Report

Teaser, summary, work performed and final results

Periodic Reporting for period 1 - ORCHESTRA (Optical peRformanCe monitoring enabling dynamic networks using a Holistic cross-layEr, Self-configurable Truly flexible appRoAch)

Teaser

ORCHESTRA is a European research project aiming to develop a hierarchical and integrated performance monitoring and control architecture for new generation optical networks. In particular, ORCHESTRA aims to close the observe-process-act network control loop, making an optical...

Summary

ORCHESTRA is a European research project aiming to develop a hierarchical and integrated performance monitoring and control architecture for new generation optical networks. In particular, ORCHESTRA aims to close the observe-process-act network control loop, making an optical network observable and then subject to optimization. Real monitoring data will be collected from coherent optical interfaces, deployed today in optical networks, which can be extended, almost for free, to also serve as software defined optical performance monitors (soft-OPM). To achieve this, novel digital signal processing (DSP) algorithms for real-time multi-impairment monitoring are being developed and will be combined with a novel hierarchical monitoring plane to handle monitoring information in an efficient and scalable manner. The ORCHESTRA network will be optimally planned to operate close to the best performance that is feasible under the current network conditions, reducing the margins typically reserved when provisioning the lightpaths. By gathering, processing and correlating monitoring information, the ORCHESTRA’s control and management plane will optimize the network by acting in response to certain provisioning tasks, physical layer degradations and/or fault events. This process will be running continuously in the background with the objective of improving the reliability of delivered services, reducing provisioning costs, and simplifying maintenance and operation procedures. The ORCHESTRA vision is to close the control loop (see figure), enabling dynamicity and unprecedented network efficiency.

Fig. 1. ORCHESTRA observe-decide-act dynamic control cycle.

The future of optical networks is coherent and elastic: operators are deploying an all-coherent, multi-format transport layer, leveraging DSP in powerful ASICs. This enables more robust transmission, allowing the shedding of redundant hardware and simplifying network design. ORCHESTRA takes advantage of the evolving trends and aggressively pursues the development of advanced DSP algorithms that add real-time impairment measurement-capability to optical transceivers; potentially, every single transceiver in the network can be used as a soft-OPM. A key point is that monitoring functions come almost for free: the coherent receivers have already deployed ASICs for DSP. In addition to algorithms for measuring and mitigating dispersion effects (present in current receivers), ORCHESTRA works on optical signal to noise ratio (OSNR) and takes on the challenge of estimating non-linear effects and crosstalk.

The ORCHESTRA network will have a plethora of soft-OPMs to extract physical-layer information. But we can do even more: an OPM at a receiver provides aggregate measures over a path that usually spans several links. ORCHESTRA’s ambitious objective is to correlate information from multiple soft-OPMs throughout the network, which opens up a multitude of possibilities such as: quality of transmission (QoT) estimation before lightpath establishment; detection, as well as anticipation, of ‘hard’ (total link failure) and ‘soft’ (QoT degradation) failures. Also, such methods make the gradual deployment of ORCHESTRA more appealing, since added value comes even from just a few OPMs.

ORCHESTRA also develops a hierarchical control and monitoring infrastructure capable of transferring and manipulating monitoring information while it goes beyond passive-monitoring operations by adding active-control functionality. ORCHESTRA’s monitoring plane will enable effective processing of monitoring information (filtering, correlation) and fault management, avoiding bottleneck issues related to centralized approaches. Active control functions include the tuning of transmission parameters of flexible transceivers (changing modulation format, FEC, power, etc), the shift in spectrum or rerouting. Control actions are examined at a local level for single connections, and then at higher hierarchy layer

Work performed

The main achievements during the first period (first 18 months) are summarized below:

WP2: Software Defined Monitoring and Dynamic Network Architecture
The reference network scenarios, which include the network framework, the technology, reference network topologies, the reference control plane architecture, and a reference ageing model were described and five use cases were identified, motivated by telecom operators. A cost model tailored to an important use case related to postponing investments was defined. The state of the art in optical performance monitoring (OPM) was also reviewed. The review included hardware OPMs and the state of the art monitoring and compensation DSP algorithms. This led to a preliminary identification of monitoring parameters to be targeted in ORCHESTRA. Experimental testbeds for design and implementation of (state-of-the-art and novel) monitoring algorithms were setup. Experiments were performed to study how filtering effects from Reconfigurable Optical Add/Drop Multiplexer (ROADM) nodes impact the performance of the transmission. Experimental assessment of literature DSP demodulation and monitoring algorithms was also carried out. In addition, the consortium defined the reference control and management plane for the ORCHESTRA network. The emerging ABNO architecture was chosen as the candidate model for the control and management (monitoring) planes, while the NETCONF protocol– including YANG model– was chose for communications in both planes. A novel hierarchical monitoring architecture was proposed for ORCHESTRA, consisting of virtual monitoring entities and agents with the OAM Handler of the ABNO controller as root entity of the hierarchy. A scalability study was performed to compare the performance of such hierarchical approach to the traditional centralized OAM solution. The main causes of hard and soft failures and implications on the services were identified which were then complemented by actions to recover from the related failures.

