Research On Energy-saving Technologies In Cognitive OFDM Systems

In this paper,the key energy saving technologies in cognitive OFDM systems are studied.As a promising technology in improving spectrum utilization,cognitive radio(CR)attracts much attention by researchers.Besides CR,OFDM is also an important research topic for next generation wireless communication systems.Consequently,cognitive OFDM system is a promising development direction of wireless communication.On the other hand,due to the added energy consumed for spectrum sensing,green communication in cognitive OFDM system is a critical problem to be focused on.The low-complexity resource allocation problem in cognitive OFDM system is firstly considered in this paper.Although there are some optimal algorithms to solve the problem maximizing system capacity under power and interference constraints,it is not feasible to apply these optimal algorithms into practical system because of the long computation time caused by the high complexity.In this paper,the successive cap-limited water-filling method with low complexity is proposed,by which the resource allocation maximizing system capacity can be obtained in a short time.Through this method,the original problem is partied into a system of successive sub-problems,each of which can be solved by a rapid cap-limited water-filling method.As the sub-problems being solved one by one,all the constraints are satisfied.The simulation results prove that the proposed method can improve computation speed by two magnitudes with less than 5%performance lost.Then we focus on the cognitive OFDM system with SIPT and the optimization objective is turned to system energy efficiency.Simultaneous information and power transfer(SIPT)technology can help the receiver to harvest energy from the received signals,based on the nature that electromagnetic wave carries both information and energy.In this paper,SIPT with power splitting architecture is introduced into cognitive OFDM system.Our objective is to maximize the system energy efficiency,which is measured by "bit per Joule" metric,by optimizing power allocation and power splitting ratio.The fractional form of the energy efficiency function and coupling of power allocation and power splitting ratio cause the non-convexity of the optimal problem.Through fractional programming and approximation,the original problem can be turned into an equivalent convex problem and solved by convex algorithm.The simulation results prove that the proposed method can approach optimal and exceed other suboptimal methods.Next,the cognitive OFDM system utilizing cyclic prefix for SIPT is studied.In traditional OFDM system,the cyclic prefix,which is used to eliminate inter carrier interference,is directly discarded by the receiver,causing a lot of waste of energy,in the view of green communication.In this paper,the cyclic prefix is used for energy harvesting and a novel SIPT architecture is applied in cognitive OFDM system.Aiming to maximize system energy efficiency,the optimal power allocation and cyclic prefix size are designed under the condition that the energy harvested from cyclic prefix should be sufficient for information decoding consumption.As the cyclic prefix size will impact the interference for primary users,it is hard to obtain the close expression of the influence on energy efficiency caused by the cyclic prefix size.Consequently,the optimal cyclic prefix size is obtained by a full search method.The power allocation problem with fixed cyclic prefix size can be turned into convex and solved easily.The simulation results prove that the proposed method can achieve higher performance compared with the traditional OFDM system.