With the proliferation of wireless communications, many wireless communications technologies coexist in the same area of an urban setting. For instance, in a cellular zone, there exist a base station (e.g., xG), various WLANs, picocells and femtocells. So what is the best choice among all these alternatives for a mobile user? How should it select the access technology for the best connectivity? And how can femtocells make themselves attractive so that mobile user connects through them at the expense of sharing the cost of femtocell with its original owner?

The aim of this thesis is to design a decision scheme that considers the demands and requirements of various parties in the system. First, the agents, their properties as well as their possible objectives (e.g., profit making for femtocells to share their spectrum while mobile user may like to have a certain bit rate) will be defined in such a heterogeneous network environment. Femtocell devices can also have some cognitive functionality to discover new spectrum or offer a time-varying spectrum usage fee for attracting the mobile user to connect via itself. Next, using the system definition and related analysis, a decision scheme will be proposed.

Basic research questions can be listed as follows:

1. What is the best technology for the mobile user among the alternative wireless technologies?

2. What are the possible objectives of the entities in the considered system?

3. Should a centralized architecture or a non-centralized architecture be applied?

4. What policies should a cognitive femtocell device apply in order to increase its profit?

5. Should a femtocell seek for white spaces or can its already available spectrum meet the demand?

6. How is the pricing mechanism designed and how to decide on a reasonable spectrum cost at the femtocell?

7. What are the key (limiting) factors (e.g., data limits at each femtocell, number/density of free WLANs) in such a system?

Tentative outline for this thesis work

• A literature survey on femtocells, spectrum sharing in two-tier cellular networks, and selecting the best connectivity link (1 month)

• System definition: agents, their properties, aims, system performance metrics (1 month)

• Design of a decision scheme (3 months)

• Implementation of the decision scheme in a simulator (2 months)

• Analysis of the proposed scheme via simulations, revision/tuning of the model(1 month)

• Thesis writing (2 months)

Required Background:

• General understanding of wireless communications

• Proficiency in one of the programming languagesA weekly meeting (possibly 1 hour duration) will be scheduled and further meetings can also be scheduled upon request.

Some useful things to know before you start:

• Mendeley for organizing your digital library

• Latex for writing all thesis related documents

• Dropbox for sharing in case of multi-user editing is required

• Please read the following document before reading scientific papers: http://blizzard.cs.uwaterloo.ca/keshav/home/Papers/data/07/paper-reading.pdf

• Please read this paper before writing your manuscript: http://www.cs.columbia.edu/~hgs/etc/writing-style.html