Optimization Frameworks For NOMA-Enabled IoT Network Resources

The demand of social and economic development and technological advances are forcing communication technologies to undergo a transformation every decade,i.e.,cre-ating a new standard.We are now in the research phase of rolling out the fifth-generation(5G)communication technologies.The commercial deployment of 5G technologies has also been accelerated around the world,with industry and academia looking beyond 5G and conceptualizing 6G.The work of this thesis was started when 5G was in full swing,hoping to make a meager contribution to the current and the next generation of wireless communications.These technologies are expected to provide three primary use-cases such as enhanced mobile broadband(eMBB),ultra-reliable low latency communications(URLLC),and massive machine-type communications(mMTC).More specifically,the objective of eMBB is to provide high data rates across the extended coverage area,whereas URLLC restricts requirements on latency and reliability for mission-critical applications such as remote health-care equipment,autonomous driving or the tactile Internet.The role of mMTC is to support a very large number of wireless devices in a small area,which may only send data sporadically,such as Internet-of-things(IoT)use cases.As the latest member of the multiple access family,non-orthogonal multi-ple access(NOMA)technology proposed in this thesis is recommended as one of the promising technologies for high spectral efficiency and massive connectivity.Recently,NOMA has been proposed for 3GPP long term evolution(LET)and envisioned to an important technology of upcoming 5G and future 6G systems.The fundamental feature of NOMA technology is to allow multiple devices in the network to use the same frequency resources with a nominal increase in the receiver complexity.This thesis presents a systematic treatment of this newly emerging technology,from the basic principles of NOMA,to its combination with IoT network,Heterogeneous net-works(HetNets),backscatter communication,as well as enhancing the energy efficiency,spectrum efficiency and addressing the security issue.The main research results of this thesis are as follows:1.First,this thesis investigates a downlink and uplink transmissions of a multi-cell NOMA network,where the BS is located in the center of each cell and serves multi-ple randomly distributed users according to the power-domain NOMA principle.The objective is to maximize the sum-capacity of downlink and uplink NOMA transmission subject to minimum rate and maximum transmit power constraints.The problem of ef-ficient power allocation in both downlink and uplink is formulated as non-convex,which is very complicated and hard to solve.To obtain a suitable solution,a suboptimal power allocation scheme based on Karush-Kuhn-Tucker(KKT)conditions is exploited,where the Lagrangian multipliers are updated using sub-gradient methods.For comparison,we also present the network performance under the non-optimal NOMA scheme and tradi-tional orthogonal multiple access(OMA)scheme.The results are obtained using Monte Carlo simulations,which unveil that the proposed NOMA scheme converges within few iterations and significantly outperforms the non-optimal NOMA and traditional OMA schemes in terms of sum-capacity2.Second,the thesis describes a new resource allocation scheme for efficient user association and optimal power control in NOMA-enabled HetNets.The objective is to maximize the energy efficiency of the proposed HetNets while guaranteeing the successful signal decoding and minimum rate of each user.The problem of user association and power control is jointly formulated as a non-convex optimization problem.The proposed optimization problem is coupled over both user association and power control which contains high complexity,thus,it is very difficult to obtain the joint solution.To obt ain an efficient solution and reduce the complexity,the original problem is decoupled into two subproblems for efficient user association and optimal power control.For any given power allocation of BSs,we first adopt dual theory for user association,and then a new sequential quadratic programing(SQP)is employed to calculate the optimal power control.Later,the benchmark suboptimal power control method is also presented which is based on KKT conditions.Monte Carlo simulation results unveil that the proposed NOMA resource optimization scheme in HetNets can significantly improve the system performance compared to the benchmark NOMA and orthogonal multiple access OMA schemes.3.Third,this thesis provides a joint power allocation problem based on multi-objective optimization for downlink multi-user NOMA system.The objective is to maximize the spectrum and energy efficiency while guaranteeing the minimum user rate,ensures successive interference cancellation(SIC)decoding process at receiver side and limits the user transmit power at BS.The problem of joint spectrum and energy efficiency optimization is formulated and also proved as convex problem.To obtain an optimal solution,we adopt the dual theory based on KKT conditions where the La-grangian multipliers are computed using sub-gradient methods.Later,a low complexity single-objective optimization problem to maximize the spectral efficiency of NOMA sys-tem is also investigated which is treated as a benchmark power allocation scheme.We provide the results based on Monte Carlo simulations and compare the proposed joint power allocation scheme with benchmark NOMA power allocation as well as the tra-ditional OMA power allocation schemes.The results demonstrate that the proposed NOMA scheme achieves significantly higher spectral and energy efficiency than that of benchmark schemes4.Forth,our thesis presents a novel resource optimization framework for max-imizing the energy and spectral efficiency of the IoT networks using power domain NOMA.The proposed framework considers a limited number of spectrum resources in the IoT network and provides an optimal resource allocation methods.Different practi-cal constraints like SIC complexity,ensuring the minimum gap of received power among different IoT equipment over the same frequency block for successful SIC operation quality of services requirements,and IoT equipment's transmit powers have also been taken into account.Accordingly,a non-convex optimization problems have been for-mulated for resource management.For energy efficiency problem,we employ the new SQP approach to obtain an optimal solution.For spectral efficiency optimization where the objective is coupled by both frequency and power allocation.To effectively solve this problem,the resource optimization problem is decoupled into two subproblems for spectrum assignment and power allocation.A suboptimal algorithm has been designed for spectrum assignment and SQP approach is adopted to solve the non-convex power control subproblem.For the sake of fair comparison,a low complexity suboptimal NO-MA power allocation scheme,based on KKT conditions,is also provided.The results unveil that the proposed optimal resource management scheme significantly outperforms the suboptimal scheme in terms of the total energy and spectrum efficiencies of the IoT network.5.Fifth,this thesis proposes a multi-cell backscatter network where a base sta-tion(BS)in each cell communicates to cellular users using the power-domain NOMA technique.A backscatter node in each cell also receives the superimposed signal from BS,utilizes this signal to modulate data and,then,retransmit it to nearby cellular user in the presence of multiple eavesdroppers.The eavesdroppers in the vicinity may try to overhear the transmission of the backscatter node due to the broadcast nature of the wireless network.Therefore,we first investigate an optimization problem to maximize the secrecy rate of the NOMA-enabled backscatter network in single-cell scenario.Then we also present a multi-cell backscatter scenario,where secrecy rate in each can be max-imized through efficient reflection coefficient in the presence of multiple eavesdroppers.The optimization problems in both scenarios are formulated as convex problems that are subjected to the maximum reflection coefficient of the backscatter node.To obtain an optimal solutions,we exploit dual decomposition methods followed by KKT condi-tions where the Lagrangian multipliers are updated by the sub-gradient method.We also present the secrecy maximization problem under traditional time division multiple access(TDMA)for the sake of comparison.Finally,the results are obtained using the Monte Carlo simulation which demonstrates that the proposed NOMA scheme signifi-cantly outperforms the traditional TDMA scheme.