Project Title:
Wireless Internet Post Office

Recipient Institution:
Department of Computer Science & Engineering
Hauz Khas, New Delhi 110016, India

Project Leader: Prof. Huzur Saran

Amount and Duration: USD $ 29,630 / 12 months

Project background and justification
Cost is the biggest impediment to providing digital connectivity to citizens of the world. For if cost was not an issue, a satellite dish, a PC computer, and array of solar panels and batteries could be used. The next impediment is the subtler issue of matching the technology and applications to the needs and abilities of the users. Putting an illiterate peasant farmer in front of a networked PC will not help him get better prices or fight the disease that is killing his crops.

This research project takes a fresh look at breaking the digital divide by considering the entire system, from internet connection to end users, addressing technical, social and economic issues. The research output will be a complete how to guide, posted on a publicly accessible web site, that small companies, universities and governments can use to build and deploy a wireless infrastructure to deliver simple but essential digital communication to millions of people across the Asia-Pacific.

The research project will focus on asynchronous text messaging (email, news reports, bulletins, news groups, etc.) for it places the lightest system burden (bandwidth, storage, processing power, latency tolerance) and provides the essential content. The majority of the research work is the design of an end-to-end system that delivers text messaging in the most economical, robust, practical and appropriate way for the largest population possible. The project is called a Wireless Internet Post Office because text messages are read and collected (off-line) in the field with a commercial hand held computer (PDA) and sent and received by synchronizing with a wireless relay stations.

Project objectives
The research project objectives are to solve the multitude of technical and system design issues that define a Wireless Internet Post Office (WIPO) capable of delivering digital access to the widest possible population at the lowest usage and capital cost. The research project will design, test and document the system components in sufficient detail to enable others skilled in the art to build and deploy the system.

Project beneficiaries
The research project is to design a complete system from internet gateway to end user. Ultimately it is the end users who are served by the WIPO, including villagers who want to contact relatives in distant locations, farmers who want pricing, access to markets, and advice on disease and pest control, small businesses who want access distant markets, educators who want teaching material, and medial and aid workers. The WIPO will also benefit people involved in the wireless network, for each component can have an economic model for income to encourage deployment and propagation. A networked PC operator provides the internet gateway connection and can provide news letters for a food coop, price and vendor list for fertilizer and fuel, news clipping and translations, and other information services to end users by subscription or request. The wireless relay station owners can charge an access fee to client PDAS. The WIPO creates an entrepreneurial opportunity for PDA owners to become micro-businesses, providing services including scribing, data collection, crop pricing, referrals, matching buyers with sellers. A literate village who can purchase a PDA can become an information worker, similar to the entrepreneurial model created by the Grameen Bank of Bangladesh, whereby villagers (mostly women) are given micro-loans (e.g. <$100) to purchase a cell phone and selling calls to villagers.

Project sustainability and outputs
The project will create a web site to disseminate the know-how to build and deploy a WIPO system. The web site will survive the project to keep the information and results publicly available, and serve as a focal point for discussion, links to similar work, supporting material, and follow-up activities. The specific deliverables are specification documents and test results that resolve the system component design issues posed in the project methodology.

Project methodology
The design goal is to deliver reliable asynchronous text communication at minimum cost with strong emphasis on what is practical and most likely to succeed in the field. The project is organized by system component. A minimum configuration of a WIPO, shown in Figure 1, is a single PC with a internet connection, a multitude of wireless relay stations that bring the digital communication out into the rural villages, and many PDAs that synchronize (up and down load text messages) with the wireless relay station. A battery powered PDA enable the owner to roam freely, visiting clients (farmers, villagers, small businesses, health clinics, homes, schools, and offices), entering and reading text messages and data with the PDA off-line (unconnected). The PDA owner must get within proximity, by cable, radio, or infra-red link of a wireless relay station to synchronize. The combination of roaming PDA and stationary wireless synchronization stations provides the lowest cost and most flexible means to collect and disseminate information in developing countries.

Figure 1: Wireless Internet Post Office Architecture

The following are the system components and some of the fundamental issues the research project will address.

