We report, for the first time to our knowledge, the experimental observation of quasi-cw nonlinear switching and multiple gap-soliton formation within the bandgap of a fiber Bragg grating. As many as five gap solitons with 100-500-ps durations were generated from a 2-ns pulse at a launched peak intensity of approximately 27 GW/cm(2). A corresponding increase in the grating transmission from 3% to 40% of the incident pulse energy was observed.

changed, i.e. it does not evolve into SS pulse even if the gain coefficient is doubled.Hereto we know 10-ps pulse can not evolve into similariton by increasing propagating distance, pulse energy, and gain coefficient, the reason lies in L D is greatly larger than fiber length z. To make the pulse evolve into similariton, we choose higher dispersion coefficient, ␤ 2 ϭ 25ps 2 /m, to get shorter L D ϭ 4 m. Given U in ϭ 10 pJ and the other amplifier parameter is not changed, the calculated empirical distance for self-similar evolution is about 6 m. Figure 4(a) gives the evolution result of 10-ps pulse at z ϭ 7 m when L D is 4 m; Figure 4(b) gives the evolution result of 10-ps pulse at the distance of 3.5 m when L D is 4 m and g is 3.8 m Ϫ1 . It can be seen when L D (4 m) is comparable with the fiber length, and 10-ps pulse can evolve into similariton within the self similar evolution distance, this result further accounts for that L D determines the SS evolution. So it is very important to ensure L D to be smaller than or comparable with the fiber length, for different initial pulse we can choose different fiber to meet this condition. Comparing Figure 4(a) with Figure 4(b), it is clear that when the gain coefficient increases, the SS evolution is accelerated. CONCLUSIONIn this article, we have numerically studied the influences of amplifier parameters on the pulse's the self-similar evolution. We find that L D is the determining factor for pulse self-similar evolution, while L NL is not. Pulse with L D near fiber length can evolve into similariton, otherwise it can not. The inner mechanism for self-similar evolution is that pulse should be under the effect of GVD, nonlinear, and gain. When L D is comparable with fiber length, the pulse can experience enough GVD effect. If L D is near fiber length, then increasing pulse energy and gain coefficient can accelerate its evolution, otherwise it can not. We can choose appropriate amplifier parameters to make L D comparable with fiber length, thus make pulse self-similar evolve. The study can help us to get similariton in fiber amplifier, and this is very useful for the SS parabolic pulse amplifier design. ACKNOWLEDGMENTThis research is supported by the National Natural Science Foundation of China (Grant No. 60677013). OPTICAL MM-WAVE GENERATION BASED ON PHASE MODULATION ALONG WITH OPTICAL FILTERING

We report the amplification of 10pJ-100pJ, semiconductor diode pulses up to an energy of 158µJ and peak powers >100kW in a multi-stage fibre amplifier chain based on a single-mode, large mode-area erbium doped amplifier design. These results represent the highest single-mode pulse energy ever extracted from any doped fibre system. Sources capable of delivering high energy pulses of nanosecond order duration at 'eye safe' wavelengths beyond 1500nm are required for a wide range of practical applications from LIDAR through to laser marking. A semiconductor seed source followed by an erbium doped fibre amplifier (EDFA) cascade would seem an attractive means to such an end, providing for a compact, reliable system. Recently progress towards this solution to such applications has been made with >100µJ, 10ns pulses being reported from a multimode, diode-pumped Er/Yb fibre amplifier [1]. However, for many of these applications a single transverse mode-output beam is a desirable, if not fundamental requirement. To date conventional single-mode EDFA fibre designs, optimised for small signal gain and high pump efficiency, have restricted the attainable pulse energies to less than 10µJ [2,3]. For applications requiring higher pulse energies source choice has therefore been limited to more cumbersome and expensive bulk laser options.To obtain significantly increased output energy characteristics one needs to consider what can be done in terms of fibre design to achieve improved energy storage within the EDFA. Increasing the mode-field and dopant areas of the EDFA should result in a greater number of ions participating in the amplification process and a reduction of the small signal gain occurring within the amplifier, which can restrict the inverted population depletion through gain saturation by amplified spontaneous emission (ASE). The improvement such increases make has been shown through the numerical modelling of low repetition-rate high-energy EDFA pulse amplification in reference [4] and in the work of reference [5] where a multimode EDFA produced pulse energies of 0.4mJ. With the restriction of maintaining a single transverse mode at the signal wavelength, greater single-mode areas can be achieved through a reduction in the fibre numerical aperture along with an increased core size, though such changes are limited by the requirement of maintaining a practical level of bend-loss. We have recently reported the development and application of a large mode area EDFA in an experimental chirped-pulse amplification (CPA) system [6] where the increased mode area also reduces the nonlinear effects that limit pulse recompression in these systems. In this paper we report an examination of the limits to pulsed energy extraction in these large mode-area fibres through the amplification of 10ns-100ns pulses, resulting in the highest single-mode pulse energies extracted from any doped fibre amplifier system, to our knowledge. igure 1 shows a schematic diagram of the experimental setup. A laser diode source, providing 1mW cw ...

We experimentally demonstrate an all-optical AND gate based on coupled gap-soliton formation in an apodized fiber Bragg grating. A switching contrast of better than 17 dB is obtained with an incident pulse peak power of 2.5 kW.

We demonstrate a scheme for optical pulse compression by cross-phase modulation that utilizes a nonuniform Bragg grating to work in reflection. Our scheme is similar to the conventional optical pushbroom, which works in transmission. This reflection geometry has the advantage of allowing the compressed signal to be observed easily, as it is spatially separate from the pump. This is to our knowledge the first nonlinear effect to be observed that requires a nonuniform grating.