Adel A. M. Saleh

All-Optical Networking—Evolution, Benefits, Challenges, and Future Vision

While all-optical networking had its origins in the research community a quarter of a century ago, the realization of the vision has not had a straight trajectory. The original goal of the all-optical network was based on keeping the data signals entirely in the optical domain from source to destination to eliminate the so-called electronic bottleneck, and to allow arbitrary signal formats, bitrates, and protocols to be transported. The latter property is referred to as transparency. When all-optical networks were finally commercialized around the turn of the century, however, a modified reality emerged; the quest for transparency was replaced by the more pragmatic objective of reducing the network cost and energy consumption. Moreover, especially for networks of large geographical extent, electronics were still present at some (relatively few) points along the data path, for signal regeneration and traffic grooming. This modified vision captures the state of todays networks, though terms like all-optical and transparent are still used to describe this technology. However, continued advancements are bringing back some aspects of the original transparency vision. In this paper, we review the evolution of all-optical networking, from the early vision to its present vibrant state, which was made possible by great advances in optical transmission and all-optical switching technologies. We describe the numerous benefits afforded by the technology, and its relative merits and drawbacks compared to competing technologies, sometimes referred to as opaque. We also discuss the remaining challenges and future directions of all-optical networking. While all-optical solutions permeate todays access, metro, and core networks, this paper focuses on the core.

Wavelength-Selective CDC ROADM Designs Using Reduced-Sized Optical Cross-Connects

As optical transport networks become more configurable, the need for reconfigurable optical add/drop multiplexers (ROADMs) that are colorless, directionless, and contentionless (CDC) increases. Either of the two major classes of ROADM architectures, i.e., broadcast-and-select and wavelength-selective, can provide the CDC properties. With broadcast-and-select ROADMs, this is generally accomplished by employing more costly or complex configurations or adding adjunct equipment. In contrast, the wavelength-selective class of ROADMs, where the central component is an optical cross-connect (OXC), inherently provides the CDC properties. However, this architecture is limited by the achievable size (i.e., port count) of the OXC. We present alternative wavelength-selective ROADM designs that require smaller OXCs while providing equivalent functionality.

Data Center Energy Efficiency:Improving Energy Efficiency in Data Centers Beyond Technology Scaling

The paper identifies four key themes that should be addressed to improve energy efficiency in data centers: These include understanding the newly emerging workload space to ensure proper matching of architecture design to application requirements; provisioning data center resources for predicted future workloads and applications, and improving the energy proportionality of individual machines to correlate power consumption to load; improving vertical integration within the software stacks of communication, storage, and runtime systems; and establishing standards for hardware and software communications to allow the integration of new technologies and component designs.

Elastic WDM crossbar switch for data centers

A fast WDM crossbar switch suitable for interconnecting electronic packet switches in scalable data centers is proposed and analyzed. It is capable of elastic multiple-wavelength interconnection of any node pair. The switch is compatible with silicon-photonic integration; thus it promises small size, power and cost.

Analysis of internal ROADM protection

Reconfigurable optical add/drop multiplexers (ROADMs) are a key component in most carrier networks. One common ROADM design utilizes a set of wavelength-selective switches (WSSs) in a broadcast-and-select architecture. All traffic at the node passes through at least one WSS. If a WSS fails, the affected traffic can be routed over spare network capacity, similar to a link failure. Alternatively, WSS redundancy can be provided internal to the ROADM. We discuss the advantages of internal ROADM protection, and analyze two architectures for providing protection for the WSSs.

Scaling-out Data Centers Using Photonics Technologies

Novel architectures are presented, employing energy efficient photonics WDM or SDM technologies, that are shown to enable scaling-out of leaf/spine data center architectures economically and to allow their elasticity, without increasing latency.

Evolution of the architecture and technology of data centers towards exascale and beyond

The continual growth of data centers cannot be maintained through incremental advances in todays technologies and architectures. An alternative architecture based on reconfigurable WDM optical networking technology is presented for realizing future, exascale, elastic data centers.

Quantifying the value of wavelength-add/drop in WDM rings

In WDM rings, the value of wavelength-add/drop rises steeply with ring size and/or internodal demand. We present a bound on attainable equipment savings, and present a methodology for performing the bundling for rings carrying uniform all-to-all traffic, and rings carrying uniform hubbed traffic.

On-chip wavelength locking for photonic switches

We present an on-chip wavelength reference with a partial drop ring resonator and germanium photodetector. This approach can be used in ring-resonator-based wavelength-selective switches where absolute wavelength alignment is required. We use the temperature dependence of heater resistance as a temperature sensor. Additionally, we discuss locking speed, statistical variation of heater resistances, and tuning speed of the switches.