Download Congestion control mechanisms in network layer and more Essays (university) Computer Networks in PDF only on Docsity!
Congestion Control algorithms :
Congestion is the state in which network performance decreases. This happens when a n/w is holding too many which are much more than the n/w’s capability.
There are various reasons for congestion to occur. Few of them are listed below.
- If the traffic on the n/w is very high.
- If the CPU’s processing speed is slow.
- If there is insufficient memory to hold the packets.
When too many packets are present in the n/w, the performance will be decreased, this situation is called as congestion.
Congestion can be caused by more and more number of packets from different input line demanding the same output line, less memory for queuing these packets at the router and slow processors at the router as well. Increase in memory may help to control congestion.
General principles of congestion control:
Congestion control is concerned about controlling traffic entry in to a n/w so as to avoid congestion by taking resource reducing steps such as reducing the rate of sending packets.
Congestion control strategy is divided into 2 types.
- Open loop congestion control
- Closed loop congestion control
Open loop congestion control: it is used for preventing congestion from happening to occur on the first place.
Various open loop congestion control techniques are as follows:
- Retransmission policy
- window policy
- Acknowledgement policy
- Discarding policy
- Admission policy
Retransmission policy : Retransmission can be main cause for increasing the congestion. By using effective retransmission policy and timer, congestion can be controlled.
Window policy : Before the transmission starts we have to specify the receiver’s window size.
Acknowledgement policy : this policy specifies that the receiver should not acknowledge each and every packet that it receives due to which the load on the network is very much reduced.
Discarding Policy : To prevent congestion routers have to adopt good discarding policy. Hence this will reduce the integrity of transmission.
Admission policy : According to this policy, new virtual connection request is not accepted if that request leads to congestion.
Pitfall : the drawback of open loop congestion control is that once the system has started running in between, connection cannot be adjusted.
Closed loop congestion control : it is used to remove the congestion, after it occurred. These congestion control strategies are based on feedback mechanism.
Different kinds of closed loop congestion control techniques are as follows:
- Back pressure.
- Choke packet.
- Implicit congestion.
- Explicit congestion.
Back Pressure : in this the congested node will stop accepting data from its previous node, due to which the node will be congested and which in turn stop accepting data from it’s previous node. This technique is used only in connection oriented n/w’s.
Choke Packets : it is a packet created by the congested node that warns the source about the congestion.
Implicit congestion : in this the sources automatically learn about congestion by observing the delay in packets acknowledgement, time out condition and retransmission of packet.
▲ Router continues to set the warning bit till it remains in the warning state.
▲ Source monitors each acknowledgement packet for the warning bit and if it set then it cuts down the traffic accordingly. In this way all the sources that get acknowledgements with warning bit set, reduces the data rate helping the congestion to overcome.
▲ Traffic is increased when no router in the path is in trouble.
▲ Choke packets : This method is faster one because it sends a special packet(known as choke packet)to the source host directly indicating the state of congestion. The steps are given below.
▲ The router sends choke packet back to the source host directly, mentioning the destination found in the packet, to which the source was sending packet.
▲ The original packet is tagged so that the next routers will not generate any more choke packet while the packets are travelling further.
▲ Source host after getting the choke packet reduces the traffic sent to the specified destination.
▲ Since the information of congestion through choke packets comes to the sender consumes sometime and mean while some of the packets with same destination are already in the line, they will also generate choke packets. Source will ignore these choke packets till certain interval of time. After that it will again listen for any new choke packet.
▲ Hop-by-hop packets : it is better modification of original choke packet technique. In this technique while the choke packets traverse back towards the source it is received
By each of the intermediate routers and immediate steps are taken by them to reduce load of the network. The steps are given below
▲ The router sends choke packet back to the source host.
▲ The choke packet is received by the previous router and it gets information of congestion.
▲ The previous router sends the choke packet back and at the same time it takes necessary steps to reduce the data rate.
▲ All the routers getting choke packets reduce their data rate immediately without waiting for the original source to reduce the data rate.
▲ Load shedding : it is the effective mechanism for controlling the load. A router which is overloaded by packets that it cannot handle just throws them away. In selecting packets for throwing it may apply following techniques.
▲ Random selection, in which the packets are randomly selected.
▲ Depending on the application old packets can be thrown away or new packets can be thrown away.
▲ Priority can be marked in packets by the source to indicate how important the packet is, Depending on this, the packet may or may not be selected for throwing away.
▲ Random Early Detection (RED) : The basic idea is to drop packet before situation becomes hopeless. Router maintains an average queue length for each of the outgoing lines. When average queue length exceeds a threshold the line is said to be congested. The router then drops packet at random from the queue and does not report it to the source. Source will notice lack of acknowledgement and will slow down transmission rate assuming that lost packet generally occur to the congestion and discard.
Leaky bucket algorithm:
Consider a bucket with a hole at the bottom. Whatever be the rate at which water enters into the bucket it leaks out at constant rate through the hole. The rate of flow is zero if there is no water in the bucket and if the bucket is full the excess water spills over and is lost.
Token bucket algorithm incorporated the feature of handling bursty traffic.
