In this paper we analyze the characteristics of tapered multimode interference (MMI)-based coherent lightwave combiners, and theoretically confirm that the stable and clear multimode interference images exist in the tapered MMI combiners. We present the output characteristics of 21 tapered MMI-based coherent lightwave combiners under various structures, which are useful to optimally design the combiners. In order to extend the combiner to a multi-port (N1, N > 2) configuration, a new structure with the border shapes of two tangent arcs is proposed, by which we show an output power imbalance of about 0.5 dB between different input ports of an 81 tapered coherent lightwave combiner. Due to the advantages of no end-facet reflection, easy extension to a multi-port configuration, high tolerance for fabrication errors and a compact size, the tapered MMI structure is a good candidate as a coherent lightwave combiner used in large-scale photonic integrated circuits.

Traffic classification has recently gained much attention in both academic and industrial research communities. Many machine learning methods have been proposed to tackle this problem and have shown good results. However, when applied to traffic with out-of-sequence packets, the accuracy of existing machine learning approaches decreases dramatically. We observe the main reason is that the out-of-sequence packets change the spatial representation of feature vectors, which means the property of linear mapping relation among features used in machine learning approaches cannot hold any more. To address this problem, this paper proposes an Improved Dynamic Time Warping (IDTW) method, which can align two feature vectors using non-linear alignment. Experimental results on two real traces show that IDTW achieves better classification accuracy in out-of-sequence traffic classification, in comparison to existing machine learning approaches.