Standards and technologies need work. Roughly half of PON products now on the market support the Full Service Access Networks (FSAN) specifications blessed last year by the International Telecommunication Union (ITU). These are based on asynchronous transfer mode (ATM) technology, so-called ATM PON (APON), and according to many sources are limited in terms of speed and flexibility. To date, just a couple of leading companies, Quantum Bridge Communications Inc. and Terawave Communications, have achieved 622 Mbit/s both upstream and downstream on PONs.

A new series of specs being explored by the First Mile Study Group of the Institute of Electrical and Electronics Engineers Inc. (IEEE) could support faster networks that support more users and are cheaper to deploy, proponents say. But Ethernet PON (EPON) solutions are still proprietary, and standards are far from "carrier ready." What's more, without the blessing of the ITU, it's unlikely large incumbent carriers, especially RBOCs, would be willing to risk big deployments.

Ironically, the best-selling PON is neither APON nor EPON. Optical Solutions has 23 paying customers who've adopted its proprietary radio-frequency PON. The company recently released an APON product called the FiberPath 400, however, and says that while it has just one paying customer for the new gear, the company anticipates most of the others will adopt it in short order.

Carriers are slow. One observer likens getting major ILECs and RBOCs to roll out anything new such as PON to "trying to turn a battleship around in the harbor." Evidence of this can be seen in the SBC trials, which have been "rolling out" for over a year, with the vendor selection announced just this past spring. Others have been in the talking stage for years. While many carriers are considering PON, they're clearly not in any rush.

Carrier projects are on hold. The telecom shakeout eliminated many CLECs that were prospective PON customers. And the key prospects that remain are strapped for cash, such as the RBOCs, ILECs, and MSOs (multiservice operators, such as cable companies that are moving to offer voice, data, and video services on their existing infrastructures). This situation has cut capex expenditures and put many new technology projects on hold.

Present access techniques aren't cutting it. Digital subscriber line (DSL) and other access technologies don't provide what carriers need to roll out a desired mix of data, voice, and video services, PON proponents say. "DSL and cable modems don't provide enough bandwidth and can't be used to create attractive services," says Mark McDonald, VP of product marketing for the Marconi Corp. PLC (Nasdaq/London: MONI) access business. CIBC's Bezoza is more direct: "DSL doesn't offer an economical solution for broadband access."

Replacing copper with fiber will drive PON. For many carriers, the cost of putting in fiber or copper is similar. But the potential payback isn't. The value of fiber to support more services is being augmented by reductions in the cost of singlemode fiber and construction of PONs.

Carriers need PON to compete. Today's carriers are going to need PON to stay competitive, particularly in the market for residential broadband services. "On the business side, cable operators are going to start attacking the business markets with a combination of voice, data, and video services," Bezoza says. This is going to force carriers to accelerate their fiber deployments via PON.

Vendors also say carriers will need PON to extend new services, such as storage and backup. Quantum Bridge, for instance, has just begun a program to work with SAN manufacturers on new storage network offerings via PON (see Quantum Bridge Joins SAN Vendors). "This will appeal to a huge amount of businesses that have more storage than a T1 can handle but less than required for Sonet services," says a spokesman for Quantum Bridge.