Routing in Mobile Ad-hoc Networks considering Human Attributes
The current explosion of batteryNpowered mobile devices has driven an immense growth in the amount of computing power and storage ability available on the planet. There are a huge number of applications (e.g., mobile social applications, mobile storage applications) built for these mobile devices that take advantage of these mobile storage and computing resources. The underlying network infrastructures and protocols are one of the keys to the success of these applications. Therefore, it is very critical for us to study the fundamental aspects of such networks (e.g., mobile adNhoc networks) and build more efficient and intelligent network protocols to onboard more next generation mobile applications.
The main object is to propose efficient strategies for establishing routing paths within such mobile adNhoc networks. There are two main directions in this proposal: Contact & Packet Dropping Prediction and Routing Protocol Design. Contact prediction is a vital part of routing in mobile adNhoc networks. Most previous work predicted contact probability only according to contact history. Many human factors and surrounding environments were ignored. We will study what human factors (e.g., interests, job title, current time, and current location) can help to improve prediction accuracy. When their resources are used up and new packets are still coming, some packets will be dropped.
One important issue is raised. Whatis the probability that one packet will be dropped by one device? Both contact probability and packet drop probability will affect routing performance. We will study how to balance these two factors to achieve efficient routing protocols.
'Task' assignment for Tacoma Water
Tacoma water studies water by sampling and testing water quality. There are a set of places where researchers will go to take samples. There are two problems are raised. 1) Given a group of researchers who will go to take samples and a bunch of places of interests, each researcher will be assigned of a place subset. That is, every researcher needs to go to the places in the subset assigned to him/her. We define journey time of one researcher as the total time this researcher needs to visit all the places assigned to him/her. How to assign these places to each researcher to minimize the maximum journey time of all researches? 2) Based on the previous sampling results, is it possible to find the result similarity among sampling places? If we can find the similarity among a group of places, we may go to one instead of all places in this group. Then, which one should be chosen in this group?