Research On Key Technologies Of Multi-connectivity In 5G Networks

Compared to 4G,5G has obvious advantages in system deployment flexibility,multi-service support,spectrum efficiency,peak rate and delay.It provides users with much more immersive experiences,such as UHD video,VR,AR,cloud desktop,mobile cloud and more.Due to the close relationship between 5G NR and 4G LTE at this stage,the rapid growth of data volume and the difference in performance requirements of users,5G needs to form a multi-standard network with existing 4G and other wireless supplementary networks to achieve "complementary interoperability"system.From the user's point of view,with the gradual development and maturity of the terminal multi-standard technology,they can freely select multiple network standards that match the service characteristics and personal preferences according to the current network environment and service type.From the perspective of operational maintenance and performance indicators,continuing to use a single network connectivity will not be able to adapt to the diverse and ultra-high access requirements of 5G.Therefore,the multi-connectivity that can make operators be easy to manage multiple networks in an integrated manner,ensure smooth switching between multiple networks,and provide a good user experience has become a key technology to be considered when constructing a future 5G network architecture.Although the 5G multi-connectivity is one of the effective solutions to improve the user experience and integrate multiple networks,it has no standardization and technical evaluation on key issues such as network adaptation and resource allocation,and still faces the challenges of key technologies.Based on the practical problems in ultra-dense scenario,multiple access scenario and D2D offload scenario in 5G HetNet networks,with the goal of improving 5G HetNet network adaptability and resource utilization,this paper models several issues with practical significance and studies how to solve these issues by some effective strategies,which makes this paper has certain theoretical and practical value.The main research of this paper focuses on the on-demand provisioning and adaptation of network connectivity and network functions,as well as real-time perception and provisioning of network resources.The specific research contents and contributions mainly include as follows:(1)Aiming at the rapid change of the received signals of users in ultra-dense scenes and the imbalance of small cell load,this paper proposes a network selection strategy based on load sensing.Unlike the fixed and static network selection strategies proposed in most existing literatures,our strategy gets rid of the constraint of the maximum number of connections on the network selection.First,the strategy filters the predetermined network candidate set according to different QoS requests of UE,and secondly dynamically adjusts the number of network candidate sets according to the S function formed by the network load.The finally determined user activity set can not only dynamically adjust the maximum number of network connections but also offload the flux on high-load networks to low-load networks.The simulation results show that,compared with the performance of the network selection algorithms proposed by most literatures,the proposed load-aware network selection strategy improve the system throughput significantly and effectively reduce the probability ofradi link failure.As the standard of 5G multi-network converged network architecture is still under exploration,in order to meet the different requirements of different users for data transmission rate,bit error rate,delay,jitter and other QoS parameters,firstly,several new multi-connectivity network architecture are proposed in this paper.The architecture upgrades the load from the traditional MAC layer to the higher layer to reduce the delay backhaul requirements.Secondly,the data offload transmission problem is transformed into the network selection and adaptation problem,and the service-aware network selection strategy is proposed according to the different mappings of different service QoS requirements on each access network.This strategy combines the characteristics of the network side and the user side to provide theoretical guidance for the actual deployment of multiple connections,which makes it has certain practical significance.(2)D2D offloading technology can offload cellular network traffic by mitigating base station load to alleviate downlink transmission pressure of cellular network,but it still has some problems in resource allocation and mode selection.Therefore,firstly,the paper proposes a user D2D mode and a cellular mode conversion criterion,which allows the user to adjust the working mode according to the service priorit dynamically.Secondly,for the problem of D2D and cellular user spectrum sharing,a strategy of resource assignment based on user priority is proposed.Different from the fair competition resources of D2D users and CUs in most traditional literatures,the proposed strategy allocates network resources according to the user's priority,and models the actual problem as a conditional convex optimization problem.Finally,the paper designs a resource allocation scheme based on Kuhn-Munkres algorithm to find the optimal solution.The simulation results show that the proposed resource allocation strategy based on user priority can significantly improve D2D access rate and system throughput.(3)An effective resource allcocation algorithm is one of the key technologies that can meet the performance requirements of 5G multi-network fusion.As the research on the allocation of uplink resources in eLWA scenarios is still in the blank,this paper proposes a resource allocation optimization algorithm for delay quality of service guarantee.At the same time,considering the five factors of user actual rate,demand rate,spectrum efficiency,BS load and fairness in ultra-dense network,this paper also proposes a resource allcocation algorithm based on PF.In the resource allocation optimization algorithm for delay quality of service guarantee,this paper focuses on the network model that the user's multiple services transmit data information simultaneously through the uplink channel of the LTE network and WLAN network.Unlike the resource optimization strategy in most literatures that maximizes system capacity,this paper uses delay quality of service as a channel characterization of eLWA network for the first time,and obtains the uplink resource allocation optimization strategy by the total effective capacity according to the maximum delay quality of service guarantee.The specific ideas are as follows.First,based on the constraints of delay guarantee when users have multiple service requests,this paper constructs a problem of maximizing effective capacity.Second,the problem is transformed into two sub-problems:power allocation optimization and bandwidth allocation optimization.Regarding the power allocation optimization problem,this paper adopts the convex optimization theory and the Lagrangian function to obtain the method of power allocation optimization.After determining the proportion of power allocation,paper obtains the method of bandwidth allocation optimization based on convex optimization theory and subgradient method,and then obtains the overall resource allocation optimization strategy.Finally,the simulation results show that compared with the common ABC(Always Best Connected)power allocation algorithm and the fixed ratio power allocation algorithm,the proposed algorithm can effectively improve the effective system capacity under the premise of delay quality of service guarantee.In the ultra-dense network,some small cells provide services for multiple users at the same time,while some small cells are idle.Therefore,the paper first proposes a new network architecture based on the characteristics of ultra-dense networks,and establishes a complete set of resource scheduling signaling procedures.Secondly,unlike the algorithms that only consider single-sided parameters in most existing literatures,the paper considers the combination of network side parameters(such as load,signal to noise ratio)and user side parameters(such as QoS request,priority,etc.)to solve resource allocation problems.It not only ensures the fairness of each UE in the system,but also promotes the load balance of the whole system.Finally,the simulation results show that,compared with the traditional PF resource allocation algorithm,the average user throughput of the proposed resource allocation strategy based on load-aware increased by 20%approximately.