Wednesday, July 13, 2005

Passive Optical Networks - Coming Back to Light

INTRODUCTION

The last mile, local loop, access network, these are various terminologies used to qualify that part of a telecom network, which provides broadband services to business and residential users. It’s this last mile that remains a potential bandwidth bottleneck between a high capacity LAN, a home network and a backbone.

There has been a deadlock kind of a situation between consumers, content providers and network operators. More bandwidth requires more revenue from the consumers. Consumers don’t want to pay for more bandwidth because they don’t receive any service that requires more bandwidth; the Service Providers are not providing any new services because there is no network infrastructure to support these kinds of services and Network Operators are not willing to provide services, because consumers are not paying for bandwidth!!

There has been a resolution to this problem as far as long haul networks are concerned with the advent of optical backbones on the long haul networks leading to incredibly high speeds of transmission and more available bandwidth, but when it comes down to the local exchange, copper still prevails because it’s the cheapest option. Thus the consumers are not able to enjoy the benefits of high bandwidth access. The solution to this dilemma is seen in the form of optical access networks called “Passive Optical Networks”.

THROWING LIGHT ON PASSIVE OPTICAL NETWORKS

A Passive Optical Network is an optical network, which takes optical fiber right to the customer premises instead of copper. The idea of an Optical Access was conceived in 1995 by a consortium of carriers and vendors called the Full Service Access Network. (FSAN) This group now comprises of companies like Alcatel, BellSouth, Nippon Telephone and Telegraph, Fujitsu, SBC, Lucent, NEC, British Telecom, French Telecom and so on.

A Passive Optical Network is called so, because it does not use any Active electronic components (devices consuming power), from the central office to consumer’s premises. The network carries a single strand of fiber, which undergoes a multiway way split to serve 32 consumer installations. This splitting is achieved by means of a passive electronic device called a splitter. The entire process is illustrated in the diagram below.





Source: SBC Communications inc.

This single strand of fiber is capable of carrying network traffic at a rate of 155.52 Megabits per second! , Which is way beyond what existing services (like DSL) offer, thus making this, such an attractive option.


THE BANDWIDTH POTENTIAL

The reason people keep rhapsodizing about the bandwidth factor, is because of potentially wide range of services that it can generate to the consumers. Capitalizing on higher bandwidth access can make the Internet a more viable business place. A few ideas that would have run for the Oscars in the sci-fi category (if there was one!) could become a part of day-to-day life! A few services that consumers could hope to have would be:
a) Video on demand: To watch your favorite blockbusters or TV shows on demand and you would be paying only for what you watch, and the providers would gain by advertisements.
b) News Reports on demand: You would have the choice of which news feed you would like to hear from an array of reporting services
c) Video Conferencing: Colleges would be able to use video conferencing as a tool for distance education.
The major paradigm shift that a higher bandwidth could initiate is a migration from the traditional push technology to a pull technology. The consumers will now have the option of choosing the service that they would like to receive and need not be constrained by what is dished out to them. In turn content service providers could provide a more eclectic blend of services, and the network operators would stand to gain by providing the necessary infrastructure, that hold so much promise. Does this not break that deadlock that we were looking at earlier?!

LIGHT AT THE END OF THE FIBER?

Communications Industry researchers inc, (http://www.cir-inc.com/) have forecasted the following revenues for the US PON Market by 2004.
Fiber to the Home 69.3 Million
Fiber to the Curb 468.5 Million
Fiber to the Building 225.1 Million
Leading to a gross revenue estimate of 762.9 Million dollars.

Passive Optical Network Technology is a part of an evolution, and has all shades of being a viable alternative to the Last Mile Bottleneck at the access layer. There are a lot of hurdles this technology has to overcome, for example the lack of availability of fiber in the access networks, lack of vendors, distance limitations imposed by the lack of amplification or regeneration on the fiber. Barring these few limitations, Passive optical Network is currently the most Hip Hep and Hyped, access layer technology in the wings.


REFERENCES

At UT Arlington there are currently two courses offered by the department of computer science that provides valuable research experience into the fields of access and metro networks and optical networks in general. These courses are Telecom Network Design (CSE 5347) offered by Prof Kalyan Basu and Optical Networks (CSE 5392) offered by Dr Gergely Zaruba.


There are lots of interesting resources available on the net which offer lot of information on passive optical networks, a few of them are:
http://www.lightreading.com/: Carries a lot of articles on optical access technologies
http://wwwcsif.cs.ucdavis.edu/~kramer/: Research on PON by Glen Kramer of UC Davis
http://www.terawave.com/: A company that is currently providing Optical Access Solutions.

No comments: