A token is passed from station to station in sequence. If any station wants to send data, first it will grab the token then transmit the data. Data passes through all stations; only the intended receiver gets a copy of data. Finally the original data is received at the sender and then the sender release the token.Here there is no collision as it using token method.If multiple stations attempts to get the token simultaneously the one which high priority will get the token.
1. Delayed Token Reinsertion
Token is released only after getting the entire data packet back at the sender
Efficiency is Low
Reliability is high
Useful under low load conditions.
2. Early Token Reinsertion
Token is released immediately after the transmission of data packet
Efficiency is High
Reliability is Low
Useful under high load conditions.
Any station can become a monitor station.If multiple stations are ready to become monitor station then station with high priority will become the monitor station.The monitor station will check the availability of token in the ring .If the token is missed or vanished the monitor station will issue a new token.
The monitor station makes identification on data packet as it passes through the monitor station.So it will not allow the same packet to pass through the monitor station again.This solves the orphan packet problem.
The monitor station checks the data packet for all the fields.If it finds any problem the data packet, then the monitor station will not allow the packet to continue in the ring.
The monitor station has to send message to all other stations in regular intervals indicating that the monitor is active.
If this message is not received a new monitor is selected.
Malfunction of Monitor Station:
Monitor Station is alive, but is unable to perform its functions.
Issues in Token Ring:
1. Token Problems
Corrupted or unhealthy token
2. Source Problem
3. Monopoly of High priority stations over the low priority ones.
4. Destination Problem : If the destination is busy or crashed the data will not be get delivered.
Destination Busy: If the destination is busy then sender retransmits the packet.
Destination Crashed: If the destination is crashed then sender does not retransmit the packet.
5. Ring Problem: Any problems in the ring affect the physical connection.
Solutions to issues in Token ring.
1. Token Holding Time (THT):
No station can hold the token beyond the maximum Token Holding Time.This avoids the monopoly of high priority stations over low priority ones.
2. Monitor Stations:
Monitor station controls the activities in the ring - Token Management, Packet Delivery,Error Detection
- Circuit Switching
- Message Switching
- Packet Switching
It creates a direct physical connection between two devices.Circuit Switching reserves bandwidth in advance.Unused bandwidth is wasted in Circuit Switching.Each Packet follows the same Path.
Example: Telephone System
Here no physical connection is established.The message to be sent is stored in the first switch and checks for errors and the retransmitted (store and forward).This requires enough disk space to store the message in each switch and is very costly.The solution is dividing the message into different packets.
The message is divided into smaller packets and also no need to establish physical path in advance. It acquires bandwidth when needed and release soon after use. So no wastage of bandwidth.
- Virtual Circuit
Each Packet is treated independently and can go through any path.The datagram arrives at the destination at any order.
It is the responsibility of the transport layer to reorder the packet before sending to the destination port.Datagram is a connection less network service
Here a route is established before packets are sent. All Packets follows the same virtual path. But it is not a dedicated path as in circuit switching. Virtual Circuit is a connection oriented network service.
Disadvantages of Ethernet in Computer Networks
1. As the load on Ethernet increases number of collision increases, therefore efficiency decreases.
2. Ethernet offers non-deterministic service, so it is not suitable for real time application.
3. No priority mechanism in Ethernet.
4. There is a restriction of minimum size of packet.
What is SubNetting?
The 32 bit IP address is divided into two parts as Network part and Host part. That is the netid indicate the network part and the host id indicate the host part.To reach a host on the network we must first reach the network by using the netid. Then we must reach the host using the hostid.
- Network id
- Host id
A network can be divided into several subnetworks (subnets).This is achieved by taking some bits form the host part.
For example consider a university where the hosts are grouped according to department. University network has a specific networkid. Inside this network, we can create different subnetworks to identify various departments and each department has a subnetid.
Thus the 32 bit IP address is divided into three parts, namely network id, subnet id, host id.The bits from the host part is borrowed to generate the subnetid.
- Network id
- Subnet id
- Host id
Now to reach a host on the network we must first reach the network by using the netid, then to the subnetwork using subnetid and finally to the host using the host id.
Borrowing n bits from host part yields 2n – 2 subnets.
Leaving n bits in host part yields 2n – 2 hosts.
Class C address is divided into network and host portions as follows:
N . N . N . H
For a class C, we can borrow from 2 to 6 bits. But not possible to borrow 1 bit or 7 bits from host part.
Why not borrowing 1 bit or 7 bits from host part in subneeting a class C IP addresss.
When borrowing 1 bit , the number of subnet possible= 2n - 2=21 - 2 = 0
When borrowing 7 bits, the number of hosts possible on each subnet = 2n - 2=21 - 2 = 0