Network Congestion And Reactive Pricing Model

With the increasing application of the multimedia service in the Internet, the network congestion has become increasingly serious. After a probe into the causes of the congestion, this thesis is going to discuss the externality effect of the network congestion. The existence of the congestion externality has resulted in inefficient allocation of network resources. In order to utilize the network resource in an economically efficient way, this thesis has proposed a response pricing mechanism, which can allocate the network resource in an efficient way.This thesis has found out that three key factors have mad network congestion more and more serious. They are: the defect in the current networks pricing system; the increasing demand of the consumers over the network service; privatization and commercialization of the Internet. The structural impact of these factors is so profound that we have to reconsider the economic system of the Internet.Congestion is an inefficient means of resource allocation, because the allocation of resource is achieved by waiting. Generally speaking, rare Internet resources will be allocated to those Internet users who have lower opportunity cost of time by this means of allocation. Therefore, network congestion is a typical non-economy of externality.The most efficient way of solving network congestion is through pricing mechanism like solving most other externality effect. An effective pricing mechanism will automatically balance the demands and supply for network resources. On the contrary, the current flat pricing mechanism, which has a zero marginal price, will make the network congestion more and more serious.If a proper feedback signal can help adaptive Internet users react to network condition accurately, then it will definitely increase the network efficiency. When the network traffic is too much and network congestion has occurred, the feedback signal will help the adaptive users to decrease the data package transmission. When the network traffic is quite moderate, the feedback signal will encourage the adaptive user to increase the data package transmission. Therefore, the network congestion can be solved.This thesis will start from the perspective of welfare economics to define what is efficiency of network resources allocation and what is quality of network service (QoS). After such definitions, this thesis will make use of qualitative and quantitative methods to study efficiency loss of network congestion. In addition, this thesis will also touch upon some aspects of game theory when examining the cause of network congestion and designing the responsive pricing mechanism.