Use this URL to cite or link to this record in EThOS:
Title: Some issues in designing edge pricing schemes
Author: Birdsall, G. J. R.
Awarding Body: University of Cambridge
Current Institution: University of Cambridge
Date of Award: 2005
Availability of Full Text:
Full text unavailable from EThOS.
Please contact the current institution’s library for further details.
Communications network operators wish to promote fair, efficient and stable use of network resources. Service providers are obligated to pass users' flow on to their destinations. They may do this directly or pay a neighbouring service provider to take the flow and pass it on to its destination. Payments are made at boundaries between adjacent service providers, hence the name edge pricing. This thesis introduces a simple model of edge pricing, and generalizes it to handle multicast flows and payments shared between sender and receiver. Edge pricing can act as a control mechanism. We discuss why such a mechanism is necessary, its important characteristics, and the advantages of edge pricing over other pricing mechanisms: edge prices may be calculated in a decentralised mechanism, requiring only locally available information. Where users have variable demand for flow, the routing is determined as a function of local prices. We explore whether global stability of prices and routings in the network could be achieved by a purely local price adjustment mechanism. An equilibrium is not always reached, or many possible equilibrium prices and routings may exist. Illustrative experiments follow, assessing whether over-simplification of the model causes the instability and the factors most important in characterizing the equilibrium. In a separate problem, it is desired to identify the most valuable nodes or links in a network. A cooperative game on the network is considered, where nodes may form coalitions to maximize their joint profits. We analyse whether the profits can be split fairly by the coalition. Under reasonable conditions, the core of the cooperative game is shown to be non-empty, hence the profits may be split in a 'subsidy-free' manner. The above two ideas are linked as a stable routing for edge pricing corresponds to a Nash equilibrium point of a particular non-cooperative game (which in turn is closely linked to the above cooperative game) with the nodes of the network as players. A mixed strategy Nash equilibrium is proven to exist, but it is demonstrated that a pure strategy Nash equilibrium may not exist. This helps to explain why global stability may not be achieved from local price adjustment mechanisms such as edge pricing.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available