Base Station

The base station is a PC connected to the internet and is the gateway between the internet and the roaming PDA. Since text messages do not consume much bandwidth or storage, a single PC can serve thousands of PDAs. The research project will investigate;

How to look like one account to take advantage of free email services around the world (e.g. a www.yahoo.com email account)
How to make asynchronous web searches with an off-line PDA (e.g. using www.google.com)
Authoring tools and services for the PC server operator (e.g. create newsletters, special interest groups, clipping service)
How WIPO system scaling with PC server, wireless relay stations, PDA and end-users
PC server requirements (processing speed, RAM, disk size, internet connectivity speed)
PC server reliability, robustness and fault tolerance
How much off the shelf and freeware software can be used for the server PC
PC server operator skills requirements
PC server installation, software maintenance and remote monitoring
Hacker and virus protection
Cost and kinds of internet access in Asia-Pacific (e.g. mobile or land line phone, ham radio packet radio, ISDN, DSL, cable, satellite).

Wireless Networking Architecture
A multi-point-to-multipoint mesh is the preferred architecture to network the wireless relay stations. This robust architecture is used by the internet so a message can get through even if nodes are down or out of range of the PC server. Since asynchronous text messages are tolerant of large lag times, many wireless relay stations (hops) may be used, spanning large distances between PDA synchronization station and base station internet connection. The important factor is reliability to insure the message gets though. Providing several redundant paths greatly increases the reliability of the communication system. The project will design and simulate the wireless networking architecture, and test wireless components using off-the-shelf high performance low-cost commercial radios (802.11b WLAN cards). The research project will also investigate;

Wireless network protocol, including packet size, error correction and detection
Topology performance, scalability, hidden-node problem, robustness
Communication security
Tradeoff of modifying radio firmware to get better performance (e.g. running a slower data rate to gain longer range), restricting design to particular vendors and radios vs. unmodified radio firmware (e.g. running standard 1 Mbps DQPSK that has been demonstrated to provide >5 km range using Yagi directional antennas)
Power management (e.g. sleep vs. active time, long recovery time)

Wireless Relay Station
The wireless relay stations bring text messaging out to the remote villages and synchronize local PDAs. They are solar powered so they can be placed anywhere, use line-of-sight directional antennas for long range (kilometers). Any number of them can be placed in series to cover vast distances since latency is not an issue for asynchronous text messaging. The wireless relay station consists of a high performance low-cost WLAN radios (<$100 each), a controller (e.g. StrongArm or digital signal processor) to implement the radio protocol, and a link (wire or wireless, to be determined) to local PDAs.

The project will design the wireless stations to be practical in the field, including easy to set-up with minimum skills and tools, robust and easy to replace. Remotely configuration and monitoring is used to study system component performance and anticipate problems and failures. Research topics include;

Radio operation, station installation and monitoring software
Antenna performance and tolerance to alignment, tower interference
PDA synchronization software
Hardware design and requirements (processor speed, DRAM and FLASH memory)
Mechanical (weatherproofing, ergonomics of installation, maintenance)
Remote firmware update mechanism
Security (eavesdropping, hack and virus protection)
Link from PDA to wireless relay station (cable, optical, RF)

Wireless Station Power Plant
The wireless station power plant consists of a solar panel, rechargeable batteries and an electronic charging circuit. The wireless station power plant provides all the power for the wireless relay station, making it self-sufficient for remote locations (e.g. mounted on a mountain ridge to bring connectivity to a village located in a valley). Research issues include;

Power budget
Solar panel and battery capacity, size, vendors, cost, lifetime, temperature range
Operation during extended no-sun conditions
Install, remote monitoring, maintenance, repair, and replace

Client PDA
The client PDA is typically owned by an individual who pays for usage time on the wireless network. Research issues include;

Basic software applications
Vendors, costs, availability, operating systems
Communication link (IRDA (infrared), RS-232 cable, USB)
Loading software
Language support
Character representation (ASCII (8 bit), Unicode (16 bit))

Economics and Business
Anticipated order of magnitude costs for the components are $5,000 for host PC and internet connection, $1000 for wireless relay stations (no installation) and $100 for client PDA. Research issues include;

• System component costs
• Usage models and scenarios for client PDA owners (e.g. village scribe, health/aid worker, business person, school teacher, government official)
• PDA registration and access payment process
• Component ownerships and payback schemes
• Prevention of monopoly, price gouging, creating and spreading misinformation 

Project time-line
The project will last one year, with requested funding to sustain the web page for an additional year.

Project Monitoring

The web site shall post quarterly research reports, a final report and serve as a repository for research result output, and conduit for discussion among colleagues, researchers, small companies, governments, universities, and individuals around the world.


Read the Abstract of Project

Read the Progress Report

Read the Technical Report (PDF, 400 kb)