基于监测数据的非参数频谱地图构造方法

Non-Parametric Spectrum Map Construction with Spectrum Monitoring Data

本文研究了频谱地图构造问题,提出了一种基于监测数据的非参数构造方法,无需用到关于发射机和传播环境相关信息。本文将感兴趣区域内任意位置的接收功率建模成幂函数的线性组合,并提出使用交替最小化方法,利用监测数据对线性组合权重、幂函数中心点位置和指数进行联合优化。仿真部分通过与常用的反距离加权法和克里金插值法进行对比,验证了本文方法的有效性。

This paper studies the problem of spectrum map construction and presents a novel non-parametric algorithm based on monitoring data which does not depends on any parameters of transmitters and propagation environment. In this paper, a linear combination of power functions centered on a specific number of centroids is proposed to model the received signal power at arbitrary location over a geographical area of interest, and then the weights, centroid locations and exponent are jointly optimized using an iterative alternating minimization method. The superior performance of the proposed algorithm is verified via simulations by comparing with inverse distance weighted method and Kriging method.