Power Allocation Algorithm Based On Interference Temperature Limits In Cognitive Networks

As the rapid growth of wireless technologies, current spectrum scarcity has become a serious problem as more and more wireless applications compete for very little spectrum.Fixed spectrum allocation strategy is applied to current radio networks,which means that the spectrum management organization allocates special licensed bands for corresponding communication technologies.However, a survey from Federal Communication Commission of America shows that the usage of licensed bands are not adequate, and it is varying in different time, location and frequency bands.Consequently, in order to raise the usage rate of wireless spectrum,as a key technology of cognitive networks, a new technology, cognitive radio(CR) presents,so as to make some communication facilities adopting cognitive technology can operate on idle spectrum bands by automatically detection.In this scenario, interference control and QoS requirement became the key problems.In this thesis, Firstly, the spectrum detection is studied, and an enhanced energy detection algorithm is deduced based on traditional energy detection method firstly. Secondly, the mechanism of spectrum sharing under Interference Temperature Limits(ITL)of multiple systems in classic interference temperature limits model is analyzed, and a new ITL model is issued for cellular networks after combination of classic ITL model and cellular networks.Afterwards, a power allocation problem of multi-user is analyzed based on power control problem in multiple ITLs model with multiple users accessed.The real cellular networks evolve into an theoretical model,from which a non-linear Mathematical Program(MP)is deduced. Then, some solutions for non-linear MP with constraints conditions are discussed based on some common mathematical method for solving non-linear MP, and some features and mechanism are analyzed for Reduced Gradient Method(RGM) and Punishment Function Method(PFM).Considering RGM can control the computation accuracy and has acceptable convergence performance, so in this thesis RGM is adopted to solve power allocation non-linear MP problem.Finally, simulation results are developed, and some theoretical analysis is commented to these results.