| With the deployment of the fifth-generation mobile communications,many new Internet of Things(IoT)applications such as smart cities,smart agriculture,and the industrial Internet,have become important infrastructures to support a modern information society.IoT is faced with many technical challenges such as ultra-reliable and low-latency communication,massive access,limited device energy,and limited spectrum resources.Compared with the traditional voice and video communications using long packet data,IoT mainly transmits short packet data such as sensing and control information.Therefore,the classical Shannon's theorem based on the assumption of infinite block-length cannot accurately describe the reachable rate and reliability of the communications in IoT.In 2010,Polyanskiy et al.deduced an analytical expression for the maximum achievable rate and reliability of signal transmission under finite block-length,providing a theoretical basis for the short packet communications(SPC)system in IoT to solve the above-mentioned challenges.This dissertation focuses on SPC technology combined with the non-orthogonal multiple access(NOMA)technology,the energy harvesting technology,and the intelligent reflecting surface(IRS)technology,and then studies the resource allocation for IoT SPC systems.Specifically,by establishing a performance metric for the tradeoff between the transmission rate and the reliability,this dissertation provides algorithm designs for the joint allocation of various wireless communication resources in the time domain,the frequency domain,and the space domain,to achieve high reliability,low latency,massive connections,low power consumption,and high spectrum-efficiency for IoT communications.The main contributions of this dissertation are summarized as follows:1)Resource allocation for NOMA-SPC systems: NOMA technology superposes multi-user information in time-/frequency-/code-domain,which can improve the number of access devices and spectrum-efficiency,thereby solving the problems of massive access and spectrum scarcity for IoT.Therefore,this dissertation studies the multi-domain resource allocation for the multi-user multi-carrier NOMA-SPC systems.Specifically,by taking effective-throughput as the performance metric to balance the transmission reliability and the rate,the weighted effective-throughput is maximized under the fairness criteria and the transmission reliability constraints by optimizing user pairing,subcarrier allocation,transmission power,and transmission rate.Then,an optimal resource allocation algorithm based on dynamic programming is proposed to explore the best throughput performance of the studied system,and a sub-optimal resource allocation algorithm with low computational complexity is developed to reduce the requirement of the computing ability of IoT devices.Simulation results show that the performance of the proposed suboptimal algorithm can approximate that of the optimal algorithm and outperforms those of the traditional algorithms under the assumption of infinite block-length,which demonstrates the effectiveness of the proposed algorithm.2)Resource allocation for energy harvesting based SPC systems: Wireless energy harvesting technology can harvest radio-frequency signal from the surrounding environment and transform it into electrical energy for information transmission.This technology can be used to solve the limited energy problem of IoT devices,thus this dissertation studies the time domain resource allocation for the energy harvesting based SPC systems.By defining the effective-throughput and effective-information-amount as metrics to measure the successful transmission information-rate and information-amount of SPC,and through optimizing the energy harvesting time,user information transmission time,and user transmission reliability,the system effective throughput maximization problem was studied under the constraints of the limited transmission time and the maximum transmission error probability.Also,the total transmission time minimization problem was formulated subject to the constraints of the minimum effective-information-amount requirement and the maximum transmission error probability of each user.Since the time optimization variable is an integer variable related to packet length,this dissertation uses the block coordinate descent method,the successive convex approximation method,and the continuous-variable integer method to solve the original problem.Simulation results show that the performance of the proposed algorithm is close to that of the optimal exhaustive search algorithm and better than the traditional resource allocation algorithm under the assumption of infinite block-length.3)Channel estimation for IRS-assisted SPC systems: IRS can change the wireless signal propagation environment by adjusting its reflection coefficients,thus the wireless signal at the intended receiver can achieve signal combination or interference cancelation.Hence,this technology is one of the key technologies to improve the energy-efficiency and spectrum-efficiency of IoT.In this dissertation,channel estimation is studied for IRSassisted multi-user SPC systems,which will be the foundation of the associated resource allocation design.Specifically,a novel pilot transmission frame structure is proposed,and then a sparse channel model is studied to transform the original channel estimation problem into a sparse matrix recovery problem,which can be solved by a single-user channel estimation algorithm based on the matching pursuit principle.In order to further reduce the training overhead,a multi-user common row-column-block sparse model was studied and a multi-user joint channel estimation algorithm was proposed based on common subspace projection.Simulation results show that the proposed algorithm can achieve high estimation performance with limited training consumption and outperform the traditional channel estimation algorithm.4)Resouce allocation for IRS-assisted SPC systems: Following the research on the channel estimation for IRS-assisted multi-user SPC systems,this dissertation studies the space resource allocation design.By optimizing the transmission reliability of each user,active beamforming at the base station,and passive beamforming at the IRS,the maximization problem of the minimum effective-throughput of all users is formulated subject to total transmission power,transmission reliability,and IRS discrete reflection coefficient constraints.Then,a spatial resource allocation algorithm based on block coordinate descent and successive convex approximation principles is proposed to efficiently solve this non-convex optimization problem.Simulation results illustrate that using IRS can significantly improve the transmission performance of the SPC system,and the performance of the proposed algorithm is significantly much better than those of other heuristic algorithms.In addition,the performance of the proposed algorithm under 3 bits quantization reflection coefficients can approximate the performance of the continuous reflection coefficients,which guides the practical applications of the IRS-assisted SPC systems. |
PDF Full Text Request