Key Research Issues For Control And Optimiz-Ation Of Heterogeneous Small Cell Networks

A key problem in the mobile communications lies in the contradiction be-tween the limited spectrum resources and explosively increasing demand of multimedia services. In order to improve the system capacity, high spectrum utilization technologies are under intensive research investigation. To enhance the spectrum utilization, this thesis considers underlaying of small cells with macrocells which has been brought forward as an effective solution for future cellular networks. There are a lot of issues in heterogeneous cells deploymen-t, however this thesis mainly focuses on inter-cell interference (ICI) problem. When small cell such as femtocell operates in close access, it creates severe in-terference problem to the macro user in the close proximity that cannot access it. Thus ICI becomes a major obstacle which can impair the potential gains from the small cells and its pattern is highly diverse. In this thesis we pro-pose power control and precoding methods for the interference management in heterogeneous networks.In power control schemes the interference signal is treated as noise and interference can be avoided by adjusting the transmitter’s power dynamically. Inspired by the game theory, this thesis propose pricing based power control in femtocell networks witch is mainly based on bi-level Stackelberg game. In the game, macro base station (MBS) acts as a leader and femto base station-s (FBSs) as followers. The leader allows the use of its spectrum resources to the followers by paying interference price, conditional to the critical limit of interference tolerance, and thus earns the revenue. FSBs as followers, calculate their power subsequently based on interference price. The purpose of the game is to find the Stackelberg equilibrium (SE) where neither leader nor followers have incentive to deviate from their strategy unilaterally. We derive closed form solutions of the game considering important constraints, thus exhaustive search for resource allocation is avoided which is usually the case with convention-al solutions. Additionally, energy efficiency of cellular network is becoming increasingly important under current trends of green communication. We also extend our work for energy efficient power optimization in two-tier femtocell networks based on Stackelberg game model.In parallel, multiple input and multiple output (MEMO) system has also attracted huge attention due to its potential capacity increment. If we deploy multiple antennas at both transmitter&receiver and design a proper precod-ing matrix, channel interference can be pre-cancelled at the transmitter thus channel reliability&data rate can be enhanced without additional transmission power. Based on MDVIO, coordinated multipoint transmission/reception (CoM- P) has emerged as an effective technique for inter-cell interference cancelation. For joint processing and transmission (CoMP-JP), dirty paper coding (DPC) can achieve maximum theoretical capacity, however, it is non-linear and brings huge computational burden. Linear precoding schemes such as block diagonal-ization (BD) are less complex, hence are more feasible in practice. We propose a modified form of BD for interference mitigation in femtocell networks and describe an algorithm to replace computationally complex SVD operations of BD with low complexity operations based generalized zero forcing channel in-version (ZF-CI), QR decomposition and lattice reduction (LR) transformation methods.Altogether, this thesis investigates fast converging power control algo-rithms and low complexity high efficient precoding methods for interference mitigation in small cells network.