No abstract

Starting from the estimation criterion and the essential problem of algorithmic computing, this paper introduces four kinds of spectral estimation methods commonly used in teaching: Periodogram to fit frequency vector under least square criterion，AR model estimation converts spectrum estimation into parameter estimation, MVDR converts spectrum estimation into filter weight coefficient estimation, and MUSIC converts spectral estimation into subspace estimation.Then we analyze and solve the equation, derive the assumption of signal and the defects caused by this assumption.

No abstract

Abstract. In this paper, we propose a subspace-based matching (SBM) approach based on fractional lower order moment (FLOM) matrix to estimate the location of near-field acoustic signals in additive nonGaussian impulsive noise environment. By matching the subspace in the fingerprint database and the subspace computed from the current received signals, we associate the received signal with its transmission location with high accuracy. The proposed algorithm has two advantages: firstly, it can locate multiple sources without estimating bearing and range of sources. Secondly, it is more robust to multipath, environment changing and noise level. The simulation results show the efficacy of our proposed algorithm.

Abstract. In this paper, we propose a new algorithm to detect outliers for RSS measurements in building environment. RSS measurements fluctuate severely in complex building environment because multi-path, obstacles, and moving people, can distort the propagation of radio wave, which leads to some inevitable large values in RSS measurement, known as outliers. Some classical outliers detection algorithm fail to detect them when outliers rate is higher. Based on this observation, we proposed a modified Hample filter algorithm to detect outliers. Simulation results show that our proposed algorithm can overcome the fluctuation of RSS in building localization environment.