Wireless LAN: Traffic and Congestion Control – Term Paper Example

The paper "Wireless LAN: Traffic and Congestion Control" is a worthy example of a term paper on information technology. This paper seeks to analyze the issues that are presented in the Wireless LAN case study Analysis. The study is mainly focused on providing an analysis of traffic and congestion control on Wireless LAN networks using various research methods and tools. This case study has highlighted a number of researches which has been conducted to demonstrate traffic in a dense wireless Local Area Network (LAN) environment. Analysis of Wireless LAN Characteristics
Various studies have been conducted in a number of Campus networks to measure network traffic using different approaches. The studies try to analyze macro characteristics of wireless LANs such as overall usage, traffic characteristics, user mobility, and application mix. User Mobility and session data in the network has been measured using the authentication logs of users. In terms of user mobility, the network performance is measured depending on the location of the user and how user movement affects the network usage. From studies carried out, a common conclusion was reached that the network users remained constant as the user tended to move around an access point.
Another characteristic that has been illustrated by the study is the user population across APs and its relation to network performance. The studies reveal that the aggregate data transfer rate of an AP did not depend on the number of users that are associated with it, but it can rather be related to the users present.
In short, the study has explored a number of macro-characteristics that are associated with network traffic. The major characteristics under study include User mobility, session duration, and overall traffic. It can also be deduced that HTTP applications and TCP packets contribute to P2P network traffic as well. The variation in the application and the bandwidth also affect the traffic congestion in the P2P networks.
Wireless LAN Environment Analysis
The second issue highlighted in this study is the analysis of the wireless LAN environment using porcupine a platform that is used in wireless data analysis. This platform is used to support research into an advanced wireless environment. It consists of sixteen separate radios, sixteen single board computers and sixteen directional antennae each of which can be configured to function together. The porcupine can be able to capture wireless packets and determine their origin in a real-time manner. The Porcupine was demonstrated to be able to sniff all 802.11 traffic (both management frames and user data).
One of the advantages of passive sniffing is that it can allow one to be able to observe the unconstrained behavior of the wireless network. This type of model allows one to detect rouge Access points, ad-hoc networks and misconfigured clients which cannot be detected by other means. One disadvantage of passive sniffing is that it can only capture wireless traffic that can be decoded properly any of the antennae. Another disadvantage of the passive sniffing of wireless data by the porcupine relates to the variety of 802.11 modes offered at the SC2004 exhibit floor. The porcupine has been designed to capture only 802.11b wireless packets. This is an implication that the effect of the presence of 802.11g traffic would be indirectly observed in the 802.11b performance, directly observing 802.11g traffic, that shares 2.4GHz radio frequency spectrum with 802.11 is not possible. It is not also possible to save all traffic of 802.11b due to
Network Protocols
The third issue this case study tries to demonstrate is the use of protocols in the analysis of traffic. The study IPv4 protocol was used to test out for logs at the network layer. The transport layer consists of two main types of protocols, these are TCP and UDP. The TCP is a connection-oriented protocol and is used to perform congestion control while the UDP is a connectionless protocol and does not perform congestion control.TCP and UDP are unicast protocols while IGMP is a broadcast protocol. Because of these differences, it is vital to understand the performance of a network that is based on the exact mix of different protocols.
Bytes and flows are traditional methods in the analysis of traffic and so the perspective of the packet is used since it better measures the popularity of a protocol in a wireless medium since channel contention can be resolved per packet, irrespective of the number bytes that a packet contains. The adoption of the use of packets in the understanding of traffic results to a number of important insights into various transport protocols. One of the conventionally accepted beliefs about the prevalence of transport protocols on the internet is that TCP carries most internet traffic while the fraction that of traffic by other transports protocols other than UDP and TCP is quite insignificant.
The other issue illustrated in this study is the HTTP application how much it accounts to the total internet traffic. Studies carried out on the wireless LAN networks shows that HTTP application does not dominate in the traffic in bytes on the internet. NetBIOS and UDP based application seem to dominate the traffic on the internet.
The micro-scale analysis that has been used in the case study reveals that there is significant performance loss for the users for a dense wireless LAN traffic. There is a question as to whether to trust protocol usages and distributions without using the micro-scale analysis. It is evident from the research that TCP application is involved in congestion control that may result in reduced sending rates. The study has shown that there is a need for some congestion control for other protocols other than TCP, otherwise TCP-based application may suffer immensely if the transport layer protocol that is observed becomes an internet reality.