Research On Resource Allocation Strategy Of Heterogeneous Cellular Networks Based On Information Freshness

With the widespread use of smart terminal devices,new communication services are emerging and users’ needs for various services are rising,putting higher demands on the carrying capacity of cellular cells and urgently improving the spectrum resource utilization of base stations and expanding system capacity.Traditional single-layer cellular communication architecture can hardly handle excessive information.Heterogeneous networking technology forms a multi-layer communication network structure by adding small station such as relay,D2D(Device to Device)or micro base station to the underlying structure of macro base station,which shortens the distance between terminal equipment and base station and thus effectively increases the system capacity.This new network architecture is the Heterogeneous Cellular Networks(HCN).In addition,with the concept of Internet of Everything,communication terminals now have higher requirements for timely information,and how to improve spectrum utilization by correctly selecting channels and base stations while ensuring timely and effective information for end users has become an important issue for HCN.A new network performance evaluation index----Age of Information(Ao I)can better measure the freshness of information,which refers to the time elapsed from the moment of generation to the current moment of the latest received information packet at the destination node.Ao I outperforms previous metrics such as latency and throughput,and is able to dynamically describe the aging process of information packets at the destination node.In this paper,we study the timeliness of information acquisition in HCNs,and we focus on the performance of the average Ao I at the base station in HCNs and the resource allocation strategy for D2 D users in HCNs,respectively.The specific research work in this paper is as follows.First,the problem based on information freshness in HCN is studied.Two models of single-hop communication network and multi-hop communication network are established respectively according to the number of relays in HCN communication,and the information freshness problem of single-hop and multi-hop cellular networks in interference environment is studied by using the image property of Ao I,Taylor series theory and Shannon’s theorem deformation.The effects of different relay nodes’ forwarding methods on the message timeliness in multi-hop communication systems are investigated.The relevant factors affecting the system freshness are analyzed,and the relationship between the transmit power of cellular source node users,spectral efficiency constraints,the number of interferences,the usage of relays,the location of relays and the average Ao I of the system in the decoded forwarding and amplified forwarding methods is explored.Simulation verifies the results of the theoretical analysis,and the advantages and disadvantages of the two relay forwarding methods and the applicable scenarios are studied in comparison.Secondly,an improved resource allocation strategy of Augmented Stochastic Salp Swarm Algorithm(ASSSA)is studied for HCN to achieve timely delivery of information.A HCN model with D2 D multiplexing is established,and the optimization problem is built and solved with the transmit power,spectral efficiency constraints and the multiplexing rules of the channel as constraints and the average Ao I in HCN as the optimization objective.In order to improve the convergence rate stability and global searchability of the smart group sensing algorithm----Salp Swarm Algorithm(SSA),the SSA algorithm is improved,and the ASSSA algorithm with the addition of Cauthy variance and random parameter selection strategy is proposed,and the ASSSA algorithm is applied to the channel and transmit power of D2 D users in the multi-hop HCN The problem of allocating resources such as channel and transmit power for D2 D users in HCN is proposed.Finally,a comparison simulation experiment with other optimization algorithms is conducted to verify that the proposed algorithm can reduce the average Ao I of the system better while converging more quickly.