Transmitter Optimization For Physical Layer Security With Multicast Service

Recently,physical-layer service integration(PHY-SI),a technique of combining multicast service and confidential service into one integrated service for one-time transmission at the physical layer,has received much attention in wireless communications.Compared with the conventional upper layer-based approach,PHY-SI enables coexisting services to share the same resources by solely exploiting the physical characteristics of wireless channels,thereby significantly increasing the spectral efficiency.However,due to the coupling of services,service messages may interfere with each other.Hence in PHY-SI,a crucial problem lies in how to establish the security of confidential service while not compromising the quality of public service.To resolve this problem,this thesis first puts forth the concept of secrecy rate region maximization(SRRM).That is,a biobjective problem is formulated to find the method for simultaneously maximizing the multicast rate and secrecy rate.Based on this concept,this thesis investigates the following problems.1)First,this thesis considers artificial noise(AN)-aided transmitter optimization for multi-user multi-input single-output(MISO)systems with PHY-SI.The goal of this thesis is to jointly design the optimal input covariances for the multicast message,confidential message and AN,such that the achievable secrecy rate region is maximized subject to the sum power constraint.This SRRM problem is a nonconvex biobjective maximization problem.To handle it,the thesis reformulates the SRRM problem into a provably equivalent scalar optimization problem and proposes a searching method to find all of its Pareto optimal points.The equivalent scalar optimization problem is identified as a secrecy rate maximization(SRM)problem with the quality of multicast service(QoMS)constraints.Further,the thesis shows that this equivalent QoMS-constrained SRM problem,albeit nonconvex,can be efficiently handled based on a two-stage optimization approach,including solving a sequence of semidefinite programs(SDPs).Moreover,the thesis also extends the SRRM problem to the case with multi-input and multi-output(MIMO).This thesis proved that the aforementioned method of scalarization also applies to the MIMO case.As for the resultant scalar problem,a suboptimal difference-of-concave(DC)algorithm is proposed to seek a lower bound on the rate region,which converges to a stationary point.2)Considering the imperfection of practical channel estimators,this thesis further allows for the transmitter optimization issues with imperfect channel state information(CSI).Two robust SRRM formulations are investigated,i.e.,worst-case SRRM formulation and outage-constrained SRRM formulation.For the worst-case SRRM formulation,the thesis shows that the worst-case SRRM problem can be handled in a similar manner as that in the perfect CSI case.The resulting scalar problem is nonconvex and semi-infinite,this thesis elaborates upon how to transform it into a convex and finite problem by using a two-stage reexpression and S-procedure.On the other hand,the outage-constrained case is more challenging to solve due to the existence of probabilistic outage constraints.Nonetheless,it is shown that an approximate solution can be efficiently computed by invoking the Bernstein inequality approach.This approximate solution is bound to fulfill the original outage constraints,and thus safe(conservative).For implementation efficiency,the rank properties of the obtained solutions,as well as computational complexity of the proposed methods are also evaluated in this thesis for our considered two robust SRRM formulations.3)Finally,to alleviate the increasing need for energy in wireless communications,this thesis studies energy efficient optimization problems emerging in PHY-SI.Two different fundamental tradeoffs are characterized.The thesis first studies the tradeoff between the secrecy energy efficiency(SEE)and the spectral efficiency,i.e.,how to maximize the SEE with QoS constraints.Due to the nonconvexity of this problem,an equivalent parametric reformulation,based on the fractional program and DC program,is proposed to recast the problem as a sequence of convex problems.By this means,the maximum SEE can be found via a root search algorithm.Next,the tradeoff between SEE and multicasting energy efficiency(MEE)is considered,the goal of which is to seek the transmit design that maximizes the EE region.Such an EE region maximization problem is a nonconvex biobjective maximization problem,and is an extension of SRRM to energy efficiency.A method of scalarization is utilized to reformulate it into a single-objective optimization problem.Despite the nonconvexity,the thesis shows that this problem can be iteratively solved by combining fractional programming and DC programming methods again.