Research On Dynamic Power Control Based On Control Theory In Cognitive Radio Networks

Although wireless mobile communication technology has profoundly changed our life,but we never stop researching to mobile communication technology with higher performance.With the ever-increasing growth of mobile data traffic,the limited spectrum resources must become the bottleneck for the development of wireless communication.In order to deal with the problem of spectrum shortage,cognitive radio(CR)technology has been considered into future wireless communication systems.In cognitive radio network(CRN),by adjusting the transmission parameters according to the characteristics of the surrounding communication environment,the unlicensed user(secondary user)can share a common frequency band with licensed user(primary user)at different time or under the interference temperature constraint at the same time.This approach can effectively utilize the scarce spectrum resource.However,the form that the primary user coexists with secondary user will result in mutual interference between them.It is one of important methods to deal with the problem that design a reasonable power control algorithm.Power control is one of the key technologies to realize the desire that primary user and secondary users share the spectrum resources in CRN.The existing power control algorithms based on optimal theory for CRN,focusing on the current time slot,is actually a static power allocation mechanism.These algorithms need to build an accurate mathematical model of optimal problem with well-known parameters,and solve a set of optimal transmit power values by optimizing the given target function used to defining communication system performance.However,that the parameters have time-varying characteristics and randomness and there maybe exist measurement errors and feedback delays about them will cause a problem that communication performance is degraded or even interrupted when apply these algorithms in practice.Some algorithms that utilize robust optimization theory and stochastic optimization method to deal with the parameter uncertainty of power allocation in CRN have been proposed by some people,but the necessary upper bound of error and information of statistical model for all of negative factors are obtained difficultly.Considering the above problems,a series of power control algorithms for CRN based on system dynamic description will be proposed in this thesis,which do not rely on the exact system parameters and can resist some disturbances like measurement error and feedback delay.The main contributions of this thesis are summarized as follows.(1)Inspired by target signal to interference plus noise ratio(SINR)tracking power control algorithm in traditional wireless cellular network,the problem of power control for CRN is modeled as a discrete linear time-invariant system with exogenous disturbances input,represented as a state-space expression,where the dynamics of the channel gain of each link and the interference plus noise among secondary users are formulated as first order Markov process.Based on optimal control theory,a controlled indicator is defined by combining the interference temperature constraint and the communication quality of service(QoS)requirements of secondary user.For the dynamic system and the controlled indicator,a linear quadratic regulator(LQR)based power control algorithm with considering channel not time-varying is proposed to demonstrate the stability of system.And the effectiveness and superiority of the proposed algorithm is also proved by theoretical analysis and numerical simulation.(2)Considering the random variation of channel parameters in real wireless communication,the exogenous disturbances input and the measurement error of feedback information are assumed white Gaussian random process with known statistical information for the case with time-varying channel state information(CSI).On the basis of the assumption,a power control algorithm based on linear quadratic Gaussian(LQG)control is proposed for the above dynamic system expressed by state-space model.In addition,by analyzing the relationship between data transmission rate and SINR of secondary user,the operations that adaptively adjust weight parameter and switch interference temperature threshold protection margin are introduced into proposed power control algorithm to further improve the algorithm performance,especially the protection for primary user is strengthened.(3)Considering the problem that it is difficult to obtain statistical information about the exogenous disturbances input and the measurement error of feedback information,the two quantities are further assumed as random variables with limited energy.Then theH?control based power control algorithm for CRN is proposed according to robust control theory,where the processes to designH?state feedback controller and output feedback controller based on Lyapunov stability lemma are addressed in detail.The corresponding simulation results show that the proposed algorithm has better protection effect for primary user and better communication performance for secondary user.(4)It is inevitable that delay exists in signal transmission and command execution in the actual communication,thus the delay quantity is introduced into the modeled dynamic system.For the new system,the technology used to treat delay problem in robust control theory is adopted to obtain the power control algorithm based on delay-independentH?control for CRN,which can select to use one of delay-independentH?state feedback controller and output feedback controller according to communication requirement,the design processes of which are given in the thesis.Simulation results show that the proposed algorithm can reduce the influence caused by system delay on performance of power control.(5)The power control algorithm based on control theory proposed in this thesis draws lessons from the idea of power update of existing cellular network users,and continues its advantages of dynamic regulation,closed-loop control,low computational complexity,simple design,low requirements for hardware equipment,and large realizability.Comparing with the approaches based on optimal theory,it is an initiative and successful attempt to use control theory to solve the problem of power allocation for CRN in a dynamic way,the advantage of which is that the system can be described dynamically,the transmit power can be dynamically allocated according to practical communication situation and the algorithm is insensitive to initial conditions.It can enhance the protection effect to the primary users in the actual communication and restrain the influence of loop delay effectively.