| In the heterogeneous cloud radio access network(H-CRAN),the remote radio heads(RRHs)can provide high data rates to user equipment in hot spots,while heterogeneous macro base station supports seamless network coverage,and achieves the separation between data information and control information.At the same time,centralized baseband processing in H-CRAN can suppress the intra-layer interference between RRHs as well as the inter-layer interference between RRHs and macro base station.Energy harvesting technology can harvest energy in surrounding environment,such as solar energy,wind energy,etc.,and convert them into electric energy.In order to meet the growing energy demand,based on the energy harvesting technology,a hybrid energy suppling scheme with multiple energy sources is adopted in H-CRAN,which can reduce the power consumption of traditional power grid.Furthermore,how to allocate resources such as power effectively and carry out energy management to realize the tradeoff between data rate and cost of system is an important problem to be solved.Therefore,for the H-CRAN with hybrid energy sources,namely energy harvesting and power grid,considering whether the capacity of fronthaul is limited and RRHs are equipped with batteries or not,the system model is established to maximize the cost efficiency of system,and the power allocation,user association and energy management of system are studied.After the problem formulated,convex optimization method is used to solve it,and the simulation verification and analysis are carried out.The main works of this thesis include:1)In the scenario of H-CRAN with ideal fronthaul,in which RRHs are not equipped with batteries,a cost-effective hybrid energy suppling model of H-CRAN is established.In this model,RRHs obtain renewable energy from energy harvesting,and make up the remaining energy gap by power grid.The energy sharing mechanism between RRHs can improve the utilization rate of renewable energy,while the macro base station is only powered by grid.The capacity of fronthaul is not limited,and the minimum data rate requirements of user equipment and the upper limit of system cost are considered,thus we take the user association status,power of transmission signals and energy managements of RRHs as optimization variables,and formulate the optimization problem to maximizing the cost efficiency of system.The objective function of original problem is in fractional form,and there are discrete integer variables in the optimization variables,so the original problem is a mixed integer nonlinear programming(MINLP)problem,which is nonconvex.The original problem can be transformed into a convex problem equivalently and solved by Lagrange dual decomposition method.The simulations show that,compared with baseline strategies,the hybrid energy suppling strategy of H-CRAN can improve the cost efficiency of system significantly.2)In the scenario of H-CRAN equipped with batteries in which the capacity of fronthaul is limited,considering the influences of batteries on the energy states of RRHs and system performance in subsequent timeslots,the resource allocation and energy management strategy of H-CRAN with hybrid energy sources in multiple timeslots is studied.Aiming to maximize the cost efficiency of system in each timeslot,the formulated problem is a MINLP problem,which can be transformed into convex problem and solved eventually.The simulation results show that the proposed optimization strategy could maximize the cost efficiency of system on the premise of considering the minimum data rate requirements of users,the upper limit of system cost and the limits on the capacities of fronthaul and batteries.Compared with other baseline strategies,the proposed strategy can reduce the cost of system significantly,thus obtaining the optimal cost efficiency.Based on simulations,the thesis further focuses on the influences of unit cost of battery storing,battery capacity and fronthaul capacity on system performances. |
PDF Full Text Request