Search

Adaptive digital beamformer (ADBF) is effective in suppressing unexpected interferences arriving at the antenna array. An ADBF intellectual property (IP) using the floating point data type has been developed with high-level synthesis (HLS) technology. This IP computes the real-time adaptive weights of each sub-array to spatially filter the arriving signals. The design has been validated in a recent airborne flight test and the result demonstrated the effectiveness of this approach.

A hybrid machine learning (HML) model combining a-priori and a-posteriori knowledge is implemented and tested, which is shown to reduce the prediction error and training complexity, compared to an analytical or neural network learning model.

We propose and experimentally demonstrate an all-optical architecture for data center interconnect networks with reconfiguration times of a few seconds. Filtering and amplification transient effects have minimal impact on BER performance.

Wavelength dependent gain is a critical parameter in erbium-doped fiber amplifiers and is the primary determinant of the channel power divergence and excursions in optical transmission systems, each of which varies with channel loading in wavelength-division-multiplexed (WDM) systems. In an optical transmission system, measurements of the wavelength dependent gain between two locations can be used to obtain estimates of optical power excursions that occur in optical circuit switching. Non-intrusive sampling of just 10% of the WDM channel positions using weak probe signals achieves less than 0.15 dB error for the gain spectrum estimation of the amplifier with 5 dB tilt. In a two-span node-to-node network with a 3 dB peak-to-peak channel power difference, the weak probe approach further shows to provide less than 0.2 dB error in optical channel power excursion prediction for changes in channel loading.

Optical transmission systems exist ubiquitously in data center, access, metro and long-haul networks, and are divided into different autonomous domains based on different carrier ownership, network equipment manufacturers, geographical location, etc. Software-defined Internet exchange methods have been proposed to solve the long-standing problem of inter-domain routing and enable control flexibility in layer-3 networks, and software-defined networking (SDN) has also been introduced to optical networks and much progress has been made to enhance optical network control capability. In this paper, we investigate a transparent software-defined exchange (tSDX) with real-time impairment-aware service level agreement (SLA) guarantees for multidomain optical networks. An SDN-based hierarchical control architecture is designed and implemented. End-to-end, impairment-aware wavelength path provisioning is experimentally realized in a multidomain network. Real-time optical signal to noise ratio based optical performance monitoring, together with inter-domain negotiation enabled by our proposed SDN control architecture, serves as an SLA guarantee for 100-Gbps polarization-multiplexed quadrature phase-shift keying signals in dense wavelength division multiplexed transmission.

The COSMOS testbed provides an open-access and programmable multi-layer beyond 5G wireless platform built on an advanced optical x-haul network supporting mobile edge cloud base band processing and applications.

In this paper, we study a cognitive unmanned aerial vehicle (UAV) relaying system (CURS) which shares the downlink frequency of a satellite network. Specifically, we propose a collaborative beamforming (CBF) scheme that maximizes the signal-to-interference plus noise ratio (SINR) of CURS, subjects to the SINR requirements of the satellite network and the individual transmit power constraints of UAV relays. Since the original optimization problem is non-convex and mathematically intractable, we first transform it into an approximate convex one by means of the Charnes-Cooper transformation and the penalty function method as well as the difference of two-convex functions (D.C.) approach. Then a convergence proven iterative algorithm is designed to find the optimal CBF solution. Simulation results are provided to demonstrate the effectiveness of the proposed scheme and the impact of UAV transmit power and satellite network SINR requirements on the CURS.

This paper investigates the outage performance of an unmanned aerial vehicle (UAV) relaying system with decode-and-forward (DF) protocol, where multiple sources access the destination via UAV. By assuming that the destination is corrupted by multiple co-channel interferers, we first obtain the end-to-end signal-to-noise ratio (SNR) expression. Then, based on the SNR expression, we derive the closed-form expression of the outage probability (OP) for the UAV relaying system, where both of the uplink and downlink channels undergo Nakagami-m fading while the interference channels experience Rayleigh fading. Finally, numerical results are given to demonstrate the effectiveness of the performance analysis, and reveal the impacts of UAV antenna numbers and co-channel interferer on the system performance.

In this paper, we propose a cooperative beamforming (BF) scheme for unmanned aerial vehicle (UAV)-assisted relay networks, where multiple UAVs are deployed as cognitive relays to assist the signal transmission between a cognitive transmitter and a cognitive receiver. Specifically, assuming that only the partial channel state information of each link is available, we first formulate an optimization problem to minimize the total transmit power of UAV relays under the constraints of the quality-of-service (QoS) requirement of the cognitive network and the interference limitation of the primary network. Then, a vector iterative algorithm is presented to solve the optimization problem so that the optimal BF weight vector is obtained. Finally, simulation results are provided to demonstrate the effectiveness of the proposed BF algorithm and show the effects of UAV number and power constraint on the considered UAV-assisted network.

