Adopting mode division multiplex (MDM) technology as the next frontier for optical fiber communication and on-chip optical interconnection systems is becoming very promising because of those remarkable experimental results based on MDM technology to enhance capacity of optical transmission and, hence, making MDM technology an attractive research field. Consequently, in recent years the large number of new optical devices used to control modes, for example, mode converters, mode filters, mode (de)multiplexers, and modeselective switches, have been developed for MDM applications. This paper presents a review on the recent advances on mode converters, a key component usually used to convert a fundamental mode into a selected high-order mode, and vice versa, at the transmitting and receiving ends in the MDM transmission system. This review focuses on the mode converters based on planar lightwave circuit (PLC) technology and various PLC-based mode converters applied to the above two systems and realized with different materials, structures, and technologies. The basic principles and performances of these mode converters are summarized.

In this paper, the resistive switching behaviors of Ru:SiO2/TiO2 based memristive devices have been investigated. It is found that the random and uncontrolled formation of conductive filaments in the Ru/Ru:SiO2/p++-Si devices are crucial to realize a filamentary resistive switching. It is also found that the resistive switching behavior of Ru/Ru:SiO2/p++-Si devices could be significantly improved via inserting a TiO2 interfacial layer as in the form of Ru/Ru:SiO2/TiO2/p++-Si device structure. In the modified device, strong and stable conductive filament formation could be realized when the top electrode is positively biased. In addition to nonvolatile memory applications, an analog-type switching behavior has also been realized in our newly proposed resistive switching device. The current obtained analog conductance modulation is essential for simulating synaptic functions in electronic devices for neuromorphic applications.