Research On Resource Allocation Of Cognitive Wireless Sensor Networks Based On SWIPT/NOMA

With the rapid development of 5G,wireless sensor networks will be more widely used in the future.Meanwhile,wireless sensor networks are confronting with problems such as limited energy and spectrum shortage,which restrict their further development.Using cognitive radio technology,secondary users can perform spectrum sensing and analysis,dynamically access the licensed spectrum of primary users,and improve spectrum utilization.Using simultaneous wireless information and power transfer technology,we can provide energy to wireless devices while sending information to them,so the lifetime of wireless devices is extended.Using non-orthogonal multiple access(NOMA)technology,multiple users can use the same channel to send information at the same time,thus we can use limited spectrum resources to achieve more user access,and improve spectrum efficiency.Therefore,this thesis studies the resource allocation strategy of cognitive wireless sensor network(CWSN)based on SWIPT technology and NOMA technology.On the basis of problem formulation and solution,the proposed resource allocation strategies are verified by simulation and analysis.The main works include:(1)In order to alleviate the problem of limited energy in wireless sensor network,this thesis establishs a CWSN network model based on SWIPT technology.The system model consists of a primary system and a sensor subsystem.The primary system includes a primary transmitter and a primary receiver.The sensor subsystem is composed of multiple sensor nodes,using the idle spectrum of the primary system user to send data.The sensor nodes in the sensor subsystem act as a relay,and decode and forward(DF)the signal of the primary transmitter to realize the information transmission of the primary user.Moreover,each sensor node is equipped with a power-splitting receiver and a rechargeable battery,which can harvest energy from the signal of the primary transmitter in a certain ratio to charge its own battery.For a CWSN network based on SWIPT,this thesis studies the resource allocation strategy,aims at maximizing the sum-rate of the sensor subsystem.Under the premise of guaranteeing the transmission rate of the primary system,an optimization problem is formulated.We solve the problem using convex optimization theory,and obtain the policies of the optimal power-splitting ratio,power and subcarrier allocation.Then,the optimization algorithm is simulated and can draw the conclusion that the sensor can obtain more available energy by using SWIPT technology for energy harvesting,thus the sum-rate of sensor subsystem is improved.(2)In order to alleviate the problem of spectrum shortage in cognitive wireless sensor network,this thesis establishes a CWSN network model based on NOMA technology.In this model,the primary system includes a primary transmitter and a primary receiver,which use licensed spectrum,and the primary receiver can directly receive the signal sent by the primary transmitter.The sensor subsystem consists of multiple groups of sensors,and each group of sensors uses NOMA technology to multiplex one subcarrier of the primary user.Due to sharing the same subcarrier,there is mutual interference between the primary user and the sensors.Since different sensor groups use different subcarriers,there is no inter-group interference.In the cognitive wireless sensor network model based on NOMA technology,on the premise of satisfying the interference constraint of the primary system and the transmission rate requirements of each group of sensors,this thesis formulates an optimization problem aiming to maximize the energy efficiency of the sensor subsystem.By using fractional programming,time-sharing strategy and successive convex approximation,the non-convex optimization problem is transformed into a convex problem.Using the Lagrangian dual method and KKT conditions,the power and subcarrier allocation strategies are obtained.On the basis of formula derivation,numerical simulations are carried out.Through simulation,we can see that our algorithm can maximize the energy efficiency of the sensor subsystem,and can achieve higher energy efficiency compared with the comparison strategies of orthogonal frequency division multiple accesses,fixed power allocation and fixed subcarrier allocation.