Question

Explain the issues related to congestion control and flow control in LANs. Be sure to identify any performance tradeoffs caused by the use of various networking components. Be sure to identify the basic layers/sublayers where management of congestion control and flow control for LANs typically resides.

Answer 1

Flow control is handled by a receiving side and it safeguards that a sender can only send something that a receiver can control or handle. Flow control handles a mechanism available to safeguard that communication flows smoothly. For example, a situation where somebody has a quick fiber connection could be sending on something or dialup similar. A sender could have the capability to send packets quickly, however that can be useless to a receiver on dialup. Thus they require a method to throttle the sending side might send. The issues related to flow control are receiver buffer can be full, design implementations of PAUSE Function, adding pause frame in transmit queues, etc.

Coming to congestion control, it is a technique which safeguards everything across the network which has a good amount of access to a network resource, within an allocated time. Suppose in a mixed-network environment, everybody requires to be capable to assume a similar general phase of performance. Like in office LAN, a number of segments are available where everyone is doing works within a LAN, though they can require to go out for a WAN link which is slower than a LAN segment. So some type of congestion control could require to be in place to safeguard that there are no issues within a greater network. Issues related to congestion control are the low loading, service policies, packet queuing, router queue, it is tough to get because the router can get several packets since several sources connected to a network.

One problem can be isolated production through monitoring traffic. Suppose there is a virtual environment where it is required to compute costs and input, output tradeoffs while creating copies of network traffic. There are two main steps in a design optimization procedure. One is performance evaluation on the basis of some parameters and another is performance tradeoffs or cost determination. By adding several components in the networks like routers, networking cables, gateways, bridges, hubs, switches, wireless access points, repeaters, etc. can create performance tradeoffs. Authoritative DNS servers can target DDoS attacks which represent some possible point of failure as DNS uses UDP. With a wire request format can be faster, though DNS allows compression of names. This is common in DNS, where a similar domain name appears more than one time, though proves fewer in requests. If wire-format hash keys are implemented, then we should need to perform a check for compression first. Then such requests are decompressed along with re-encoded to uncompressed a wire-format for hashing. There is a choice among optimization of a common case, with the help of wire-format hash keys, or optimization for worst cases. The format is faster, but later it can be robust for a DDoS attack which reflects performance tradeoffs.

The network layer provides routing and switching technologies, which make logical paths, called virtual circuits, to transmit data from one node to another. The routing functions of the network layer maintain or handle the congestion control, addressing, packet sequencing, and error handling. So congestion control is known as a network layer problem.

At the data link layer, the data packets are decoded and encoded into bits. It also handles errors, transmission protocol management and knowledge, frame synchronization, and flow control. The layer is also segregated into two sublayers: Logical link control layer, and the media access control layer. The LLC handles error checking, and flow control. When a data is transmitted over through a data transfer medium, the transport layer is responsible for the flow control in a network.