Optical transmission systems with high spectral efficiency require accurate quality of transmission estimation for optical channel provisioning. However, the wavelength-dependent gain effects of erbium-doped fiber amplifiers (EDFAs) complicate precise optical channel power prediction and low-margin operation. In this work, we examine supervised machine learning methods using multiple artificial neural networks (ANNs) to build models for gain spectra prediction of optical transmission line EDFAs under different operating conditions. Channel-loading configurations and channel input power spectra are used as an a posteriori knowledge data feature for model training. In a hybrid learning approach, estimated gain spectra calculated by an analytical model are added as an a priori input data feature to further improve the EDFA ANN model performance in terms of prediction accuracy, training time, and quantity of training data. Using these methods, the root mean square error and maximum absolute error of the predicted channel output power can be as low as 0.144 dB and 1.6 dB, respectively.

Disaggregating optical communication systems can impact physical layer control. Recent progress on multi-domain transmission control and machine-learning provide capabilities for adaptation and development of engineering rules in the field with potential benefits for disaggregated systems.

Machine learning based modelling of Erbium-Doped Fiber Amplifiers (EDFA) is used to determine wavelength dependent gain for use in optical transmission systems, and achieves root mean square error (RMSE) of 0.08, 0.18, and 0.27 dB under input ranges of +/− 3, 6, 9 dB.

We experimentally demonstrate self-adaptive coded 5×100 Gb/s WDM polarization multiplexed 16 quadrature amplitude modulation transmission over a 100 km fiber link, which is enabled by a real-time control plane. The real-time optical signal-to-noise ratio (OSNR) is measured using an optical performance monitoring device. The OSNR measurement is processed and fed back using control plane logic and messaging to the transmitter side for code adaptation, where the binary data are adaptively encoded with three types of low-density parity-check (LDPC) codes with code rates of 0.8, 0.75, and 0.7 of large girth. The total code-adaptation latency is measured to be 2273 ms. Compared with transmission without adaptation, average net capacity improvements of 102%, 36%, and 7.5% are obtained, respectively, by adaptive LDPC coding.

Greater network flexibility through software defined networking and the growth of high bandwidth services are motivating faster service provisioning and capacity management in the optical layer. These functionalities require increased capacity along with rapid reconfiguration of network resources. Recent advances in optical hardware can enable a dramatic reduction in wavelength provisioning times in optical circuit switched networks. To support such operations, it is imperative to reconfigure the network without causing a drop in service quality to existing users. Therefore, we present a system that uses neural networks to predict the dynamic response of an optically circuit-switched 90-channel multi-hop Reconfigurable Optical Add-Drop Multiplexer (ROADM) network. The neural network is able to recommend wavelength assignments that contain the power excursion to less than 0.5 dB with a precision of over 99%.

End-to-end, impairment-aware wavelength path provisioning is experimentally realized in a multi-domain network using a transparent software-defined exchange (tSDX) with real-time OSNR monitoring as an SLA guarantee for 100Gbps PM-QPSK DWDM transmission.

Recent advances in software and hardware greatly improve the multi-layer control and management of reconfigurable optical add-drop multiplexer (ROADM) systems facilitating wavelength switching. However, ensuring stable performance and reliable quality of transmission (QoT) remain difficult problems for dynamic operation. Optical power dynamics that arise from a variety of physical effects in the amplifiers and transmission fiber complicate the control and performance predictions in these systems.We present a deep-neural-network-based machine learning method to predict the power dynamics of a 90-channel ROADM system from data collection and training. We further show that the trained deep neural network can recommend wavelength assignments for wavelength switching with minimal power excursions.

Using a dual-wavelength source, a single optical signal is distributed over two wavelengths. This approach is used to reduce power excursions due to optical circuit switching in ROADM systems. In a multi-hop optical transmission system with 100Gbps PM-QPSK signals switched over five ROADMs and 265 km of single mode fiber (SMF), power excursions are kept within ± 0.2 dB using dual-wavelength sources. A cumulative distribution function (CDF) of channel power excursions is generated from measurements of over 100 random channel loadings for four different channel loading plans. Using dual-wavelength sources, power excursions of up to ± 6 dB are reduced below ± 1.5 dB for 99% and ± 0.5 dB for 71.4% of the cases while switching up to 30 channels.

Fast switching of a dual laser PM-QPSK transceiver is used to mitigate channel-power excursions in amplified optical transmission under wavelength reconfiguration.

An SDN controller is developed for both testbed management and experimentation for the optical x-haul network in the COSMOS testbed providing a service-on-demand and reconfigurable platform for 5G wireless experiments coupled with edge cloud services.

Development of optical x-haul networks is driven by 5G wireless radio requirements. The potential of a mesh optical x-haul architecture merging WDM-PON and DWDM-ROADM networks is examined with respect to 5G requirements in metropolitan networks.

Hysteresis-based margin-allocation for adaptive coding is experimentally investigated against PDL-induced OSNR fluctuations in a SDN-enabled multi-domain optical network. Up to 90% network outage reduction can be observed under different margins.