Research On High-Speed Optical Vector Signal Processing Technology For Flexible Optical Networks

In recent years,emerging IP services such as the Internet of Things(IoT),high-definition(HD)video,and real-time conferencing,driven by the construction of the "Arithmetic Network",the development strategy of the Industrial Internet,and 5G technology,have greatly contributed to the development of the capacity and architecture of optical networks.Compared to the total number of subscribers and the surge in network traffic,the annual growth rate of fiber optic communication capacity is no longer sufficient to meet the current growth rate of business volume.In super 100G communication systems,a combination of higher-order modulation formats and multiplexing methods is used to improve spectrum efficiency and expand system capacity.In addition,changes in service characteristics have put forward new requirements for the flexibility of optical networks,prompting the evolution of optical networks towards multi-functional,multi-layered and flexible optical networks,which undoubtedly brings new challenges to the node’s signal processing technology.The current digital chip-based signal processing technology has rate bottlenecks,low optical-electrical conversion efficiency and other problems,it is difficult to meet the development needs of flexible optical networks.Optical vector signal processing technology can directly in the optical domain to provide a variety of flexible signal processing,effectively avoiding inefficient photoelectricity optical conversion at the same time,but also has a fast response and high-speed processing capabilities,can maximize the advantages of all-optical networks,in the future flexible optical network will play an important role.This paper is oriented to the development needs of flexible optical network,for the lack of flexibility in the current optical network,focusing on the highspeed optical vector signal processing technology to carry out research,with the goal of developping flexible,efficient and scalable all-optical processing methods in a variety of scenarios as the goal,proposes the all-optical aggregationde-aggregation method for flexible modulation conversion,the all-optical constellation add/drop multiplexing method for flexible optical switching,and the all-optical processing method for multi-wave signals,the main research work and innovation points are as follows:(1)For flexible all-optical modulation conversion,an all-optical aggregation and de-aggregation method from QPSK signal to two-channel BPSK signal is proposed to solve the problem that existing aggregation needs to introduce intensity signals and de-aggregation requires multiple nonlinear processing.This paper uses a delayed interference mechanism to avoid introducing additional intensity signals during the aggregation process,and then uses XPM to achieve all-optical aggregation of two BPSK signals to one QPSK signal.In addition,all-optical de-aggregation of QPSK signals to two BPSK signals is achieved in a single nonlinear process through a loop de-aggregator.The experimental results show that when the BER reacheed the FEC threshold,the required OSNR of the aggregated QPSK signal is 13.2 dB,which is a 2.62 dB degradation compared to the OSNR of the B2B conditionthe,and the deaggregated BPSK signals are degraded by 1.32 dB and 1.56 dB compared with the reference BPSK signals.(2)For flexible optical switching,an all-optical constellation add-drop multiplexing method with flexible granularity is proposed to address the problems of large switching granularity and lack of flexibility in existing optical switching.In order to realize optical switching with smaller granularity,this paper proposes two all-optical constellation add-drop multiplexing methods based on different modulation formats from the constellation dimension:a)An all-optical constellation add-drop multiplexing method for MPSK signals is proposed.An all-optical constellation add-and-drop multiplexing method for MPSK signals is proposed.Taking QPSK(M=2)and 8PSK(M=4)signals as examples,an all-optical loop de-multiplexer,an XPM-based constellation multiplexing and constellation drop based on all-optical phase erasure are used to realize all-optical constellation add-drop multiplexing with BPSK as the minimum switching granularity,which breaks the optical switching paradigm where wavelength is the minimum granularity.Simulation results show that the OSNR of the multiplexed QPSK and 8PSK signals reaches 13 dB and 19.6 dB at the FEC threshold;b)A constellation add-drop multiplexing multiplexing method for 16-QAM signals is proposed,which utilizes PSA-based constellation demultiplexing,saturation amplitude compression-based constellation drop,and FWM-based constellation multiplexing to realize all-optical constellation adddrop multiplexing with QPSK as the minimum switching granularity.Simulation results show that the OSNR required for multiplexed 16-QAM signal reaches 24.6 dB at the FEC threshold.(3)For multi-channel processing,an all-optical de-aggregation method for multi-wave signals is proposed in response to the problems of obvious interference and high complexity of existing all-optical processing schemes for multi-channel signal processing.In order to be able to effectively process multi-wave signals,this paper proposes a high-speed all-optical de-aggregator for multi-wavelength QPSK/16-QAM signals by combining the mechanism of polarization-assisted,which is characterized by the introduction of spatial separation between signals,effectively avoiding the adverse effects of mixing interference during signal processing.Simulation results show that the proposed method is capable of realizing the deaggregation of 4-channel 25/50 Gbaud QPSK/16-QAM signals,and the OSNR difference between the 4-channel deaggregated signals is no more than 1.28 dB and 1.48 dB under the FEC threshold.