| Nowadays wireless communication has made substantial progress and users’expect is gradually changing. Firstly, users’requirement for higher data rate increases day by day, which is a great challenge for the scare spectrum resource. Secondly, more and more users tend to use communication equipment indoors and it leads to the increasing number of base stations. Thirdly, the traffic is becoming more and more fluctuated in real networks. To fulfill users’ demands and reduce the CAPEX (Capital Expenditure) of operators, cognitive femtocell is raised by many experts. Cognitive femtocell is based on the cognitive radio and femtocell. On one hand, cognitive femtocell can be deployed by users to improve their communication experience. On the other hand, in order to optimize the network performance, cognitive femtocell is able to sense the surroundings and adapts its parameters as well as algorithms accordingly. Taking the price and coverage distance into consideration, there can be billions of cognitive femtocells all around the world in several years, which will result in the increase of energy consumption greatly. Therefore, it is of great importance to study on energy consumption reduction in femtocell systems.This dissertation is based on the Huawei Technology Foundation "The Key Technology of Cognitive Radio" and the contents include:(1) The introduction of cognitive radio, femtocell networks and cognitive femtocell networks is given at first. Then detailed descriptions of spectrum sensing, self-optimization and energy saving methods in cognitive femtocells are given.(2) Cooperative spectrum sensing (CSS) can deal with those problems caused by multipath, shadow and hidden terminals and it is widely used in cognitive femtocells. However, more secondary users is needed to carry out local spectrum sensing compared with traditional sensing with single user, which restricts the application of CSS. To deal with this, this dissertation proposed an energy efficient two stage cooperative spectrum sensing (TSCSS) method. In TSCSS all secondary users are divided into two groups firstly. Whether those users in the second group carry out spectrum sensing or not is based on the sensing result of the first group so as to reduce the average number of secondary users taking parts in CSS and the corresponding energy consumption. In order to maximize the energy efficiency of TSCSS, particle swarm optimization algorithm is introduced. Simulation results show that the TSCSS has a good performance in both energy efficiency improvement and the convergence.(3) The self-optimizing ability of cognitive femtocell enables the system to adapt the resource allocation algorithm and base station status according to the current circumstance so as to save energy. The focus of this paper is cell sleeping based on self-optimizing. As a simple but efficient energy saving technology, cell sleeping is becoming a hot topic in green communication. However, traffic in cognitive femtocells is always under high fluctuation, which gives unbearable complexity for those traditional cell sleeping. To overcome this, a traffic prediction based sleeping (TPBS) mechanism with low complexity in femtocell networks is proposed in this dissertation. Firstly, artificial neural network (ANN) is used as a cognitive method to predict the traffic in cognitive femtocell. Then based on the results of traffic prediction, a sleeping algorithm with low complexity is proposed. Simulation results show that the TPBS mechanism has a good performance in both energy saving and algorithm complexity. |
PDF Full Text Request