Why do this now?

Timetable:   The timetable for this project is being driven by the BOREAS project partner institutions who want to have BOREAS operational ASAP.   The goal is connectivity to Chicago by spring 2007.   By collaborating, we improve the probability that we can successfully acquire dark fiber, as well as provide a cost savings in lighting the fiber.   The benefits of this type of collaboration will probably not be available to us in the future.

Reduction in cost: By participating in a larger consortium, each institution should be able to negotiate a volume discount with the selected vendors.   We'll also be able to get greater capacity & better survivability at more reasonable expense than we'd be able to get separately.

Anticipated fiber cost increases: Experts project that fiber costs will increase 200-300% in the next 24 months.   Purchasing fiber now will help avoid this additional cost.

What is our current network connectivity?

Each institution has two 1 Gigabit/sec external connections that connect with the national networks in Chicago.   One of these connections is dedicated for Internet2 traffic (network activities with national research organizations and other academic institutions in the . edu domain). The other connection provides a path to the commercial Internet( e.g., Google, Yahoo, etc).

The Universities are also members of Internet 2 and (through the CIC) the National Lambda Rail.

How does this connectivity compare with our peers?

The table below compares current capability with peers:  

Institution	External network connection speed (Gigs/second)
U of Illinois	30
Ohio State	10
Indiana University	10
Purdue	10
U of Chicago	10
UT Austin	10
UCLA	10
U of Wisconsin	10
Michigan	5
Michigan State	5
U of North Carolina	2
Iowa	2
U of Minnesota	.622
Iowa State	.2

How will BOREAS improve our off-campus capability?

The BOREAS network is composed of two physical components: 1) “dark fiber,” and 2) electronics. Each of these components makes a unique contribution toward improving off-campus networking capability.

Dark Fiber  

What is “dark fiber?”
Telecommunications companies have deployed optical fiber cables around the world to connect the Internet. The capacity of this fiber to carry information is called “bandwidth.” The bandwidth of an optical fiber can be divided into multiple independent network circuits. The telecommunications companies usually use their equipment to "light" or activate the fiber and then lease part of the bandwidth to customers. However, large organizations sometimes purchase the fiber directly and light it themselves, thereby bypassing the telecommunications companies and controlling all of the bandwidth themselves. These purchased optical lines are called dark fiber because they are unused and unattached to optical equipment owned by the telecommunication companies.

What are the advantages of purchasing dark fiber?
Flexibility: With access to our own dark fiber, we can change the capabilities of our network by altering the electronics that control the signals sent over the fiber. This allows us to consider new paradigms for using the network without requiring cooperation from or agreement of the telecommunication companies.

Investment: The expected life span of optical fiber is 20 years. There are no technologies on the horizon to challenge optical fiber as the way to create very high bandwidth networks. Despite much change in the telecommunications industry, optical fiber use for high-speed communications has remained constant. Technologies such as DWDM (see below) are continuing to extend optical fiber capabilities, decreasing the need for any alternative media. The bottom line is that by purchasing optical fiber, it is possible to pay a fixed cost for media with extensible bandwidth. This is preferable to the current arrangement with telecommunications companies in which costs go up as bandwidth increases. (Adapted from http://www.osc.edu/oarnet/tfn/faqs.htm)

Network Electronics

What are these electronics?
Network optical electronics (sometimes called optronics) are the pieces of equipment that send and receive light via the fiber to define the capabilities of and control access to the network. Modern network electronics technology (such as DWDM, see below) provides a tremendous boost in performance of the fiber and allows for new ways to use the network that have not yet been considered. The electronics will be the most expensive part of the BOREAS network.

What is DWDM?
Dense Wavelength Division Multiplexing (DWDM) technology allows information to be sent down glass strands through pulses of colored light. The light can be divided into different wavelengths, or lambdas. Each lambda can currently sustain bandwidth of 10 Gigabits per second. Since the capabilities of a DWDM based system can be enhanced through the upgrade of the electronics and ongoing evolution of the technology, DWDM-based optical fiber networks have a vast and growing bandwidth capacity.

What Types of Projects will benefit from BOREAS? (draft)

Projects that will benefit from the BOREAS network are ones that include the following activities:

• Distributed Data Management
• Massive Computations
• Access to Remote Instrumentation
• Grid Computing
• On-line Meetings
• New Network Paradigms

Distributed Data Management

Data sets are becoming larger, more complex, and distributed.   Many of the disadvantages of being geographically distant to other data sources are eliminated by the BOREAS network.   Projects of this sort include the need to:

• Access large databases at other institutions.
• Access and aggregate data from multiple sources around the country.
• Provide off-campus access to databases located on campus.
• Participate in collaborative management of inter-institutional distributed databases.

Massive Computations

The ability to access computational resources (such as clusters and supercomputers) at other institutions and centers around the country is greatly enhanced by BOREAS.   Projects of this sort include the need to:

• Make use of the TeraGrid (supercomputing centers).
• Reduce the computational latency introduced by the speed (or lack) of the inter-institutional network.
• Participate in inter-institutional computational grids. (BOREAS is a necessary step to becoming a TeraGrid member organization, for example.)

Access Remote instrumentation

Several investigators work with instruments (spectrometers, microscopes, detectors, etc.) at other institutions.    The need for traveling to these instruments is reduced if they are network accessible and the network bandwidth is sufficient to support remote control of the instrument.   Often, the datasets from these remote instruments are quite large. BOREAS will enable or enhance the transfer of this type of data to each campus.  

Grid Computing

Grid Computing is a new technology that may produce the next networking “killer app.”   Grids enable access to a number of distributed computer resources through a simple user interface. Iowa researchers participate in a growing number of regional and national grids.   BOREAS will provide the network performance required to participate in these inter-institutional projects at the same level of connectivity as other institutions.  

On-line, inter-institutional meetings

The enhanced network performance provided by the BOREAS network will improve capability to participate in on-line inter-institutional meetings and to use distance learning to teach courses.   Projects that will benefit in this area include those that require:

• Richer content — BOREAS provides the ability to transmit more bandwidth intensive content such as Digital Video DVTS, 3D interactive graphics, and shared data and documents.
• Enhanced fidelity — BOREAS will improve the quality of the video signals used in video conferencing.
• Increased capacity – BOREAS will allow the size of virtual meetings to increase.   This is particularly useful when doing “virtual lectures” involving dozens of students.

New networking paradigms

Using dark fiber and DWDM(Dense Wavelength Division Multiplexing) technology, network technology researchers will be able to test new paradigms for using the network.   This could involve unique applications running on a dedicated channel.