Research On Energy Cost Minimization Problem In Heterogeneous Cellular Networks With Hybrid Energy Supplies

Owing to the exponential increase of smart-phones and bandwidth-hungry applications,wireless cellular networks(WCNs) have experienced immense growth in traffic load, which leads to much more energy consumption and greenhouse gas emission. In cellular mobile communications, using green energy and adopting heterogeneous networks have been recently promoted as a strategic shift for reducing energy consumption and greenhouse gas emission.In this thesis, we consider a green heterogeneous network, consisting of both macro base stations(mBSs) and pico BSs(pBSs), with hybrid energy supplies. In such a network, mBSs are powered by on-grid energy, whereas pBSs can use either green or on-grid energy supplies. As renewable green energy is often considered much cheaper than on-grid energy, a new interesting problem is how to minimize the total energy consumption cost of the network while guaranteeing users' quality of service(QoS) requirement and network performance for such a green heterogeneous network.We investigate two kinds of scenarios: Basic scenario only takes the access network into consideration; While complex scenario considers not only the access network but also the backhaul links when modeling the traffic and the power consumption of BSs. In both scenarios, we first formulate a constrained total energy cost minimization(ECM) problem,under the constraints of transmission power budget of all BSs, users' data rate demand,available bandwidth budget of each BS, limited energy supplies and so on. It can be shown that the ECM problems in both scenarios are NP-hard. Thus, it is desirable for us to design feasible solutions with low computation complexity to them.In the basic scenario, we decompose the ECM problem into two components: the user association(UA) problem and the optimal bandwidth allocation(OBA) problem, which are complementary to each other. Next, a novel OBA algorithm is proposed to minimize the power consumption of each BS under a given UA scheme. For the UA problem, we propose both a centralized and a distributed UA algorithm to find feasible and near-optimal solutions based on the OBA algorithm, making the efficient utilization of the green energy storage for energy cost minimization. In the complex scenario, we propose a centralized and a distributed algorithm that exploits the available context-aware information of the whole network and the available context-aware information of each BS's own, respectively, based on user association and re-association. Finally, a series of simulation experiments are conducted to verify the effectiveness of our algorithms.