Accelerating time series subsequence matching on the Intel Xeon Phi many-core coprocessor

The problem of time series subsequence matching occurs in a wide spectrum of subject areas. Currently Dynamic Time Warping (DTW) is the best similarity measure but despite various existing speedup techniques it is still computationally expensive. Due to this reason science community is trying to accelerate DTW calculation by means of parallel hardware. There are implementations of DTW-based subsequence matching on GPU and FPGA but there none for accelerators based on the Intel Many Integrated Core architecture. This paper presents a parallel algorithm for time series subsequence matching based on DTW distance adapted to the Intel Xeon Phi coprocessor. The experimental results on synthetic and real data sets are presented and confirm the efficiency of the algorithm.