Resource Management In Heterogeneous Cellular Networks

The world-wide explosive growth in the amount of mobile traffic and the number of mo-bile devices has posed new challenges to mobile network operators.To cope with this data deluge in mobile networks,the denser deployment of network nodes,which enables a higher spectrum reuse efficiency,has been considered as a key technology in next generation mo-bile networks.Nevertheless,the irregularly-deployed network nodes would inevitably cause more severe inter-cell interference,in particular for cell-edge users.Therefore,the interfer-ence in heterogeneous cellular networks has posed serious constraints on the improvement of spectrum efficiency and network capacity,and must be suppressed in demand.Through signaling exchange across networks,the Inter-Cell Interference Coordination(ICIC)technique distributes resources to users with good channel conditions,and mean-while avoids the severe interference to cell-edge users.Therefore,ICIC can mitigate in-terference and thus improves the spectrum efficiency effectively.In the interference co-ordination stage,multiple transmitters coordinate and eventually decide their resource man-agement strategies,thus distributing orthogonal resources to different heterogeneous cellular networks,combined with the power control.It follows that ICIC is in essence an interference avoidance-based resource management scheme,without complex multi-antenna configura-tion as well as advanced signal processing.This dissertation mainly focuses on the interfer-ence avoidance-based resource management;meanwhile,to leverage the simultaneous trans-mission between multiple wireless links via the multi-antenna technology,this dissertation further involves the combination of resource management with interference alignment,to mitigate interference in spectrum-scarce scenarios.Particularly,aiming at Quality of Service(QoS)guarantee,the QoS-guaranteed user number maximization as well as multi-operator deployment in heterogeneous cellular networks,the main contributions of the dissertation are as follows:1.We put forward a QoS-based resource management scheme in small cell networks,to enable the flexible regulation on network resources via the joint optimization over band-width and power.Firstly,focusing on the QoS differentiation between different users in Orthogonal Frequency Division Multiple Access(OFDMA)-based small cell networks,we formulate a bandwidth-power product optimization model,to minimize the resource con-sumption subject to the QoS guarantee constraints.Secondly,to differentiate the scarcity of different resources,we apply different emphases on consumed power and bandwidth com-ponents,and formulate them into the optimization objective.Then,by leveraging the Suc-cessive Convex Approximation for LowcomplExity(SCALE)algorithm,we reformulate and decompose the original problem into a series of convex subproblems,each of which can be solved effectively via existing convex optimization algorithms,and meanwhile the monotonicity as well as convergence of the proposed scheme are also proved via theoretical analysis.Finally,simulation results validate the theoretical analysis,and explore the effects of different weight settings on the system performance.2.We propose a user number maximization-based resource management scheme in small cell networks,to maximize the number of QoS-guaranteed users in the resource-scarce small cell networks via a hybrid subchannel allocation scheme.Firstly,aiming at the OFDMA-based small cell networks and assuming that network resources are not suffi-cient to guarantee the QoS requirements of all users,we strive to maximize the number of QoS-guaranteed users,subject to the subchannel allocation constraints.Secondly,to over-come the delay issue in signaling exchange,we develop a mixed-time two-level resource management scheme,which runs optimization algorithm respectively in Radio Resource Management(RRM)controller and small base stations.Finally,simulation results can vali-date the advantages of the two-level scheme over single-level ones.3.We present a multi-operator deployment-based resource management scheme in het-erogeneous cellular networks,to enable the integration of network virtualization,content caching as well as D2D technology in the Software Defined Networking(SDN)paradigm.Firstly,aiming at the tendency toward the multi-operator deployment,we develope a software-defined information-centric D2D-assisted cellular network virtualization framework,to re-alize the dynamic resource(spectrum,base stations,D2D users)allocation and content caching via the SDN controller.Secondly,in the proposed framework,involving the im-perfectness of both the channel estimation and measurement,we formulate a joint virtual resource allocation and caching refreshment optimization model,with the objective of max-imizing the totol utilities of Mobile Virtual Network Operators(MVNOs),and then solve it via the classical Discrete Stochastic Approximation(DSA)algorithm.Finally,simulations results reveal that MVNOs can benefit from not only the sharing of substrate networks,but also the ubiquitous caching functionalities in mobile devices.4.We raise a multi-operator deployment-based and interference alignment-combined resource management scheme in heterogeneous cellular networks,to enable the isolation between virtual slices in the spatial domain.Firstly,aiming at the tendency toward the multi-operator deployment and on the basis of spectrum-scarce network scenarios,we pro-pose a framework of heterogeneous cellular network virtualization combined with interfer-ence alignment,to realize the virtualization of substrate networks by aligning interference from other transmitters into the reduced dimensional subspace of each receiver.Secondly,involving the imperfectness of both the channel estimation and measurement,we formu-late a joint virtualization and interference alignment optimization model,with the objective of maximizing the totol utilities of MVNOs.Then,to solve the formulated problem,we propose a decomposition algorithm,to design the interference alignment scheme for each feasible virtualization combination,and then to search over the virtualization combination space for optimal one.Finally,simulation results reveal that MVNOs can benefit from both the interference alignment and wireless network virtualization.