Cross-layer techniques for dynamic optical networking

Related Research

Engineering Research Center (ERC): Center for Integrated Access Networks (CIAN)

Dynamic Optical Networking

Engineering Research Center (ERC): Center for Integrated Access Networks (CIAN)

Ever-growing demand for speed and bandwidth and increasing energy consumption in today's networks are driving the need for intelligent next-generation networking architectures that can overcome fundamental spectral and energy limitations. CIAN, an NSF-funded engineering research center is a multi-university consortium that aims to overcome the bottle-necks the existing static, tiered aggregation network by transforming it into a dynamic, on-demand service-aware network. Our role, as the CIAN networking thrust lead, is to develop this envisioned intelligent access/aggregation network that will seamlessly interface with the edge/core and dynamically deliver most demanding high-bandwidth applications at low-cost and extreme energy efficiency.


One of the key building blocks of our intelligent access/aggregation network is a hybrid electronic and optical switching node that we have designated as the cross-layer enabled optical network element (CLONE). The CLONE provides real-time introspective access to the optical layer in conjunction with awareness of higher layer network constraints (e.g., application, quality of service requirements, and energy consumption) to make more informed routing, regeneration, modulation, and power control decisions. This intelligent interaction between the network layers will result in improved network performance and enable scalable aggregation of heterogeneous service traffic. Another major effort is the network emulation test-bed which is used to replicate the transmission impairments and dynamics for networks of variable size and topology. This test-bed allows us to stringently study any system of interest under real network conditions, enabling us to evaluate and validate the performance of the evolving CIAN box and other novel photonic technologies.


Related publications


Top

Dynamic Optical Networking

Dynamic optical networking, where bandwidth is allocated on-demand in the physical layer in response to changing traffic demands and network conditions, is a promising solution to address the energy and capacity challenges in the current network. However, rapidly reconfiguring light-path connections causes unpredictable transmission impairments and result in network instability.


Physical Layer Techniques


To achieve dynamic networking capabilities, it is first necessary to fully understand the additional network instability and uncertainty that will be caused by dynamic operation so that compensatory methods can be developed. We use the network emulation test-bed to study and characterize the transmission impairments and dynamics that affect WDM circuit switching and all optical aggregation networks. Since the impairments under dynamic operating conditions are unpredictable, the optical layer has to be monitored real-time so that the appropriate compensatory techniques can be rapidly executed. Ubiquitous deployment of monitoring devices is thus key to achieving a dynamic optical network. However, this will be a feasible solution only if the monitoring devices and cost-effective, energy efficient and supports multiple data formats. We are focused on developing fast monitoring solutions for advanced data formats and enabling these devices to function autonomously. We are also investigating methods of re-configuring light-paths such that accompanying dynamics do not cause data loss.


Algorithms


Our collaborator, Dr. Gil Zussman, has developed cross-layer network control algorithms to stabilize optical signals in rapid light-path reconfiguration scenarios. These algorithms builds on the capabilities of new devices and will allow efficient use of optical resources through dynamic network configuration, and physical layer power, bandwidth, and modulation control. The algorithms are used in the network emulation test-bed along with new monitoring devices to investigate the time to converge to a stable operating point in dynamic networks of various size, topology and complexity.



Related publications


Top