DAS is a promising architecture towards 5G networks. Under the DAS architecture, all the
signal processing tasks are executed in the centralized processing unit (CPU), and the remote
access units (RAUs) are only responsible for simple signal transmission and reception. Hence,
the system is more amenable for centralized signal processing, such as Coordinated Multiple
Points Transmission/Reception (CoMP), joint user scheduling and flow control, etc. In addition, due to its simple functionality, its size is smaller than that of the conventional small cell BSs and more densely deployed in the cell with low operation cost. Furthermore, due to the recent development of cloud computing, DAS architecture is an ideal platform to support some emerging information technologies, such as NFV, SDN and AI. This dissertation mainly focuses on the transmission design for DAS, with the aim of improving the spectral efficiency and energy efficiency of DAS.
Firstly, a transmit covariance optimization method is proposed to maximize the energy efficiency for a single-user distributed antenna system, where both RAUs and the user are equipped with multiple antennas. Unlike previous related work, both the rate requirement and RAU selection are taken into consideration. Here, the total circuit power consumption is related to the number of active RAUs. Given this setup, we first propose an optimal transmit covariance optimization method to solve the EE optimization problem under a fixed set of active RAUs. More specifically, we split this problem into three subproblems, i.e., the rate maximization problem, the EE maximization problem without rate constraint, and the power minimization problem, and each subproblem can be efficiently solved. Then, a novel distance-based RAU selection method is proposed to determine the optimal set of active RAUs. Simulation results show that the performance of the proposed RAU selection is almost identical to the optimal exhaustive search method with significantly reduced computational complexity, and the performance of the proposed algorithm significantly outperforms the existing EE optimization methods.
Secondly, a jointly selecting the fronthaul links and optimizing the transmit precoding matrices method aiming at maximizing the energy efficiency of a multiuser multiple-input multipleoutput aided distributed antenna system is proposed. The fronthaul link’s power consumption is taken into consideration, which is assumed to be proportional to the number of active fronthaul links quantified by using indicator functions. Both the rate requirements and the power constraints of the remote access units are considered. Under realistic power constraints some ofthe users cannot be admitted. Hence, we formulate a two-stage optimization problem. In Stage I, a novel user selection method is proposed for determining the maximum number of admitted users. In Stage II, we deal with the energy efficiency optimization problem. First, the indicator function is approximated by a smooth concave logarithmic function. Then, a triple-layer iterative algorithm is proposed for solving the approximated energy efficiency optimization problem, which is proved to converge to the Karush-Kuhn-Tucker conditions of the smoothened energy efficiency optimization problem. To further reduce the complexity, a single-layer iterative algorithm is conceived, which guarantees convergence. Our simulation results show that the proposed user selection algorithm approaches the performance of the exhaustive search method. Finally, the proposed algorithms is capable of achieving an order of magnitude higher energy efficiency than its conventional counterpart operating without considering link selection.
Thirdly, a QoS driven power-and rate-adaptation scheme aiming at maximizing the effective
capacity of the user subject to both per-RAU average-and peak-power constraints is proposed,
where the EC is defined as the tele-traffic maximum arrival rate that can be supported by DAS
under the statistical delay-QoS requirement. We first transform the EC maximization problem
into an equivalent convex optimization problem. By using the Lagrange dual decomposition
method and satisfying the KKT conditions, the optimal transmission power of each RAU can be
obtained in closed form. Furthermore, an online tracking method is provided for approximating
the average power of each RAU for the sake of updating the Lagrange dual variables. For the
special case of two RAUs, the expression of the average power to be assigned to each RAU can
be calculated in explicit form, which can be numerically evaluated. Hence, the Lagrange dual
variables can be computed in advance in this special case. Our simulation results show that the proposed scheme converges rapidly for all the scenarios considered and achieves 20% higher
EC than the optimization method, where each RAU power is independently optimized.
Fourthly, we consider the energy efficiency maximization problem for a single-user distributed
antenna system over the Nakagami-m fading channels, where EE is defined as the ratio
of the average SE to the average total power consumption. In addition to the conventional peak
power constraints for each RAU, the average power constraint for each RAU is also taken into
account in the optimization problem due to the limited power budget. We first adopt the fractional programming method to transform the original fractional objective function into a more tractable subtractive form. Then, the dual variables associated with average power constraints are introduced to decompose several independent subproblems, the solution of which can be obtained in closed form by analyzing the KKT conditions. The subgradient method is used to update the dual variables, and the online algorithm is adopted to track the average power in order to calculate the subgradient. For the special case of two RAUs, the closed-form expression of the average power is derived, which facilitates direct applications without the need for training. Simulation results demonstrate the fast convergence of our proposed algorithm and the performance gain of nearly 40% over the the existing algorithms in terms of the EE performance.
Fifthly, we solve the energy efficiency maximization problem for multicast services in a
MISO distributed antenna system. A novel iterative algorithm is proposed, which consists of solving two subproblems iteratively: the power allocation problem and the beam direction updating
problem. The former subproblem can be equivalently transformed into a one-dimension
quasi-concave problem that is solved by the golden search method. The latter problem can be
efficiently solved by the existing method. Simulation results show that the proposed algorithm
achieves significant EE performance gains over the existing rate maximization method. In addition, when the backhaul power consumption is low, the EE performance of the DAS is better
than that of the centralized antenna system.