Research On Physical Layer Security In Dense Heterogeneous Cellular Networks With Imperfect CSI

The broadcast nature of dense heterogeneous cellular network(HCN)causes secure issue of eavesdropping.Physical layer security(PLS)guarantees secure transmission in dense HCNs by using the uniqueness and reciprocity of wireless channel.Perfect channel state information is important premise of PLS.Due to the characteristics of random and dynamic arrangement of dense HCN nodes and complex interferences,imperfect estimation is apt to be found in the real communication scenarios.Otherwise,the limit of channel estimation algorithm,difference of nodes and indefiniteness of actual communication scenario may also give rise to imperfect CSI,which influence the PLS performance of the system.As topology of HCNs is changing dynamically,the transmission scheme in traditional networks can not be directly used in HCNs to ensure secure communication with imperfect CSI.Specifically,there are three main problems: 1)There is a lack of analytic model in PLS theory.2)Current PLS secure transmission scheme in centralized Multiple-input Single-output(MISO)network can not be directly applied for dense HCN.3)The secure transmission scheme in cooperative communication system cannot be directly used in dense HCNs.To solve this problem,this paper considers the dynamic topology change caused by the non-planned deployment in dense HCNs.First,we propose the analytic model in dense HCNs with imperfect CSI.Based on the this model,we design the artificial noise aided transmission scheme for MISO communication cellular,and coordinated multipoint(CoMP)aided transmission scheme for cooperating communication cellular.The details of our achievements can be summarized as:1)We proposed the analytic model in dense HCNs with imperfect CSI.Considering the dynamic topologies of dense HCNs,we modeled the system by stochastic geometry,and provided the relationship between the estimation channel and real channel.We deduced the security outage probability of K-tier dense HCNs to evaluate the secure performances of dense HCNs.Then,we analyzed the influence of partial parameters on system performance,including channel estimation accuracy,density of picocell base station,density of eavesdroppers,secure outage threshold and the number of antennas.Finally,the validity of theoretical derivation is verified by simulation.2)To ensure secure transmission in MISO communication cellular in dense HCNs with imperfect CSI,we provided an artificial noise(AN)based secure transmission scheme,and proposed a power split factor optimization algorithm.We applied the AN to interfere eavesdroppers considering the dynamic topologies of dense HCNs.Meanwhile,we considered the influence the system interferences and channel estimation errors.To increase the secrecy throughput of the system,we chose the value of power split factor under different channel estimation accuracy.To satisfy the security and reliability requirements of users,we put forward a power split factor optimization algorithm.The numerical results verified that the AN transmission scheme could increase secrecy throughput by about 15%.3)To ensure reliable and secure communication in cooperating HCNs with imperfect CSI,we proposed a CoMP transmission scheme based on dual-threshold optimization,and provided a dual-threshold optimization algorithm to further support the performance.The main idea of the scheme is to select corporation base stations(BSs)with good channel conditions for CoMP transmission.We provided candidate BSs formation policy and CoMP networking strategy to select corporation BSs,and decrease the influence of channel estimation errors.We put forward a dual-threshold optimization algorithm to satisfy the security and reliability requirements of users,and employed a two-dimensional search method to get the optimal results of the dual thresholds.Numerical results were presented to verify our theoretical analysis,where the proposed scheme could increase secrecy throughput by about 0.13 bit/s/Hz.