Key Technologies For Dense Transmission And Resource Allocation In Smart-Duplex Network

With the rapid development of wireless communications,the demands of data transmission rates in future networks explode to meet the emerging new intelligent services.Ultra-dense network(UDN),which aims to deploy a large number of small cells with low-power access points(APs)in the target area,is regarded as one of the key technologies to achieve seamless coverage,high-speed communication,and massive access in future networks.However,the density of APs in UDN results in irregular small cells' structures,which also makes the interference distribution in the coverage area more complex.In order to achieve the higher speed,lower latency,and lower power consumption for the network designing,it is necessary to fully utilize the wireless resources in dimensions of space,time,and frequency to achieve the high-density wireless transmission.Based on the idea of dense transmission,Based on the idea of dense,this dissertation combines UDN and smart duplex(SD)to study the key technologies for dense transmission and resource allocation in SD Networks.First,to meet the strong co-channel interference in full duplex(FD)UDN,this dissertation proposes an amplify-and-forward(AF)protocol in-band receiver cooperations scheme,and also studies the capacity of FD interference channel.In particular,the FD AF protocol is adopted to build up the receiver cooperations: each receiver can simultaneously receive the signals from the two transmitters and two receivers in the same frequency band,and after self-interference(SI)cancellation and message decoding,it forwards them to its counterpart.With the above scheme,the equivalent channel model is analyzed,and the statistics of the accumulated residual interference and noise(ARIN),generated by the imperfect SI cancellation and the AF scheme at the receivers,is calculated by mathematical induction.Then,the achievable rate regions of both the single-user and joint decoding schemes are derived.It is proved that one-side cooperation,i.e.,only one of the two receivers forwards its counterpart's signal,is optimal to achieve the best system performance.Next,to characterize the obtained rate regions,the rate maximization problems are formulated and approximately solved by a sequential parametric convex approximation method.Simulation results show that the proposed scheme can improve the achievable rate compared to the conventional non-cooperative scheme in several typical scenarios.Then,considering the high processing cost in the SD UDN,this dissertation divides all small cells into several clusters based on the base station(BS)-centric copperative transmission scheme,and investigates the intelligent resource allocation algorithm to balance the system performance and clustering cost of the SD UDNs.In the considered SD UDN,the BSs flexibly switch between the half-duplex(HD)and FD modes according to the dynamic wireless environment.A Markov decision process(MDP)problem is formulated to maximize the average weighted sum of network throughput and clustering cost for all clusters.To approximately solve this problem,this dissertation first adopts an affinity propagation method to determine the number of clusters and the center of each cluster.Then,by treating small cells as agents,the original MDP problem is proved to be equivalent to a multi-agent MDP to maximize the average reward of all small cells.Next,a multi-agent deep reinforcement learning(DRL)is proposed to jointly implement the dynamic clustering for non-center small cells,resource allocation,and duplex mode selection.Simulation results show that SD has prominent advantages over both the HD and FD in UDNs,and the proposed multi-agent DRL outperforms other clustering schemes under the considered scenarios.Next,to meet the high-speed data transmission rates of a large number of users in UDN,this dissertation further proposes a dynamic multi-user association and power allocation based on a more flexible user-centric multi-APs cooperative architecture.Under the limited frequency and transmission power budget,each user flexibly accesses multiple APs according to its dynamic requirements,and the accessed APs form a small network to jointly serve the user.This dissertation investigates a multi-user access and power allocation problem in real time,and designs the reward function for each user based on its accessing costs and transmission rates.A MDP problem is formulated to maximizes the average user satisfied ratio of UDN in the long-term time scale by treating the APs as agents.To solve the MDP problem with large discrete action space,this dissertation proposes a multi-agent tree-structured policy gradient(MATSPG)method based on the tree-structured policy gradient recommendation architecture.It maps the large discret action space of the multi-user access and power allocation problem in the considered UDN to the tree-structured action space with two-layer non-leaf nodes.Next,this dissertation also proves that the proposed MATSPG algorithm has lower time and space complexity compared to other traditional DRL algorithms.Simulation results show that the proposed MATSPG algorithm has significant advantages in the problems with large discrete action spaces.This dissertation focuses on the issues of interference management,resource allocation and networking architecture in UDN.Under the dynamic wireless environment,the dense wireless transmission architectures and resource allocation algorithms are designed for SD networks,where their performances are also analyzed.The above studies can provide theoretical knowledge for the various intelligent applications of future networks.