Reference no: EM13947804 , Length:

Q1: Reliable Data Transfer

‘Standing in the middle of the road is very dangerous; you get knocked down by the traffic from both sides.'
-Margaret Thatcher

Not all network traffic will reach its destination in a timely fashion; sometimes data becomes corrupted by noise, interference or by being forced to buffer behind competing traffic.

Fortunately, Reliable Data Transfers (RDT) ensure that parts of a message will eventually reach its destination undamaged. Your reference materials mention three kinds of RDT protocols (RDT1, RDT2 and RDT3) and their subsequent error detection methods. In this Discussion, you will discuss benefits of the RDT3 protocol vs. RDT1 and RDT2.

To complete this Discussion:

Post: Submit a 350-500 word initial post in which you compare the benefits of the RDT3 protocol vs. RDT1 and RDT2, evaluate and explain whether or not a timer is required for round-trip delay messages under protocol RDT3, and under what conditions, and summarise the criteria that determine whether or a packet is mission-critical or whether it can be lost when using RDT3.

Q2: TCP Congestion Control

When designing, implementing and enforcing flow on your network information highway, you needn't worry about handing out speeding tickets to non-law-abiding bytes, but you do have to worry about traffic congestion and bottlenecks. Congestion issues are managed by TCP congestion control; when it receives signals of increased traffic, its algorithm manages traffic by redirecting flow and redefining threshold. In this Discussion, you will explore how TCP handles congestion control.

To complete this Discussion:

Post: Submit a response to this scenario: N-number TCP connections are present over some bottleneck link of rate R bps. All N connections have a huge file to send in the same direction, over the same bottleneck link. Discuss what happens if the N transmissions all start simultaneously. Explain and reason how TCP congestion control will react: for example, describe what transmission rate TCP would give each of the connections. Also explain how the bottleneck grows for each addition to N-number of TCP connections. For each additional connection, is latency growth logarithmic, linear or exponential? List possible solutions to the TCP congestion control problem.