the project
Project Overview

Over the last decade, we have witnessed an explosion in wireless access to the Internet. In 2002, revenue from IEEE 802.11 (WiFi) network cards and access points totaled an estimated $2.1 billion on 23.9 million devices with 73% growth predicted for 2003 alone. Moreover, advances in the physical layer and media access protocols have enabled transmission rates of 54 Mb/sec in IEEE 802.11a, and even higher rates are projected in future revisions. To date, the overwhelming majority of deployed WiFi networks are in the enterprise or home. Consequently, high speed wireless data communication can only be achieved from a small subset of wireless "islands," and we remain in the infancy of achieving a long standing vision of a high-speed ubiquitous wireless web.

system architecture
Figure 1: System Architecture

There are two simultaneous efforts to providing wireless Internet outside the home or office. The first is deployment of "WiFi hot spots." For a service provider like T-Mobile, a hot spot consists of an IEEE 802.11b access point and a Cisco router with a T1 (1.5 Mb/sec) interface to T-Mobile's wired backbone. While each WiFi hot spot arrives with great fanfare, the preliminary results are not encouraging. At the end of year 2002, the U.S. had approximately 3,000 hot spots, with T-Mobile operating the most with approximately 1,600 (Europe has 1,150 all together). The total U.S. subscriber base for 2002 was estimated at 20,000, resulting in a net revenue of a mere $2 million for 2002, and yielding a large net loss given fixed costs as high as $10k per hot spot and recurring costs of approximately $400/month per hot spot. Perhaps ironically, the overwhelming costs of providing wireless hot spots is due to fixed and recurring costs of the wired infrastructure.

As a consequence of such low deployment, the resulting "coverage" is dismal. Even optimistically estimating that each hot spot covers 100x100 m2, the 3,000 hot spots amount to a mere 30 km2 in the entire U.S., or an average of 0.6 km2 per 50 metro areas covered. Even adding Cometa's plans for an additional 20,000 hot spots, coverage will increase only to approximately 4 km2 per metro area, representing approximately 0.4% of the area of a moderate sized city such as Indianapolis. Thus, today's hot spot architecture is slow to deploy, costly, and unscalable, and is not on any path to provide large-scale coverage.

A second major effort is 3G and fixed wireless services such as LMDS. In both cases, speeds are typically 2 to 3 orders of magnitude slower than WiFi, with maximum per-user speeds in the 100s of kb/sec range. Moreover, because of multi-billion dollar spectral license costs and high infrastructure costs, such systems have proven costly to deploy and hence lead to expensive, yet moderate speed, wireless Internet services. Thus, while having the promise of near-ubiquitous coverage and allowing high mobility speeds, such technologies have significant performance and cost limitations. Moreover, given their small Internet subscriber base, scalability to many data users remains unproven. Therefore, despite a decade of strong progress in wireless data communication, it is clear that with the current evolutionary path, a large-scale high-speed wireless web is not on the horizon.

Home  |  The Project  |  Architecture  |  Research  |  Presentations  |  Participants  |  Funding  |  Contact
Copyright 2006 Transit Access Points (TAPs).
All Rights Reserved. Problems with this page? Contact us.
Website designed by
Transit Access Points (TAPs)
ECE Department, MS 380
Rice University
6100 Main Street
Houston, Texas 77005-1892