The first of the network performance parameters to be analyzed was the time elapsed for establishing a connection between the IP of the user’s computer and the host TCP (Transmission Control Protocol). Usually, this can be measured qualitatively for an application by looking either at the extent of utilization of the CPU (Central Processing Unit) or the stream of data packets being exchanged between the User Internet Protocol and the Host Transmission Control Protocol. However, the qualitative analysis has a limitation for two reasons:
1.Most of the applications are closed, though some essential applications such as the subnet mask tracer as well as the dynamic IP socket (for most wireless LAN connections, as is the case with the hardware configuration used for the purpose of this experiment, the address is a dynamic IP and not a standard, fixed Internet Protocol address provided to a registered user. However, this cost has been, in the case of this experiment, outweighed by the added benefit that the experiment can be conducted for several times of the day over a shared network to better understand the node traffic configuration as well as the node traffic congestion to what extent and to what impact it is a part of the network being analyzed.
2.The second reason is that a mere qualitative analysis of a network with virtually no applications running has little consequence as compared to a benchmark-clamping method that relies on the quantitative network performance parameters.
In order to understand the network performance parameters being parameterized in this experiment, some context is in order.
The first parameter, as mentioned above, is the most crucial and relates to the establishment of a connection between the user Internet Protocol Address and the Host Transmission Control Protocol. Once the protocols acknowledge each other’s addresses, the connection is deemed to be complete. When the two simultaneous processes in question wish to communicate, the first and foremost event, whose time complexity is of much consequence to the order in which the data sharing process is streamlined in the form of packets, their TCP’s must first establish a connection with each other, i.e. initialize the address, network state, network configuration, status information and a set of other trivial parameters on the other side. It has been observed in the course of this experiment that as the connections must be established between two potentially unreliable hosts and over a network that spans over a predominantly unreliable internet connection system of communications.