Capacity Efficiency and Restorability of Path Protection and Rerouting in WDM Networks Subject to Dual Failures

Resilient optical networks are predominately designed to protect against single failures of fiber links. But in larger networks, operators also see dual failures. As the capacity was planned for single failures, disconnections can occur by dual failures even if enough topological connectivity is provided. In our approach the design of the network minimizes the average loss caused by dual failures, while single failures are still fully survived. High dual failure restorability is the primary aim, capacity is optimized in a second step. For WDM networks with full wavelength conversion, we formulate mixed integer linear programming models for dedicated path protection, shared (backup) path protection, and path rerouting with and without stub-release. For larger problem instances in path rerouting, we propose two heuristics. Computational results indicate that the connectivity is of much more importance for high restorability values than the overall protection capacity. Shared protection has similar restorability levels as dedicated protection while the capacity is comparable to rerouting. Rerouting surpasses the protection mechanisms in restorability and comes close to 100% dual failure survivability. Compared to single failure planning, both shared path protection and rerouting need significantly more capacity in dual failure planning.     

Lower Bounds on the Redundancy of Link-Recovery Mechanisms

The paper derives lower bounds on the redundancy, the protection capacity to working capacity ratio, needed by linkprotection or link-restoration. Homogeneity conditions, on which the well-known redundancy bound 1/(d?1) holds for a network, are also identified, where d is the average nodal degree.     

Optical Network Design for a Multiline-Rate Carrier-Grade Ethernet Under Transmission-Range Constraints

Ethernet is a success story in local area networks (LAN). Efforts for extending its boundaries beyond LAN to the carriers' backbone networks are in progress. We study the problem of designing reliable and cost-efficient high-rate (100 Gbit/s) carrier-grade Ethernet in a multiline-rate optical network under signal transmission-range constraints. Reliability is achieved using shared-path protection at the connection level (Ethernet tunnel in this study). We construct an auxiliary graph, called mixed topology (MT), using which it is possible to: 1) identify traffic grooming possibilities; 2) select a path which requires the minimum amount of 3R regeneration; and 3) effectively choose the data rate of the channel to be established. Our algorithms, tested on the 17-node German network, resulted in lower network cost and higher resource utilization compared with other schemes.     

On Hamiltonian cycles as optimal p-cycles

Using Hamiltonian p-cycles, it can be shown that p-cycle design is able to reach the logical redundancy bound of 1/(d~-1) where d~ is the average node degree. We formulate two conditions on which the design is able to reach this bound if and only if Hamiltonian p-cycles are used.     

The integration of alternative contact cleaning techniques for future DRAM technology nodes

Conventional DRAM contact to Si cleaning methods are using wet HF-based chemistry to remove the native SiO₂ from the contact bottom and reduce the contact resistance [M.R. Baklanov, et al., J. Electrochem. Soc. 145 (9) (1998) 3240-3246]. With further scaling of the contact dimensions the distance between the contacts to the bitline (CB) and the adjacent gate conductors (GC) is decreasing. This is leading to increased parasitic capacitance between the CB and the GC and increased total bitline capacitance. The high bitline capacitance results in large signal delays, poor transfer ratio and thus bad signal margin and worse retention. In addition it is a general limitation for the number of cells attached to this bitline. For future DRAMs the need for small chip size is driving the reduction of the bitline capacitance and keeping the distance between the CB and the GC becomes a critical challenge. Typically, the material separating the CB from the GC is LPCVD TEOS. Furthermore, the selectivity of the wet HF-based chemistry of the as-deposited LPCVD TEOS to thermal SiO₂ is in the order of 3:1. This is resulting in high removal amount of LPCVD TEOS for a given removal of native SiO₂.     

Integrated design and operation of a transparent optical network: a systematic approach to include physical layer awareness and cost function [Topics in Optical Communications]

The manner and exact timing of the evolution of the widely used DWDM infrastructure to a transparent optical network is judged on economic circumstances and network performance. In this context this article elaborates on a methodology developed within the Project 1ST NOBEL, for integrating cost functions and other design constraints into the planning of a DWDM network while considering this evolution perspective to a transparent optical network. This systematic approach builds upon the experience of static network design and is substantiated by specific examples in this article     

Configuration of p-Cycles in WDM Networks with Partial Wavelength Conversion

The p-cycle concept offers a capacity-efficient and rapid protection mechanism for mesh-restorable networks. This work investigates the configuration of span protecting p-cycles in wavelength division multiplexing (WDM) networks with limited wavelength conversion. An important point of view is the relation between the costs associated with the number of required wavelength converters, and the protection capacity-efficiency achieved. We formulate mathematical models and solve the respective optimization problems for a pan-European network as a test-case. An interesting finding is that the total number of converters required for the network as a whole can be greatly reduced, with only a small increase in spare capacity for protection by a strategy of associating wavelength converters with the access points between a pure wavelength path (WP) working layer and a set of pure WP p-cycle protection structures.     

Automatic protection switching for p-cycles in WDM networks

p-Cycle recovery relies on a protection switching protocol. We detail several issues for such a protocol taking the evolution from ring networks to p-cycles into account. In particular, we propose and evaluate a protocol enhancement to provide means for node failure protection. For the evaluation, we describe an integer linear program, which is applied to network design case studies, and formulate availability models for p-cycles. The case studies show that the protocol enhancement improves availability at marginal additional design cost.