bitwise of every eight bytes.
The sophisticated technology used by ARCNET plus is to use amplitude modulation to squeeze out 4 bits of information at every 200ns interval.
A positive or negative sine wave has eight possible ranges from 0 to 12 V, and a pulse group with a total number of 2x8 = 16 is given.
4-bit data.
Therefore, the data rate of ARCNET plus is 4 bits × 5 Mbps pulse/second = 20 (MB/s). However, this data rate is not included
Additional Overhead: if more accurate overhead is used, the effective data rate is 16.8 Mb/s.
During ARCNET plus initialization, a workstation sends a special signal to notify other sites to use arcnetplus
When the workstation transmits the token, it also sends this special signal. When the ARCNET plus workstation and
When the ARCNET workstation communicates, it is reduced to 2.5 Mb/s.

Click to open the link for LAN and repeater, router, bridge-related just see http://www.he.edu.cn/learn/wlxy/netlan/index.html