Broadband confronts cost barrier

It is no secret that the telecom industry as a whole is in full retreat. Prices are falling faster as demand growth slows. Most would agree that the challenge in broadband is to address what had been a supply-side problem, rather than the belief that the demand is constrained.

To be sure, broadband demand isn't unlimited, given the current price point of $40 to $50 per month. And as long as there isn't a killer, bandwidth-consuming application that drives more users to switch from dialup to broadband service, broadband demand will plateau. Aside from interactive online games, few see the justification to pay higher prices for broadband. Even though most users who have experienced the exhilaration of broadband online would not go back to dialup, high prices and the lack of a need will keep many users on the sidelines. At this stage of the market economy, price elasticity for broadband service remains high: The lower the prices, the higher the adoption rate.

Fixed wireless access has made a moderate but significant penetration into the broadband market, particularly the underserved suburban and rural markets. The technology has matured in recent years, and reliability is improved. Of all the technologies on the market, fixed wireless is the most economical means to deliver broadband service to underserved markets. In communities where cable infrastructure is nonexistent and DSL does not reach, fixed wireless may be the only real alternative. Moreover, it inherently has cost advantages. Wireless service eliminates the need to dig trenches and lay wires. Time to installation can be measured in days, not weeks or months.

Fixed wireless is best-positioned to provide the most cost-effective broadband reach to suburban and rural areas-markets that had been underserved by large independent local-exchange carriers (ILECs) and regional carriers. As the cost of equipment and deployment continues to fall, so should the price of broadband services, which in turn would fuel demand. As broadband goes mainstream, it will spur content development of bandwidth-intense applications. There is ample opportunity, provided the industry can first deliver on online users' quality and reliability expectations while meeting service providers' business models.

If fixed wireless is to penetrate the market in a meaningful way, suppliers must deliver reliable performance as well as provide a compelling business model for carriers and service providers. Previously, major rollouts have been in the licensed multi-user, multipoint distribution system frequency spectrum, costing the carrier billions of dollars upfront to acquire it. Although it created a high barrier of entry for competitors, the real problem is the high cost burden placed on carriers themselves in the early stages of rollout without a large subscribers base to support it. It became impractical to amortize the huge license fee within a reasonable time.

With the licensed band in flux, carriers are rethinking the application of the 2.5-GHz licensed frequency for 3G cellular. Doing so would leave the fixed wireless market with little alternative but to migrate to the unlicensed bands. Although 2.4 GHz is available, it is very congested with traffic from wireless LANs (802.11b and HomeRF), cordless telephones, Bluetooth and even microwave ovens. Many ISPs thus shy away from that interference-prone band, leaving the 5.8-GHz and 5.3-GHz Unlicensed National Information Infrastructure (U-NII) bands as the only viable frequencies for broadband deployment.

Unlike the 2.4-GHz band, which is practically a free-for-all to use, the U-NII bands only allow the high-powered, point-to-point and point-to-multipoint broadband access systems to operate. All other medium- and low-power indoor wireless devices have been allocated to other subbands. Thus, interference is expected to be low.

Most subscribers are indifferent to the technology that delivers Internet connectivity to their home or office; they simply want reliable service. Consumers don't easily forget incidences of failed connections, slow transmission speeds and service outages. Reliable service and consistent performance rank high among their priorities. This is particularly true for business users who rely on their Internet connection for vital transactions.

An ISP can thus gain a competitive advantage by investing to ensure that the equipment deployed is up to par. That includes the customer premises equipment. The cost of replacement consists of one or more truck rolls and possibly the cost of replacing the equipment itself. So the stakes are high and the margin of error small.

Better connections

Wireless faces still other challenges that wired networks don't. Tall buildings, tree foliage and terrain may obstruct radio signals, degrading a connection's integrity. From the perspective of the radio, multipath interference has to be addressed.

Anything that will increase the signal strength from the transmitter to the receiver will improve connection reliability and coverage. That could reduce the need for truck-roll installation, allowing for a lower-cost business model based on installation by the user.

Manufacturers are racing to develop technologies and device schemes to bring that about. Attempts to date have included variations of orthogonal frequency-division multiplexing and vectored OFDM, with simultaneous modulation used on multiple carriers to transmit large amounts of data. The systems have achieved measurable signal gain improvements under multipath environments-at a trade-off of more complex electronics and increased equipment cost. The technology has seen some deployment but has fallen short of expectations because of the excessively high cost.

Others are developing smart-antenna technologies to perform beam forming so that signals are directed to the target with high precision. The attempt is to increase the signal gain at the receiver. Unless the gain improvement is significant, it is not likely to help to overcome the true non-line-of-sight problem facing network builders. For example, a 6- to 10-dB improvement in signal gain will not be sufficient to offset a 25- to 30-dB deficit after a signal passes through a stucco wall of a typical home or office. Most of the technologies are in various developmental phases; few are actually in field trials.

Others are attempting to leverage off the 802.11b LAN standard by creating a "mesh" architecture. Here, as access nodes are deployed, each node has built-in capabilities to route Internet Protocol packets through to other access points, effectively relaying the packet data to an adjacent node. Thus, as a network builds out with more and more nodes, packet relaying increases exponentially.

In concept, at least, the network expansion is self-perpetuating. This is not unlike a large backbone network, in which IP packets "hop" through several routers before reaching the destination. The difference is that it is implemented in a microcell architecture. Mesh network technology is in its infancy. Field trials are under way.