Frame relay is a virtual wide area network based technology. It basically works on two layers of OSI model i.e. physical layer or data link layer by using a packet switching methodology. ISDN infrastructure is basically done by frame relay. Data is entered into the variable size chunks of units called frames. It has a very low chance of errors. When an error is detected in a frame, then it simply drops that frame. Frame relay provides both the virtual circuit and switched circuit.
- Frame relay operates at a speed of 1.544 Mbps to up to 44.376 Mbps.
- It operates on two layer of OSI model.
- Frame relay allows busty data and a frame size of 9000 bytes.
- It is less expensive as compared to others.
- There is an error detection mechanism at the data link layer of OSI model.
- There is no flow and error control mechanism.
- If a frame is damaged in any case then it is silently dropped because there is no transmission policy.
Permanent Virtual Circuit
There is a very simple connection in PVC. There is an outgoing DLCI is given to the source and another DLCI code is given to the destination. There are two drawbacks:
- It is very much expensive because there is two parties pay connection even when there is not in use.
- The connection is only from the single source to destination. If a single source wants a connection to may destinations then it can’t be possible.
Switched Virtual Circuit
It creates a temporary connection only when the data is to be transferred from the source to destination.
Frames Relay Layers
Frame relay operates on the two layers i.e. data link layer or physical layer.
Any protocol that is recognized by ANSI standards can be supported by physical layer.
Data Link Layer
It supports an error detection mechanism.
- Address DLCI Field
DLCI is an acronym for Data Link Connection Identifier. DLCI is a 6 bit of first fields and the other is of 4 bits.
- Command and Response
C/R specifies the upper bit which tells us that the frame is either a command or response.
- Extended Address
It specifies that the current is either the final byte of the address or not. If EA is ‘1’ that means it is a final bytes else there are some more bytes.
- Forward Explicit Congestion Notification
This is a 1 bit storage information bit stores at destination if congestion is occurred. The delays and losses of packets are known.
- Backward Explicit Congestion Notification
BECN is of 1 bit and tell sender that the congestion is occurred.
- Discard Eligibility
It indicates the priority level of the frame. If there is congestion in network then the DE discard the frame.
Routing is a process in which a path is selected for the traffic in network, between the multiple networks. It is performed for many types of circuits like packet switching, and circuit switched network. In routing, the routing tables are used to keep the records of the data that is transferred over network. A router is a device which establishes an intelligent connection between two devices but with different cables. There are various routing techniques:
- Mobile Routing
Now-a-days everyone has their portable computers and it is as much easy to work anywhere in the world. These computers are known as mobile hosts. To route a packet from mobile host firstly we have to find it.
The world is divided into different areas. Each area has one or more foreign agents. The foreign agents who keep track all the mobile users.
- When a packet is travel from mobile user, it first routed to the user’s LAN.
- The packets are intercepted by user agent.
- Home agent will then find the current location of the mobile’s user and then find its address corresponding to foreign agent.
- The home agents then encapsulated the packet and send it to the foreign agent.
- Hierarchical Routing
Hierarchical routing can be used in telephone networks. In hierarchical routing, the routers are divided into different regions. The routers knew each and everything about other routers. There are four levels of routing:
Two-level Hierarchical Routing
- Level-1: Regions
- Level-2: Clusters: it is basically a group of regions.
- Level-3: Zones: zone is a set of clusters.
- Level-4: groups: it contains May zones.
It is sufficient for a small size network. When a network containing 3 regions and a full routing table with 9 entries.
The destination 1A, 1B, 1C are lies in region 1. Whereas the 1B lies in region 2 and the rest is region 3.
Flooding is an algorithm in which every packet that is coming and sent it out to the outgoing but not from the path from which it can be arrived. There are various techniques which used in flooding such as: A hop counter is used. It keeps track of all the packets which was flooded.
Congestion is a technique or a mechanism by which congestion is prevented before and after its occurrence. We can divide the congestion control into two categories:
- Open Control
Open-loop is a kind of policy which is applied before the congestion is occurred. Congestion can be controlled at either side at the source or at destination.
- Retransmission Policy is applied when the sender knows that the sent packet is lost or corrupted then the packet needs to be retransmitted.
- Window policy is a GO-back-N window that is better, when the packet times out then it should be resent again.
- Acknowledgement policy is a policy in which the receiver acknowledges every time when it receives any packet.
- Discarding policy is a policy which prevents congestion and doesn’t harm its integrity.
- Admission policy is a quality of service which also helps in to prevent the congestion.
- Close control
When the congestion is already occurred into the network then the close control loop mechanism is applied.
- Backpressure is a close loop congestion control mechanism in which the congested node receives the data from the upstream node. It is node-to-node congestion.
- Choke packet is a technique in which a packet is send to the source, if congestion is occurred and the node is send with the help of an external router known as choke packet.