为了最大化OFDM系统用户上行链路的有效吞吐量,提出了一种新的用户上行资源分配策略.与传统策略相比,该策略通过引入自适应的子载波分段数和数据包OFDM符号数,增加了设计自由度.为降低计算复杂度,有效吞吐最最大化问题被分解为2个较简单的子问题,即用户有效资源分配和数据包参数设置.针对这2个子问题,分别通过分析目标函数的单调性特征和放宽整数约束条件,各提出了一种低复杂度的算法,即L和M搜索算法,并从理论上证明了L算法的最优性.进一步结合以上2种算法,提出了一种用户上行有效吞吐量最大化算法(UUGM).仿真结果表明:由于设计自由度的增加,UUGM算法与已有的算法相比,提高了有效吞吐量,尤其在瑞利信道下,可提高5％～186％,接近于最优的枚举法；而UUGM算法的复杂度,虽然略高于传统算法,但远低于枚举法.%A new user-uplink resource allocation strategy to maximize the user-uplink goodput in an OFDM system was proposed. In comparison with the traditional strategy, by introducing self-adaptive sub-carrier number and OFDM symbol number per data packet, the strategy increases the design freedom. In order to lower the calculation complexity, the goodput maximization problem is decomposed into two simple problems; allocation on user's effective resource and the setting on data packet parameter. A low complexity search algorithm called L was implemented for the user effective resource allocation problem based on the monotonicity analysis of the objective function. A search algorithm called M was presented to resolve the data packet parameter setting problem, relaxing the integral constraints. Theoretically, the optimization of L algorithm was certified. Furthermore, a user-uplink goodput maximization (UUGM) algorithm was produced, combining the two proposed algorithms. Theoretical analysis and simulation results illustrate that, due to the increase of design freedom, comparing with the existing algorithm, the goodput was effectively increased. Under Rayleigh channel, the goodput may be increased by 5% ～186% , the performance of the proposed UUGM algorithm is almost the same as the enumeration method. However, though the complexity of UUGM algorithm is a little higher than the traditional algorithm, it is far lower than the enumeration method.
展开▼
