Moving Toward Ubiquitous Computing

William Papp, March 1998


    Mobile computing is on the increase as people find that they need to be "connected" in a variety of locations. Connectivity must be both easy and reliable. New devices are being developed and used to provide access to the internet and local networks. The increase in demand for connectivity and the introduction of new devices require that new computing paradigms be developed. This paper presents work being done in Wireless (mobile and nomadic) and Ubiquitous Computing. Technology including software, hardware and protocols are discussed.

Table of Contents

1. Introduction

There are many trends in technologies today that will effect how we interact with computers and how we access data. This paper will present some of these technologies, their implementation, and their limitations. These trends and efforts vary widely in the types of problems they are attempting to address. Some are extending the current metaphor or paradigm while others try to visualize more futuristic methods of integrating computers into our lives. Our research suggests that current work and technologies can be categorized in three ways. These categories include Wireless, Nomadic, and Ubiquitous Computing.

Wireless, Mobile and Nomadic computing - Wireless or Mobile computing endeavors to extend the reach of computing to mobile computers and other hand-held devices such as hand-held computers or PDAs, notebooks and mobile phones. This is being accomplished, in part, by designing new hardware, software, protocols and networks. Nomadic computing is similar to mobile or wireless computing. In Nomadic computing a user "wanders" across or through different networks while maintaining their user profile and environment.

Ubiquitous computing - A user has access to many computing devices. These computing devices will be unobtrusive and provide seamless access to a wide variety of data and allow the user to perform tasks as needed, where needed. The objective of ubiquitous computing is to move interaction with computers out of a personís central focus and into the userís peripheral attention where they can be used subconsciously. Ubiquitous computing is often characterized by the attributes of mobility, interconnectivity and context-awareness.

This is a dynamic document and will be amended as new efforts and technologies are discovered. As always, comments and suggestions are welcomed and solicited.

2. Wireless, Mobile and Nomadic

2.1. Background

People are becoming more and more mobile and are requiring access to various systems, data, and software. As a result, the wireless market is growing rapidly. Analysts anticipate that by the year 2005 as much as 50 percent of all communication terminals will be mobile [1]. Individuals want to take the same tools and data they have at their desks wherever they may go. In addition, they want to access a variety of information from wherever they or the data are located. Access to tools and data must be easy, transparent, reliable and secure.

When one is "mobile" and using a computer to perform various tasks and access data, connecting to remote sites should be simple and as transparent as possible. No special tasks should have to be performed. No special environments need be in place. Using notebooks and PDAs (with Windows CE) today, it is very easy to remain connected to any number of systems while in remote locations. Generally, all that is needed is a modem and you're connected. Mobile or nomadic systems present special problems. As a connected mobile host moves from network to network issues such as mobile IP [2] , caching, changing resources and possible disconnect time must be addressed [3] .

Reliability and security are also important and necessary for mobile computing to be useful. Connections should be solid and trouble free. The mobile user must have reliable access from anywhere at any time. While connected to systems and retrieving or transmitting data, security is at times extremely important. Secure connections may not be necessary when browsing the news on MSNBC, for example, but are desirable when connected to corporate databases or when electronic commerce is undertaken.

Mobile computing is currently in a state of transition. There are competing technologies and standards. With the help of the European Union, we are seeing greater cooperation among hardware / software developers, manufacturers and services providers [4] .  In the U.S. we are experiencing similar benefits as systems like CDPD [5] over CDMA are developed and introduced.  A consensus on a digital telephony standard has still not been reached between world markets and North America. In fact, there is no consensus even among North American vendors. This may change with the introduction of systems such as PCS 1900 and future generations of mobile networks such as UMTS (Universal Mobile Telecommunications Standard) or the International Mobile Telecommunications 2000 (IMT2000).

2.2. Technology

There are a wide variety of technologies that are being researched, developed and deployed for wireless and mobile computing. There are essentially two groups (three if you consider the Japanese PHS network) developing mobile networks. These are the systems being developed by the European Union and those being developed by groups in North America.

