Ph.D. position on Incentive Mechanisms in Delay-Tolerant Networks

The SARA team of LAAS-CNRS (http://www.laas.fr), Toulouse, France, is inviting applications for an open Ph.D. position on incentive mechanisms in Delay-Tolerant Networks.

Description

Delay-Tolerant Networks (DTNs) rely on the store-carry-and-forward principle to ensure connectivity, that is, each node acts as a relay by storing messages temporarily and forwarding them to other nodes it meets on the way. A message thus hitch-hikes on the back of other nodes until it arrives at the destination. Relaying of messages induces energy and storage costs, which could be substantial for terminals with small-sized batteries. Thus, some nodes might be tempted to free-ride by not relaying messages while relying on others to relay their messages. Such selfish behaviour can have an adverse impact on the network's efficiency, and result in a possibly disconnected network. The main question of interest in such a scenario is: how to persuade nodes to relay messages for others?

The thesis will focus on the analysis of an incentive mechanism in which the source promises a reward to every relay it meets. The amount of reward can vary as a function of the meeting time and the state of the buffer of the relay node. The relay that is the first one to deliver the message is given the promised reward. While this mechanism provides an incentive for relays to participate, it also leads to competition between them for the rewards on offer. Since each relay is unaware of the state of the other relays, this competition can be modelled as a stochastic game with partial information, which shall serve as the basic framework for analysis.

The main objective of the thesis is to develop analytical tools for investigating the economy of DTNs operating under this incentive mechanism. Questions related to the distribution of wealth, to the influence of the reward on the cooperation and hence on the throughput, and to the impact of information given by the source (on the number of existing copies, or the age of these copies) on the performance of this mechanism will be studied. Finally, the utility of this mechanism in other competitive contexts such as peer-to-peer networks will also be explored.

Key words: Stochastic games, mobile communication networks.

Desired profile of candidate: The candidate has (or is in the final year) of an engineering
degree or a master's degree in Computer Science or in Applied Mathematics. She/he has an inclination towards mathematical analysis, and will have acquired skills in basic optimisation theory and stochastic modelling. Familiarity with programming languages (C/C++/Matlab) will be a plus.

Application material should include a CV, a letter of motivation, and names and contact
details of two referees.

Contacts (for sending the application material or for queries):
Olivier BRUN (brun@laas.fr, homepages.laas.fr/brun) and Balakrishna PRABHU
(balakrishna.prabhu@laas.fr, homepages.laas.fr/bala).

Nr of positions available : 1