Study On Cognitive Pilot Channel Design And Self-Organizing Techniques In Cognitive Wireless Networks

Driven by the demands for better user experience and various types of service, wireless network technologies are developing rapidly recently with the capabilities of providing high data rate services with good quality of service (QoS). However, challenges still exist as different kinds of heterogeneous radio access technologies (RATs) are coexisting with each other, which make the network convergence a complex problem. Moreover, fixed spectrum allocation and management schemes are not efficiently adapt to the dynamically changing radio environment and user demands. So the challenges for future wireless network development are summarized as:how to design an efficient technique for heterogeneous network information delivery, how to decrease the complexity of heterogeneous network radio resource management, how to improve the spectrum efficiency and increase network capacity.In order to solve these challenges and improve radio resource usage efficiency, cognitive wireless networks (CWNs) technologies are considered as candidate solutions, which are based on the software defined radio (SDR) technology and cognitive radio (CR) technology. The CWNs should have the abilities to reconfigurate the parameters, protocol stacks and working modes, which are based on the sensing techniques and learning ability of existing knowledge in database, in order to adapt to the changing radio environment. The CWNs should also have the abilities to obtain the radio environment information, self-optimize radio resource and dynamically tune the parameters and protocols to adapt to the changing environment.This thesis focuses on novel theories and techniques for efficient heterogeneous network information delivery, simple heterogeneous network radio resource management, improved spectrum usage and increased network capacity in CWNs.Facing to the challenge of efficient heterogeneous network information delivery, the cognitive pilot channel (CPC) design theory and techniques are proposed in this thesis, which is regarded as a common pilot channel for accurate and efficient heterogeneous network information delivery. The key contributions include two parts.First, both the out-band CPC and in-band CPC modes are introduced briefly, and both the pros and cons between them are analyzed thoroughly. Next, key parameters for heterogeneous network information delivery using CPC are summarized as terminal position shift error and multi-RAT overlapped effects. The close-form theoretical analyses are achieved based on the mathematics model for the probability calculation of information error caused by the GPS based terminal shifting effect. Then, based on the Analytic Hierarchy Process (AHP) and Grey Relational Analysis (GRC) algorithms, novel mesh division schemes are proposed by considering both the terminal position shift error effect and heterogeneous network information, which is an achievement in this field as nobody has proposed complete solutions with proofs before.Second, to further improve the efficiency of traditional broadcast CPC, the differential mesh information coding (DIC) scheme is proposed in order to reduce the redundancy of network information between adjacent meshes by using image processing techniques. Then, to optimize the traditional broadcast CPC delivery mode, a homogeneous meshes grouping (HoMG) scheme has also been proposed which can reduce duplicate network information delivery in adjacent meshes, assuring an efficient and accurate network information delivery with novel CPC design. Therefore, the proposed optimal mesh division scheme and novel broadcast CPC flow are the fundamental basis for the accurate and efficient heterogeneous network information delivery in CWNs. Due to the increase of heterogeneity and complexity for the operation and maintenance in CWNs, the traditional techniques for network planning, operation and maintenance are not efficient and applicable in CWNs, which are human interaction dependent, time consuming and cost inefficient. Therefore, self-organizing networks (SON) techniques, including self-planning, self-configuration, self-optimization, self-management and self-healing, are analyzed in this thesis for efficiently autonomous network parameter configuration, optimization and maintenance in CWNs, to improve the radio resource usage efficiency and decrease the CAPEX and OPEX. To improve the network capacity, coverage and spectrum efficiency for indoors and hotspots, self-configuration and self-optimization techniques for femtocell networks and cognitive cooperative relay networks are taken into account in this thesis. The key contributions include three parts.First, both the downlink capacity model and key parameters of three frequency allocation schemes for femtocells deployment in the hierarchical network are designed and analyzed with close-form solutions. Next, a novel location-based frequency information acquisition scheme is proposed which includes the technique to obtain geo-location based macrocell resource occupancy information, the efficient delivery process using CPC to femtocells. Then, the frequency self-deployment scheme for femtocells is designed with the optimal grid zone size selection for femtocells by leveraging the spatial reuse of macrocell resource to improve system capacity.Second, to improve the coverage indoors and increase hierarchical network capacity, the joint antenna pattern selection and dynamic power allocation scheme for coverage self-optimization scheme in femtocells is proposed. The effects of both the static power allocation and dynamic power allocation schemes to coverage holes indoors and interference outdoors are analyzed theoretically. Next, the optimal coverage radius for femtocell indoors is analyzed in three scenarios with close-form solutions, such as the center position, corner position and midpoint of side wall. Then, the geo-location information of femtocell is obtained via CPC channel, in order to locate the indoor femtocell position and the distance to macro base station. Thus, the joint antenna pattern selection and dynamic power allocation scheme is modeled by using the artificial neural network (ANN) training the learning techniques, to fulfill the femtocell coverage self-optimization object.Third, to further extend the coverage and improve network capacity, the dynamic time slot allocation scheme based cognitive cooperative relay network capacity self-optimization scheme is designed. The capacity model for two hop cooperative relay networks is modeled in three different scenarios. Next, considering the difference of channel capacity on various links due to the randomness and time-variant effects, the optimal dynamic time slot allocation scheme is proposed and proved theoretically with close-form solutions, which can improve the relay link capacity by self-tuning transmission time slot between two relay links. Then, based on the vacant spectrum information via CPC, the joint relay node selection, channel allocation and dynamic transmission time slot allocation scheme is designed which is regarded as the capacity self-optimization solution in CWNs by using the optimization theory and graph theory. In summary, the proposed frequency self-deployment, coverage self-optimization in femtocells and capacity self-optimization in cognitive cooperative relay networks prove the efficiency of SON techniques to decrease the complexity of CWNs management and improve the spectrum usage, capacity and coverage.Finally, the summary is given at the end of this thesis and future research directions in related fields are also pointed out.