Research On Resource Allocation In D2D And Full-Duplex Enabled Heterogeneous Cellular Networks

To deal with the explosive growth of mobile internet service as well as providing seamless wireless coverage,the 3rd Generation Partner Project(3GPP)brings in heterogeneous cellular networks(HCNs)in LTE-Advance systems,which deploy low-power small base stations(SBSs)in the coverage of traditional macro-cells to enhance the overall performance of the system.Compared with the traditional cellular networks,HCNs can significantly improve the spectrum efficiency(SE)and the coverage of the system.On one hand,by sharing the spectrum resource with the macro-cell,small cells can improve the space utilization of the spectrum resource,which leads to an improvement of the system SE.On the other hand,deploying the SBS in the hot spots of macro-cell can improve the coverage of the macro-BS and provide user terminal with more ubiquitous access services.As the research further develops,how to further improve the throughput,extend the coverage,and reduce the energy consumption becomes the major concern of the HCN.Since Device-to-Device(D2D)and full-duplex communications can significantly improve the SE and energy efficiency(EE),integrating them with HCNs has been considered as a potential way to build a high efficient wireless communication system.But bringing D2 D and full-duplex communications into HCNs can also make resource allocation more challenging.In this dissertation,we mainly focus on resource allocation issues in D2 D and full-duplex enabled HCNs,which includes the following four parts:(1)Mobile association in HCNs with single D2 D relay node(DRN);(2)Mobile association in HCNs with multiple DRNs;(3)Resource allocation for full-duplex enabled HCNs with orthogonal spectrum sharing;(4)Resource allocation for full-duplex enabled HCNs with non-orthogonal spectrum sharing.The first part investigates mobile association in HCNs with single DRN,and proposes a cross-layer resource allocation algorithm to maximize the SE and EE of uplink transmission.The proposed algorithm sequentially operates in the following three steps:access point selection,transmit power control,and mode switching.In each step,the closed-form solution of the corresponding problem can be easily obtained.It shows through computer simulation that the proposed mobile association algorithm can significantly improve the SE and EE of uplink transmission.It also shows the cell edge user terminal can achieve higher SE and EE by adopting the relay mode than using the direct mode.Besides,the tradeoff between the minimum rate requirement and EE performance is also illustrated.The second part investigates mobile association in HCNs with multiple DRNs.Since the random deployment of DRNs,the estimation of channel state information(CSI)may incur high signaling overheads.When the number of DRNs is large,these signaling overheads are unacceptable.Therefore,how to design a mobile association algorithm with low signaling overheads is worth further exploring.To deal with this issue,a distancebased mobile association algorithm is proposed.This algorithm first derives the optimal relay location and then chooses the optimal DRN according to the relative position between the optimal location and DRNs.At last,it performs optimal power allocation and mode switching.The simulation results show that the proposed algorithm can achieve a good tradeoff between the uplink EE and complexity.The third part studies resource allocation in full-duplex enabled HCNs with orthogonal spectrum sharing and explores the joint optimization of uplink-downlink subcarrier assignment and power allocation to maximize the overall throughput of a small cell with the consideration of power limits at user terminals and the SBS.By using the Lagrange dual method,the original optimization problem is decomposed into a primal problem and a dual problem.Since the primal has a non-convex structure,it is solved by the concave-convex procedure(CCCP)through sequential convex approximations.For the dual problem,the sub-gradient method is utilized.Simulation results show that the proposed algorithm can always achieve better throughput in comparison with the existing algorithms.The fourth part studies resource allocation in full-duplex enabled HCNs with nonorthogonal spectrum sharing.Unlike the orthogonal spectrum sharing scenario,there exists inter-tier interference in the non-orthogonal spectrum sharing case.Therefore,besides maximizing the throughput of the small cell,the transmission in macro-cell should also be guaranteed from severe inter-tier interference.Due to the random deployment of SBS,the CSI of inter-tier interference channels may not be perfectly known.Therefore,channel estimation error is considered and the robust optimization theory is adopted to formulate a joint uplink-downlink optimization problem.Similar to the orthogonal case,the original optimization problem in the non-orthogonal case is decomposed into a primal problem and a dual problem by using Lagrange dual method.The CCCP is adopted to solve the non-convex primal problem through sequential convex approximations and the dual problem is solved by the sub-gradient method.Simulation results show the effectiveness of the proposed algorithm and demonstrate the impact of channel uncertainty on the system performance.