Existing digital technologies were developed to enhance voice services of AMPS by providing the ability to send faxes, emails, connect to the internet, to provide short message service and other digital data features. In North America, we find companies that are deploying CDMA, TDMA (and variants), PCS-1800 and PCS-1900 networks. In Europe and other parts of the world GSM is being deployed as the digital standard for networks.  For an overview of GSM, the reader is referred to [6] and to [7] for market information.

North American networks and European networks may be heading for convergence with the approval of the next generation GSM standard. This new proposal combined the work of two formerly opposing groups by incorporating main elements and benefits of Wideband-CDMA (W-CDMA) and elements of a hybrid technology referred to as Time Division - Code Division Multiple Access (TD-CDMA).

Other technologies are in use for mobile networks these include:

  • Packet Data - A wireless packet switching technology. Providing data-only capability. There are three competing standards including Mobitex from Ericsson, DataTAC from Motorola, and Messenger from Racal.
  • Satellite - Examples include Qualcomm's OmniTRACS, Alcatel/Qualcomm's EutelTRACS, Inmarsat-C, and Motorola's Iridium,
  • Mobile Radio - This technology is used primarily for dispatch in emergency services, transportation and utilities. Examples include the Specialized Mobile Radio (SMR) found in the U.S. and Public Access Mobile Radio (PAMR) found in Europe and Asia.
  • PHS (Personal Handyphone System) - Developed in Japan. This technology uses cordless phones in congested city environments where cellular phones experience shielding problems.
There is a desire to introduce new multimedia functionality into the digital network. Therefore, it is desirable to extend the present applications of voice and data transmission beyond the 9.6kbit/s rate. The development of High Speed Circuit-Switched Data (HSCSD) will help when deployed in 1999.

Part and parcel of the desire to introduce multimedia functions to mobile computing is the concept of the mobile Internet. Efforts are underway to merge Internet (IP) technologies into wireless services. Two such efforts are General Packet Radio Service (GPRS) and Cellular Digital Packet Data (CDPD).

General Packet Radio Services (GPRS) - GSM will add this new transmission mechanism for use in high-bandwidth data access applications. It will be particularly suited to E-mail and database access services. GPRS will enable users to connect to IP networks (TCP/IP) such as corporate LANs and the Internet. In addition, users can have messages delivered to their phones without a normal full end to end connection.  Since GPRS is TDMA based we can expect data rates ranging from 14.4Kbps (using a single slot) to 115Kbps or more (using all time slots).

Cellular Digital Packet Data (CDPD) - CDPD uses the AMPS media to provide packet data networking services to mobile hosts. CDPD is a connectionless network service supporting multiple protocols (IP, CLNP, and IPv6 [8] ). Like GPRS, CDPD provides a wireless network extension to current data networks. CDPD applications have been implemented in various domains, These domains include financial (Mercantile Bank, Empress II), mobile office (American Airlines, Boston Edison, Walt Disney Resorts), public safety (various police departments), telemetry (Columbia Gas, Clawson Concrete), and transportation (AAA, Salvation Army, United Airlines).

Mobile computing presents special problems not found in networks comprised solely of static hosts. Some of these problems are being addressed by the technologies mentioned above. Past assumptions about static hosts no longer hold and this requires that aspects of software, networking, and hardware devices be modified.

Software must recognize whether it is running on a remote or local host since resources are likely to be different [9,10] . Resource constraints may include display resources, power consumption, memory, disk space, input devices, caches, and variable connect times. TTML and WML [11] are examples of current new enhancements that have been made in software to support new features on digital phones. Current billing plans use both connection time (circuit-switched public networks like PSTN and GSM) and volume (packet-switched networks like CDPD) as a basis for billing for data transmitted. A new billing philosophy needs to be thought out as well. It is likely that a volume-based scheme will be used versus the current time based scheme. This makes particular sense when we are considering a mobile computing model where it is assumed that a user's device is always on and always connected. A volume based billing scheme may also make more sense for a business that intends to use a push model for data subscription and advertising. In addition, for truly nomadic computing to take place, software will have to ensure that user environments are consistent across networks.

Networks must be adapted to the concept of mobile hosts. Some form of support for mobile IP addresses must be devised. Mobile hosts may use agents (home and remote) to provide tunneling (encapsulation) solutions [12] to problems resulting from increased mobility. Mobile networks will have to maintain a consistent Quality of Service (QoS) across QoS parameters such as packet delay, packet loss rate, delay jitter, and throughput. In short, mobile networks will have to provide seamless and transparent access to mobile hosts. The deployment of new network types such as GPRS and CDPD may help resolve these issues.

