Research On Interference Management And Resource Allocation In D2D Cellular Network Based On Stochastic Geometry

With the development of multi-media services,the explosive growth of mobile data is an indisputable fact.In order to alleviate the pressure of current cellular networks,D2D(Device-to-Device)communication technology has been widely used in traditional cellular networks due to the characteristic of short-distance direct communication and the advantage of multiplexing spectrum resources.However,how to perform effective interference management and resource allocation in D2 D cellular networks remains one of the key issues that plague researchers.Meanwhile,with the trend of denser,randomized and complicated network nodes,the traditional regular hexagon grid model is no longer applicable.Based on this,the stochastic geometric model is employed in this paper to model and analyze the network nodes.By combining cognitive radio and edge computing technologies,the interference environment in cognitive D2 D cellular networks and edge computing D2 D cellular networks are studied in this work,respectively.Corresponding interference management and resource allocation schemes are also proposed.The specific research content and main contributions are as follows:First,a cognitive interference coordinate strategy based on the beacon constraint and the CSMA/CA protocol is proposed for the scenario of sharing the underlying spectrum in the D2 D cellular network.During the sensing stage based on the beacon constraint,the cognitive D2 D transmitters judge whether they could become candidate D2 D users by detecting the beacon signal on the control channel,which determines those who can share the cellular uplink channel resources.To avoid severe co-channel interference caused by random selection of subchannels between D2 D users and adjacent D2 D users,a regional CSMA/CA channel competition mechanism is further adopted,which requiring all adjacent candidates to compete for the same subchannel in a backoff manner.The network node positions are all modeled as Poisson point process models with different densities.By applying stochastic geometric methods,the access probability of cognitive D2 D transmitters and the uplink probability of cellular users and D2 D users are derived in this paper.Simulation results show that the proposed interference coordination strategy can efficiently decrease the co-layer and cross-layer interference,thereby effectively improving the outage performances of the D2D-cellular network.Then,a novel performance analysis framework is proposed for the D2 D cellular network with MEC.Assuming that the AP(Access Point)nodes and D2 D nodes on the edge of mobile users allhave computing power,mobile users can offload their computation tasks to them according to the corresponding task offloading modes.In order to further reduce the co-channel interference between cellular links and D2 D links,a prioritized channel access policy is adopted,in which the multi-channels either for uplink or downlink are divided into two sets: the dedicated set for cellular links only,and the shared set for both cellular links and D2 D links.Besides,cellular links first select dedicated channels for communication,and shared channels are allocated to cellular links only when all dedicated channels are occupied.The successful edge computing probability is defined as the probability that a mobile user successfully finished the task computation and completed the communication transmission.Based on stochastic geometry and queuing theory,the successful edge computing probability of mobile users is derived,which can simultaneously reflect the communication performance and computation performance of the network.Simulation results show that compared with traditional random access,the priority channel access policy can effectively reduce cross-layer interference and increase the communication probability of users,thereby improving the successful edge computing probability and the quality of user edge computing.In addition,the relationship between the density of D2 D nodes and the distance threshold of the offloading mode reveals the tradeoff of system performance.