WP3: Development of Flexible Optical Transceiver and Software-Defined Optical Performance Monitoring Algorithm Suite
The development of a flexible transmitter to be used in dynamic reconfiguration experiments, along with the associated transmitter side DSP has begun. Two hardware setups based on commercial-off-the-shelf components were conceptually designed and assessed with a set of laboratory experiments. The arbitrary waveguide generator (AWG)-based transmitter architecture was selected and is currently further developed. In parallel, for the receiver side the RxDSP demodulation platform architecture was identified. Existing DSP algorithms were extended with demodulation algorithms addressing Polarization Demultiplexing, Frequency Offset Estimation and Carrier Phase Recovery. The demodulation algorithms were verified through numerical simulations and validated experimentally with data collected during the ORCHESTTRA early field trial. Moreover, after the extensive literature review, the targeted impairments were identified and the development of related DSP monitoring algorithms has begun. Existing monitoring algorithms for linear impairments (e.g. CD, PDL, PMD) were developed and a novel monitoring algorithm for filtering effects is currently developed.

WP4: Cross-layer Algorithms: Multiple-Rx Correlation and Optimization
WP4 is developing correlation algorithms that use information of multiple-Rx (soft-OPM): (i) for estimating the QoT of unmonitored (or new) lightpaths and (ii) identifying the causes of quality of transmission (QoT) problems (soft failures). Two QoT estimation frameworks were developed and an initial study of using correlation for localizing failures was carried out. Development has reached a mature phase and performed studies showcased the benefits that can be obtained. Moreover, based on the ORCHESTRA use cases defined in WP2 we identified the relevant optimization problems, classified them in static and dynamic, and started

Final results

ORCHESTRA is progressing well beyond the current state of the art in several fronts. A competitive analysis carried out by the consortium showed that in addition to Telecom Operators (such as Orange and British Telecom) several optical technology vendors have started to show interest in approaches similar to ORCHESTRA’s use cases (provisioning with low margins, network adaptation), confirming the continued and increasing relevance of the project objectives and expected impact. To address the related market, three ORCHESTRA “components” were identified by partners as candidate for exploitation: monitor enabled receivers, a network planning and operation tool module, and Yang-based control/management modules and hierarchical monitoring. Moreover, detailed techno-economic studies to support the exploitation of the ORCHESTRA concept in the metro, continental-size and in core WDM networks were also provided. A patent in relation to the developed control plane is in the final phase of being filed. Also, numerous publications in high-impact journals and conferences have been achieved that are being acknowledged by the community.

More details regarding the performed activities and the way these progress the state of the art are presented in the following.

Software Defined Monitoring and Dynamic Network Architecture
The monitoring capabilities of optical networks are rather basic, while monitoring information is used only for failure management and is not used in optimization decisions. ORCHESTRA explores the possibility of an optical network that harvests and makes a more efficient use of the monitoring capabilities that are already present in the optical networks, and evaluates the potential of introducing new/advanced monitoring features.
ORCHESTRA explores the idea of using real monitoring data, collected using DSP from coherent optical transceivers, which can be extended, almost for free, to also serve as software defined optical performance monitors (soft-OPM). The use of coherent receivers as active monitors of the performance of the lightpaths is an improvement with respect to the actual monitoring capabilities available in the management suites used today in optical networks. These management suites use hardware monitors or digitally framed overhead bytes or analog optical power monitoring points. For the evaluation of the ORCHESTRA advanced monitoring concept, TILAB, described a number of use cases in different network contexts (metro, core). An important use case, that of provisioning lightpaths with reduced margins, showed that the ORCHESTRA monitoring features can postpone the investments and achieve CAPEX and OPEX costs saving. Field experiments were performed o provide input to the various research directions and support the development of the DSP monitoring algorithms.

Development of Flexible Optical Transceiver and Software-Defined Optical Performance Monitoring Algorithm Suite
NTUA and NKBL are involved in ORCHESTRA developing a flexible optical transceiver, a DSP-based suite enhancing the modulation-demodulation operation with optical performance monitoring capabilities.
Coherent systems have become a reality for the deployed optical networks enabling the upgrade to 100G and 200G rates employing DP-QPSK and DP-16-QAM formats over a single wavelength. The forthcoming upgrade to 400G systems requires efficient hardware implementations and software techniques in the optical transceivers in order to achieve this significant increase in capacity while minimizing the margins for operating the system in its full potential. Furthermore, optimum utilization in terms of capacity and spectral efficiency is also required considering the different services demands and the system adaptation due to degradation between BOL (begin of life) and EOL (end of life).
The ORCHESTRA transceiver copes with the above challenges by implementing an AWG-based transmitter solution with a commercial IQ modulator offering a fle

Website & more info

More info: http://www.orchestraproject.eu.