Energy-Efficient Power Adaptation over a Frequency-Selective Fading Channel with Delay and Power Constraints
Energy-Efficient Power Adaptation over a Frequency-Selective Fading Channel with Delay and Power Constraints
复制标题
DOI：
10.1109/twc.2013.080113.120767
复制
发表时间：
2013-08
影响因子：
10.4
通讯作者：
A. Helmy;Leila Musavian;T. Le-Ngoc
中科院分区：
文献类型：
--
作者：
A. Helmy;Leila Musavian;T. Le-Ngoc
提出了一种基于时延中断概率约束的多载波链路在频率选择性衰落信道下的能量高效功率分配方法。功率自适应最大化系统能量效率(EE)，其表示为实现的有效容量(EC)与总消耗功率的比率，包括传输功率和与速率无关的电路功率。我们证明了该目标函数在传输功率上是拟凹的，并用分式规划方法求出了全局最优解。基于所得到的解，我们提出了一种功率自适应算法，该算法包括两个步骤：(I)在没有发射功率约束的情况下，建立与最大可达到的EE相对应的最优平均功率水平；(Ii)对于给定的功率约束，基于该约束和在第一步中找到的最优功率水平，联合分配功率在时间和频率上的分布。分析结果表明，所提出的基于EE的功率分配具有与最大化EC的分配相似的结构，但具有不同的截止门限。与基于频谱效率的联合功率分配方案和基于独立子载波的联合功率分配方案相比，我们提出的联合EE-最优功率分配方案提供了显著的EE增益，其中，随着频率选择性的提高，速率-能量权衡变得更加明显。
This paper presents an energy-efficient power allocation for a multicarrier link over a frequency-selective fading channel with a delay-outage probability constraint. The power adaptation maximizes the system energy efficiency (EE), formulated as the ratio of the achieved effective capacity (EC) to the total expenditure power, including both transmission power and rate-independent circuit power. We prove that this objective function is quasi-concave in the transmission power, and derive the global optimum solution using fractional programming. Based on the obtained solution, we develop a power adaptation algorithm consisting of two steps: (i) establishing the optimum average power level corresponding to the maximum achievable EE with no transmit power constraint, and then (ii) for a given power constraint, jointly distributing the power over time and frequency based on the constraint and the optimum power level found in the first step. Analytical results show that the proposed EE-based power allocation has a structure similar to the allocation that maximizes the EC, but with a different cut-off threshold. Our proposed joint EE-optimal power allocation provides significant EE gains over both the joint spectral-efficient and independent-subcarrier EE-based power allocation schemes, where the rate-energy tradeoff becomes more pronounced with higher frequency selectivity.