Design and Evaluation of Scheduling Algorithms for LTE Femtocells

Inhalt:
With the continuous progress in technology, more complex and sophisticated mobile devices are available to end-users. Furthermore, services such as IP-TV and the introduction of cloud computing result in increased bandwidth demands. Two of the major challenges faced by mobile communication systems such as LTE-Advanced (LTE-A), trying to serve the demands, are the limited frequency spectrum and poor indoor coverage. Femtocells are capable to improve indoor coverage and provide high data rate services to customers. The un-planned deployment of femtocells by end-users can create high and difficultly to predict femto-to-femto- and femto-to-macrocell interference. This paper evaluates the potential gains in spectral efficiency achieved when femto-cells exchange information in order to enable centralised Radio Resource Management (RRM) by joint scheduling decisions. In order to show the degrees of freedom concerning radio resource assignment strategies, a detailed description of femtocell-to-femtocell interference scenarios is presented, outlining the constraints dictated by the LTE-A standard. After describing the problem in mathematical terms and showing its NP-hardness, the problem is re-duced to the so-called Linear Assignment Problem. Based on this idea, centralised uplink assignment strategies were designed with the focus on maximising Cell Spectral Efficiency (CSE). In total four different strategies are presented including an optimal solution for upper bound results as well as heuristic algorithms with reduced runtime. These heuristics are based on methods used in Economics for the cognitive process of decision making under uncertainty. Presented results prove that optimised joint scheduling decisions can significantly increase the spectral efficiency.