Research On Soft Frequency Reuse Based Adaptive Interference Coordination Method

As one of the most successful applications of the wireless communication technology,cellular networks have made much contributions to deal with the explosive growth of mobile traffic worldwide.Frequency reuse(FR)is a major breakthrough to solve the problem that the number of users is increasing while available spectrum resources is limited in the cellular network.No matter the widely deployed 4th generation mobile communication system(4G)or the 5th generation mobile communication system(5G)which has just frozen the first international standard,how to efficiently and reasonably reuse valuable spectrum resources is an issue.If the spectrum is reused unreasonably,it will become interferences that affects the system performance.Thus the essence of these interferences is the conflicting utilization of resources.In the future,cellular networks will be dense,heterogeneous,and irregular,and multiple types of interference will coexist.It puts forward higher requirements on the inter-cell interference coordination(ICIC).How to utilize research results available to guide future work? How to design an interference coordination scheme that meets the needs of future networks?These are worth further exploring.According to the above discussions,this dissertation focuses on the design of adaptive soft frequency reuse(SFR)based inter-cell interference coordination methods for the downlink of both homogeneous and heterogeneous cellular networks.We aim to enhance the adaptivity of the standard soft frequency reuse method to cope with complex network environments with minimal effort.The contribution of this dissertation can be listed as follows:1.For a multi-cell homogeneous cellular network composed of several macro base stations(MBSs),an adaptive coordinated soft frequency reuse method is proposed.It is suitable for the case that some cells are cell edge overloaded because of non-uniform user distribution.This method is based on simple and effective standard soft frequency reuse,while the orthogonality of the major bands of adjacent macro base stations is not required.It makes full use of the space of the network to achieve more reasonable frequency reuse.In this dissertation,both the commonly used hexagonal cell structure and all available resources are finely divided.We define two different types of user priorities and design two factorized scheduling methods for cells being cell edge overloaded and its neighbors,respectively.A simple coordination mechanism is designed to implement coordinated adjustments between cells.The target is to maximize the sum data rate of the system while ensuring the minimum rate requirements of all users.The simulation results demonstrate that the coordinated soft frequency reuse method can effectively reduce the outage probability of users when a cell is being cell edge overloaded in a multi-cell network.In addition,we discuss the application of this method in situations that several cells being cell edge overloaded having the same neighbor cell and that the cell shape is irregular.2.For a two-tier heterogeneous network(Het Net)composed of several MBSs and small base stations(SBSs),cell boundaries are redefined due to the deployment of SBSs.In this case,the user classification problem related to soft frequency reuse is more complicated.In general,a fixed reference signal received power(RSRP)threshold or a fixed distance threshold is utilized for user classification.However,these fixed thresholds are vulnerable to the actual user distribution and the varying network.Therefore,we design an adaptive soft frequency reuse method with dynamic user classification rules.An MBS is visualized as two dummy base stations(DBSs).With this simplification,the user classification problem is transformed into the user association problem in a Het Net.Considering the sparsity,the complicated multi-cell problem is decomposed into several single cell sub-problems.By taking the integrity gap into consideration,we propose a load aware user association algorithm based on literature available.Since the problem decomposition may have some losses,a simple mechanism is designed for compensation.It keeps monitoring the network and will trigger an inter-cell handover when necessary.The simulation results show that the soft frequency reuse with dynamic user classification can adapt to the dynamic network better.It is superior to other schemes using fixed thresholds in terms of system utility,fairness index,and sum rate.Not only the user classification results but also the corresponding cell structure are more reasonable.3.For a two-tier Het Net composed of several MBSs and SBSs,only some cells have SBSs forming a non-uniform topology.It requires an effective ICIC method to cope with different types of interference.We propose a hybrid ICIC method by partitioning the resource in the time domain and the frequency domain jointly.It is based on the division of resources in the frequency domain of standard soft frequency reuse.The major band and the minor band are configured with ABS independently in the time domain.Take the characteristics of different types of resource blocks into consideration,they are integrated to form a novel DBS structure for an MBS.We formulate the optimization of parameters of the hybrid HICIC method as a logarithm utility maximization problem and provide a semi-distributed algorithm comprising three stages to solve it.Numerical results demonstrate that our proposed method can enhance the capability of the system to mitigate interference.Not only the sum rate of the system increases,but the performance of users who have geographical disad-vantages is improved.The most critical point is that our work has proved that interference coordination techniques in the frequency domain and the time domain are not inconsistent in Het Net.Nevertheless,combining these technologies simply cannot make the full utilization of them.These must be a reasonable joint optimization design to reuse the precious spectrum resources efficiently.