Joint Time And Power Allcation For Relaying Networks Using SWIPT

Conventional relay networks are usually powered by batteries with limited capacity.It is of-ten impossible to replace batteries frequently in many harsh or special environments,which seriously restricts the communication quality and the lifetime of devices.Consequently,si-multaneous wireless information and power transfer(SWIPT)technology was introduced into conventional relay networks to prolong the lifetime and improve the communication quality of networks.Consider the existing resource allocation schemes based on SWIPT relay networks do not fully utilize the time or power resources,this work investigates the optimal joint time and power allocation schemes under the statistical channel state informa-tion(CSI)and instantaneous CSI scenarios according to the available CSI at relay nodes.For the statistical CSI scenario,we derive the analytical expressions of outage probability and ergodic capacity with two relay protocols,namely,decode-and-forward protocol and amplify-and-forward protocol,and further analyze the effect of time allocation(TA)and power splitting(PS)coefficients on the performance of the system.Under the decode-and-forward protocol,the closed-form expressions of outage probability and ergodic capacity are derived for arbitrary signal to noise ratio by means of Gauss-Chebyshev and exponential integrations.And the closed-form expressions with high signal to noise ratio are obtained by a certain approximation method under amplify-and-forward protocol.Based on the closed-form analytical expressions,the optimal TA and PS ratios can be determined offline to design the optimal joint time and power allocation schemes.For instantaneous CSI scenario,in order to design the optimal joint time and power alloca-tion schemes,we formulate two joint optimization problems to minimize the outage proba-bility and maximize the instantaneous channel capacity,and both of the multi-dimensional joint optimization problems of TA and PS coefficients are non-convex which can not be solved directly.In the decode-and-forward relay protocol,the two-dimensional optimization problem is transformed into a one-dimensional optimization problem by using information causality and the existence of a unique solution in the feasible region is proved for minimiz-ing the outage probability problem,and then the split-step iterative method is used to solve the problem.For the maximum instantaneous channel capacity problem,the maximization of the minimum rate problem is transformed into the maximizing achievable rate of the sys-tem by introducing slack variables.It is proved that the original optimization problem is a convex problem for another variable when any variable is fixed,and an alternate convex optimization method is employed.In order to provide analytical solutions for optimal values under the amplify-and-forward relay protocol,we consider the tight upper bound of the two optimization problems to obtain the approximations for optimal TA and PS coefficients.Finally,simulation results are provided to verify the analytical results and compare the pro-posed joint time and power allocation schemes with the existing schemes.It is shown that the proposed joint allocation schemes can effectively reduce the network outage probability or increase the system capacity comparing with the existing resource allocation schemes.