| With the rapid development of mobile communication technology,the requirement of network quality and data rate is higher and higher.The emergence of femtocell could solve the problem of weak coverage of cell edge effectively,provide a good diversity gain to certain extent and improve the QoS of users.Therefore,the resource allocation in femtocell system is a hot research topic currently in telecommunication field.From the co-tier interference,the resource allocation algorithms of femtocell networks are studied in this thesis.Taking into account the macro-femtocell network,the combination of the maximal clique,cluster(maximal independent set)and femtocell are analyzed and studied,and three resource allocation algorithms are proposed in this thesis.To sum up,the main contents are as follows:(1)A novel graph-based resource allocation algorithm is presented in two-tier OFDMA femtocell network.The main objective of the proposed scheme is to assign the available spectrum to femtocells dynamically,so as to obtain the maximization of system throughput.The proposed algorithm is divided into three steps,i.e.triangulation,clique degree and dynamic resource reuse,before which the neighboring information is all collected to build an interference graph with interference restricted.The triangulation transforms an ordinary graph into a triangulation graph to collect all the maximal cliques in a polynomial time.Then,the clique degree of all femtocell users are acquired to list the users in a decreasing order.Finally,the remaining spectrum is allocated based on the descending order of clique degree.As shown in our simulation results,the aggregate and average throughput of femtocells are increased obviously with the consideration of fairness for femtocell users.(2)The main objective is to manage the interference and improve the spectrum efficiency,namely the system capacity,despite the severe interference between the aggressor and victim in femtocell networks.For this purpose,a novel resource allocation scheme based on clustering strategy with Quality of Service(QoS)ensured is proposed,which consists of three parts,i.e.maximal cliques enumeration(MCE),cluster formation and resource allocation.The proposed algorithm assigns subchannels in the unit of cluster acquired by the maximal cliques,during which the femtocell suffered from less interference is designated with more subchannels.As shown in our simulation results,the aggregate and average throughputs are improved obviously compared with its counterparts.(3)In order to mitigate inter-femtocell interference(IFI)and utilize spectrum resource efficiently,a dynamic clustering-based sub-band allocation scheme is proposed.Firstly,the scheme constructs a weighted interference graph base on the user measurement report mechanism.Secondly,according to the interference graph,the disjoint IFI-minimizing clusters are formed by a Max k-Cut clustering algorithm.Thirdly,a SINR-maximizing heuristic algorithm assigns sub-bands to each cluster.Finally,due to the time variation characteristic of the femtocell networks,a cognitive sub-band self-management mechanism is provided,especially to allocate the sub-band for the newly added femtocell,a reuse entropy based hybrid centralized/distributed interference-aware sub-band allocation algorithm is presented.Simulation results show that the DCCSA scheme can improve the throughput of femtocells and suppress the IFI significantly while guarantee the required QoS. |
