CAJ | 학술논문

最优多用户检测具有指数级的计算复杂度,不能实际应用.为了缩短多用户检测时间,达到实时性的要求,提出一种将最优多用户检测的解转化成特征向量的方法,直接利用幂法计算特征向量.首先利用拉格朗日乘数法将具有约束条件的最优多用户检测优化模型转化为无约束的优化模型,然后将无约束的优化模型转化为一个代数方程组,最后对该代数方程进行同解变形,从而将最优多用户检测的解转化为特征向量.仿真结果表明,该方法能大大缩短检测时间,而在误码率、抗远近效应和增加系统容量方面的性能较之传统方法和粒子群算法也有一定的提高.
최우다용호검측구유지수급적계산복잡도,불능실제응용.위료축단다용호검측시간,체도실시성적요구,제출일충장최우다용호검측적해전화성특정향량적방법,직접이용멱법계산특정향량.수선이용랍격랑일승수법장구유약속조건적최우다용호검측우화모형전화위무약속적우화모형,연후장무약속적우화모형전화위일개대수방정조,최후대해대수방정진행동해변형,종이장최우다용호검측적해전화위특정향량.방진결과표명,해방법능대대축단검측시간,이재오마솔、항원근효응화증가계통용량방면적성능교지전통방법화입자군산법야유일정적제고.
The optimal multiuser detection has the exponential complexity, so it cannot be used in practical applications. The authors presented a method to shorten the time of multiuser detection and achieve the real-time requirement. First, the Lagrange multipliers was used to transform the constrained multiuser detection optimization model into a non-constrained one. Second, the non-constrained optimization model was transformed into an algebraic equation. Finally, the multiuser detection solution was transformed to the eigenvectors of a matrix by the deformation of the algebraic equation and the power method was used to calculate the eigenvectors. Simulation results show that the algorithm can significantly shorten the time of detection, reduce the Bit Error Rate ( BER), resist near-far effects and increase the user capacity of the system compared with traditional algorithm and particle swarm optimization.