Mobile IP Telephony (MIPTel)

Abstract

The MIPTel project aims to develop and support mobile telephony applications over IP networks. Providers are in great need of scalable, extensible, flexible and transparent charging and accounting methods, which take into account the specific attributes of wireless networks and requirements of diversified services. A wide range of accounting, charging and pricing schemes have been analyzed.

We found that the increasing user mobility requires a new approach to the existing charging and accounting concepts. While the users demand for increased service coverage, the providers aim for low infrastructure costs. Multi-hop cellular networks are a promising concept, which suits both needs. It combines the advantages of two worlds: the dynamics of mobile ad hoc networks and the reliability of wired networks. Mobile ad hoc networks are spontaneously created by interconnecting nodes operated by individual users. In order for the network to function, the nodes are required to participate in the packet transmission process, in particular they must forward packets transmitted by other nodes. This behavior is also known as cooperation among nodes. How to ensure the cooperation in civilian networks in the absence of a single authority is a challenging research question, which we tried to answer in this project. The project has been successfully completed with the PhD thesis of Attila Weyland, which presents the cooperation and accounting strategy for hybrid wireless networks called CASHnet.

CASHnet introduces charges and rewards to the packet transmission process using a hybrid accounting architecture. The charging and rewarding is done decentralized on the nodes and the refill of the virtual currency account is done at service stations operated by a provider. It uses separate accounts for charging as well as rewarding and the exchange of the virtual currencies requires the help of the provider. Compared to fully decentralized accounting schemes like Nuglet, CASHnet ensures a constant cash flow and keeps the provider in control of it. Unlike completely centralized accounting schemes, CASHnet does not put additional signaling load on the links toward the base stations. It also allows cost sharing between sender and receiver located in different multi-hop cellular networks. Further, CASHnet supports the provider in the network planning process. We implemented and evaluated CASHnet in the network simulator ns-2 and developed and tested a prototype implementation of CASHnet under Linux.

We see the main contribution in the specification, development and evaluation of a complete cooperation and accounting architecture, which has several unique features: It retains as much flexibility as possible of the multi-hop communication paradigm through its hybrid accounting and decentralized security architecture. It keeps the provider in control of the cash flow through its two virtual currencies and accounts. It also ensures the security of the accounting mechanisms through non-repudiation.

Project Details