Research On Dynamic Interference Coordination Scheme Based On Effectiveness And Fairness

With the increment of the number of mobile terminals and the user requirement on wireless transmission rate,the 5th generation mobile communication system puts forward a higher request for transmission rate and connection density.Due to the scarcity of the spectrum,the same frequency resource should be reused among adjacent cells to improve the system's spatial capacity,which unfortunately causes severe inter-cell interference in the meantime.Dynamic inter-cell interference coordination(ICIC)has been studied in depth thanks to its high flexibility,rapid response,and obvious coordination effect.With dynamic ICIC schemes,resource blocks in each cell could scheduled aiming at reducing interference,which improves the system's capacity or fairness.The user on the best status is allocated to the worst resource in the scheme of Max-Min fairness,which could be considered as the fairest strategy.But according to our research in the past,users close to the adjacent base station may cause severe interference to the users in adjacent cells,which may severely affect their transmission even though the users are with the best useful signal.Therefore,resource allocation(RA)with Max-Min fairness is never fair enough for those users who are allocated to severely interfered resource blocks.To make the allocation fairer,an ultra-fair dynamic ICIC framework is proposed in this dissertation,each interference coordination period is divided into several time slots and optimization in each slot is conducted on the basis of prior slots.In this way,we could adjust all the users' obtained resources to be consistent ultimately,hence achieving a fairer allocation than the scheme with Max-Min fairness.Another ICIC problem studied in this dissertation is for the scenario of massive and dense mobile terminals and Internet of Things equipments.Io T applications demand that the future mobile communication system should support the access of massive and dense equipments simultaneously and allocate resources rapidly.Due to high time complexity,traditional RA scheme are not suitable for such a scenario,making it important to look for efficient RA schemes.Compressive sensing(CS)technique could reduce the dimensionality of a problem by compression and recover it by nonlinear reconstruction algorithms,so the scheme of RA using CS is studied.Firstly,the feature of users and resources are comppressed by a measurement procedure.Then,the allocation process is conducted in the compressed domain.Finally,the allocation result in the compressed domain is reconstructed to be the RA strategies as a traditional RA scheme obtains.According to our simulations,the time cost of the CS-based RA scheme decreases significantly without obviously degrading the system performance.