Measurement-based optimizing algorithm for modulation and coding scheme selection in downlink LTE self-organizing networks

The traditional wireless network management becomes no longer efficient because of the new features and technologies in the next generation radio access network. The main purpose of self-organizing network (SON) is to meet new technical requirements, simplify wireless network design and operation, and implement the self-configuration, self-optimization and self-healing of the network. Adaptive modulation and coding (AMC) is one of the key technologies of link adaptation in TD-LTE system to achieve high spectral efficiency on the fading channels [1]. A key factor to determine the performance of AMC is the threshold values of MCS applied at eNode B, which decides the suitable MCS selection in the next transmission time interval (TTI). Unsuitable MCS selection will impair the system performance seriously. Relevant research on the optimizing algorithms for MCS selection includes [2]–[3]. In [2], an improved threshold adjustment to precisely guarantee the predetermined frame error rate for AMC was proposed. A novel signal-to-noise ratio (SNR) estimation algorithm which reduces computational complexity in obtaining channel quality indictor (CQI) and corresponding MCS was investigated in [3]. None of these took fully into account of the feedback and periodic performance measurement simultaneously, and the interference was not considered. Meanwhile, eNode B can't obtain the value of signal to interference plus noise ratio (SINR) directly. In this paper, a measurement-based optimizing algorithm (MBOA) for modulation and coding schemes (MCS) selection to determine an appropriate MCS for initial transmission is proposed. Differing from the traditional algorithm which selects the MCS only by channel quality indicator (CQI) feedback, the proposed algorithm can adjust the threshold values of MCS according to the performance measurements and CQI feedback, including fast and independent adjustment (FIA) and slow and centralized adjustment (SCA). FIA optimizes the threshold values by the initial transmission success ratio (ITSR). Then SCA changes threshold values by the measurement of average cell throughput and the block error rate (BLER). It is shown from the system level simulation results that MBOA yields higher average cell throughput and transmission success ratio compared with the traditional algorithm.