Joint User Association and Power Allocation for Hybrid Half-Duplex/Full-Duplex Relaying in Cellular Networks
Joint User Association and Power Allocation for Hybrid Half-Duplex/Full-Duplex Relaying in Cellular Networks
复制标题
DOI：
10.1109/jsyst.2018.2850861
复制
发表时间：
2019-06
影响因子：
4.4
通讯作者：
Gang Li;Hongbin Chen;Jun Cai
中科院分区：
文献类型：
--
作者：
Gang Li;Hongbin Chen;Jun Cai
针对中继辅助蜂窝网络中的混合半双工/全双工中继，提出了一种联合用户关联、中继模式选择和功率分配的方案，在满足最小频谱效率要求和发射功率约束的前提下，最大化能量效率.该组合优化问题是一个混合整数非凸组合优化问题，其原始形式难以求解。为了避免这个问题，首先将问题转化为一个等价的目标函数是在一个参数减法形式。然后，通过松弛指标变量，用逐次凸逼近法转化非凸性，构造辅助变量，将等价问题转化为凸问题。这样，在每次迭代中，等价的非凸问题被近似为一个凸问题，并采用对偶分解技术来求解凸问题。一个迭代算法进一步发展，以达到原始问题的解决方案。仿真结果表明，该迭代算法的性能可以非常接近的最优解的唯一搜索在只有几个迭代次数。此外，所提出的方案优于现有的，只坚持用户关联或中继模式选择，而不是两者兼而有之。
For hybrid half-duplex/full-duplex relaying in a relay-aided cellular network, a joint user association, relay mode selection, and power allocation scheme is proposed to maximize the energy efficiency under the minimum spectral efficiency requirement and transmission power constraints. This combinatorial optimization problem is a mixed-integer nonconvex one, which is difficult to be solved in its original form. To circumvent this issue, the problem is first transformed into an equivalent one whose objective function is in a parametric subtractive form. Then, the equivalent problem is reformulated as a convex one by relaxing index variables, converting the nonconvexity by the successive convex approximation, and constructing auxiliary variables. In this way, the equivalent nonconvex problem is approximated by a convex one in each iteration, and the dual decomposition technique is employed to solve the convex one. An iterative algorithm is further developed to reach the solution to the primal problem. Simulation results show that the performance of the iterative algorithm can be very close to the optimal solution obtained by the exclusive searching within only a few number of iterations. Moreover, the proposed scheme outperforms existing ones, which only adhere to either the user association or the relay mode selection but not both.