Study On Resource Allocation Strategy For D2D-NOMA Communications Via Genetic Algorithm

With the arrival of the fifth-generation mobile communication,the demand for the network capacity and the data traffic is increasing sharply.The current network will face many problems such as limited system bandwidth resources,difficulty to accommodate a large number of connections.Combining D2D(Device-to-Device)technology with NOMA(Non-Orthogonal Multiple Access)technology can improve the frequency utilization and increase the number of network connections.However,the deployment of this new heterogeneous network is based on the introduction of interference and the complexity of the electromagnetic interference environment.So,how to make full use of frequency,power resources and suppress interference effectively through resource allocation has become a research hotspot.Based on the above discussion,the thesis studies the resource allocation problem of D2 D groups reusing the cellular users' wireless resources to communicate in cellular network.This thesis firstly considers two cases,which are respectively D2 D link with perfect Channel State Information(CSI)and D2 D link with imperfect CSI,then proposes algorithms based on both cases.Through establishing the model,designing algorithm,verifying simulation and other means to research.The main work of this thesis are as follows:1.Aiming at the problems of limited bandwidth and power resources,large transmission interference and high computational complexity of resource allocation in D2D-NOMA communication system,this thesis proposes a resource allocation algorithm based on Genetic Algorithm(GA)to maxmize the throughput.Specifically,under the premise of ensuring the requirements of Quality of Service(Qo S)of cellular and D2 D users,and the constraints of maximum transmitting power,the thesis establishes the D2 D system throughput maximization resource allocation problem.Due to the many shortcomings of traditional GA Algorithm,for instance,it is easy to fall into local optimum and search slowly,the thesis uses penalty function method and hill climbing algorithm to improve the algorithm,and proposes a new Hill-Climbing Adaptive Genetic Algorithm(HAGA).Considering that the optimization problem is a Non-deterministic Polynomial(NP)problem,the thesis decouples it into two sub-problems: channel matching and power allocation.The simulation results show that the proposed resource allocation scheme based on HAGA has good convergence and search performance.2.Aiming at the perfect CSI is difficult to obtain in actual D2D-Noma system,which leads to the disturbance of uncertain parameters due to channel estimation error,this thesis further proposes an union robust resource allocation algorithm based on HAGA(Union Hill-climbing Adaptive Genetic Algorithm,U-HAGA).Specifically,the thesis considers the Imperfect Channel State Information(ICSI)of D2 D intra-group communication links,inter-group communication links and communication links between cellular user and D2 D recipients,the thesis firstly modeling the optimization problem as a nonlinear programming problem based on bounded channel uncertain parameters.Next the thesis uses the taylor series expansion to transform it into a deterministic optimization problem,then uses HAGA algorithm to evolutionise channel matching identification and power allocation factor respectively.Finally,it can obtain the optimal solution scheme through mutual iteration between channel matching and power allocation.The simulation results show the good robustness and search performance of the proposed scheme based on U-HAGA.