Features :
▲ It allows hosts to send bursty traffic when required, up to the maximum size of the bucket say n. This implies that burst of up to n packets can be sent at once. Allowing some burstyness in the o/p stream and giving faster response to sudden burst of inputs.
▲ It throws away all the tokens up to the transmission capacity. When the bucket is filled and still new burst of packet is coming from transmission, but it never discards packets.
▲ Byte version of this algorithm is also available.
- Each token is allowed to send K bytes.
- A full packet can only be sent if there are enough tokens are available to cover the whole packet of the length.
- Fractional tokens- kept for future use.
▲ Sometimes a leaky bucket is used after token bucket for smoother traffic. The rate of leaky bucket is higher than ‘r’ and lesser than ‘n’ of token bucket.
Algorithm :
A variable counts token.
▲ Counter incremented by one in every DT second.
▲ Counter is decremented when packet is sent.
▲ If counter=0 then no packet is sent.
▲ In case of byte count version
- Counter is incremented by K bytes in every DT second.
- Counter is decremented by the length of the packet sent.
- If counter=0 then no packet is sent.
▲ Length of burst is a design issue.
IPV4 :
- Version: Version no. of Internet Protocol used (e.g. IPv4).
- IHL: Internet Header Length; Length of entire IP header.
- DSCP: Differentiated Services Code Point; this is Type of Service.
- ECN: Explicit Congestion Notification; It carries information about the congestion seen in the route.
- Total Length: Length of entire IP Packet (including IP header and IP Payload).
- Identification: If IP packet is fragmented during the transmission, all the fragments contain same identification number. to identify original IP packet they belong to.
- Flags: As required by the network resources, if IP Packet is too large to handle, these ‘flags’ tells if they can be fragmented or not. In this 3-bit flag, the MSB is always set to ‘0’.
- Fragment Offset: This offset tells the exact position of the fragment in the original IP Packet.
- Time to Live: To avoid looping in the network, every packet is sent with some TTL value set, which tells the network how many routers (hops) this packet can cross. At each hop, its value is decremented by one and when the value reaches zero, the packet is discarded.
- Protocol: Tells the Network layer at the destination host, to which Protocol this packet belongs to, i.e. the next level Protocol. For example protocol number of ICMP is 1, TCP is 6 and UDP is 17.
- Header Checksum: This field is used to keep checksum value of entire header which is then used to check if the packet is received error-free.
- Source Address: 32-bit address of the Sender (or source) of the packet.
- Destination Address: 32-bit address of the Receiver (or destination) of the packet.
- Options: This is optional field, which is used if the value of IHL is greater than 5. These options may contain values for options such as Security, Record Route, Time Stamp, etc.
IPV6 :
IPv6 fixed header is 40 bytes long and contains the following information.
▲ Version ( 4-bits): It represents the version of Internet Protocol.
Types of Bridges:
Three types of Bridges are used in network.
▲ Transparent Bridge (or) Spanning tree Bridges ▲ Translational Bridge. ▲ Source-route Bridge.
Transparent Bridge: Transparent Bridges is invisible to the other devices on the network. Transparent Bridge only performs the function of blocking or forwarding data based on MAC address. MAC address may also be referred as hardware address or physical address. These addresses are used to built tables and make decision regarding whether a frame should be forward and where it should be forwarded. Translational Bridge: Translational Bridges are useful to connect segments running at different speeds or using different protocols such as token Ring and Ethernet networks. Depending on the direction of travel, a Translational Bridge can add or remove information and fields from frame as needed. Source-route Bridge: Source-route Bridges were designed by IBM for use on Token ring networks. The source routing Bridge derives the entire route of the frame embedded within the frame. This allows the Bridge to make specific decision about how the frame should be forwarded through the network.
Transparent Bridge vs. Source Routing Bridge:
Issue Trasparent Bridge Source Routing Bridge Orientation It provides connectionless networking. Its frames take independent routes to reach the destination.
It provides connection oriented networking. It uses a discovery grame to findout it’s route and then this route is followed next time. Transparency These are completly transparent ans hidden hosts.
These bridges are not transparent, as hosts are aware of bridging strategy. Configuration It is automatic. it doesn’t require network management. automatically the bridges configure themselves to the topology.
It is manual. In this, the duty of the n/w manager to install the LAN and BRIDGE numbers. It is difficult to detect mistakes from loop of frames. Routing Theses bridges uses sub optimal routing.
Uses optimal routing.
Locating Backward learning is used to find the destination, the drawback is it makes the bridge to wait for a frame from a specific node or host to come across about that host.
Discovery frames are used to find the destination. The drawback is that exponential explosion is moderate to large for internetworks.
Failure It automatically handled by the bridge with out noticing to the host.
These are handled by hosts. Bridges give failure information notice to hosts, but hosts unaware of the problem they don’t know if the problem is with the current route, until another discovery frame comes to the hosts. Complexity These bridges put extra complexity. As the bridge automatically configure themslves to the topology.
These routing bridge puts extra complexity in the host,as they have to send discovery frames to find the destination.