Lastly, hardware devices will need to be modified. Changes will have to be made to mobile stations (hosts) to support new network technologies and architectures; e.g. you will need a new GPRS telephone. Devices will have to be designed for better power management. Displays may need improvement as well.

Mobility will provide benefits that will be reflected in increased productivity and reducing errors by sharing up to date data for example. These and other benefits will be realized but not without cost.

There is a plethora of ongoing work in the areas of mobile, wireless and nomadic computing. Rather than describe each in detail, pointers to numerous papers and sites are provided to the reader. This is particularly useful since the technology and status of this work is likely to change frequently.

See Appendix A for links.

3. Ubiquitous Computing

3.1. Background

Ubiquitous computing is where we'd like to go - but it's the hardest place to get to. Reaching the technology of ubiquitous computing requires that researchers and developers think in new ways. Much of the technology (networks, hardware, software) needed to implement truly ubiquitous computing does not yet exist. Aside from one or two small implementations of ubiquitous systems, there seems to be mainly small research projects being done. There have been other systems that integrate various devices such as alarms and household appliances with computers. Smart-Home is an example. In such systems, the devices themselves are not "intelligent" but are controlled by a computer. Ubiquitous computing can be characterized by several key attributes. Important ubiquitous computing attributes include mobility, interconnectivity, and context awareness [13,14] . When combined, these attributes should provide an environment where interacting with various devices becomes transparent or in the periphery of our awareness.

One of the key attributes of ubiquitous computing is mobility. An individual will carry or wear multiple devices and move about the office, home or public places. It's obvious that these devices should be both easily transportable and simple to interact with. User interfaces, whether voice, pen based or key based, will present particular challenges in mobile devices.  Frugality will be important when mobile devices are designed since resources of mobile devices are minimal [15] . Displays of mobile devices have limited graphics capabilities and are quite small. Similarly, memory is limited, disks may be nonexistent, and software availability may be restricted.

Another key attribute of ubiquitous computing is interconnectivity. Current mobile hosts such as notebooks, PDAs and cellular phones [16] provide connectivity but lack interconnectivity. With these devices a user may connect to another system on a point to point basis. The ubiquitous computing paradigm will take this a step further. "Ubiquitous" devices, like existing devices, will have the ability to connect to systems on a point to point basis. They will have additional capabilities too. Devices will be aware of each other and know how to exchange information among themselves. In addition, they will be able to control each other as needed. Just how this will be accomplished remains to be seen. Suggestions including the use of infrared (IrDA), low power RF, or even inductive (EMF) communications have been proffered.

For Ubiquitous devices to be truly useful they must support the concept of context awareness. Context aware devices will be able to adapt their behavior to the environment in which they find themselves. Devices should recognize when they transition to a new network or one that uses a different protocol and adapt appropriately; e.g. switch from using low power RF for communication to using an IrDA port instead. Context aware devices will also recognize the software, hardware and other resource constraints of devices they are interacting with. For example, a device should know or be able to discover whether a device it wishes to interact with can support a certain protocol or windowing system. This is often referred to as resource qualification.

Realizing the goal of Ubiquitous computing and creating devices and systems that incorporate the attributes of mobility, interconnectivity, and context awareness is still in the future. The new paradigms and technologies (MBone, WaveLAN, GPRS, CDPD, etc.) that are emerging will eventually enable ubiquitous systems to be designed, implemented, and widely deployed.

3.2. Technology and Implementation

There is a benchmark early ubiquitous system that is frequently held up as an example and has been deployed in localized settings. This is the ParcTab project [17] at Xerox PARC. A second project, the Active Badges Project done at Olivetti, does address some of the issues raised in ubiquitous computing. It's worth a discussion of each to point out what approaches were used and what was learned.

The ParcTab project is the brainchild of Mark Weiser, the godfather of Ubiquitous Computing. ParcTab's purpose was to highlight design and implementation issues involved in building mobile systems for office settings. Since it is difficult to predict the size and types of devices that will exist in ten years or more, the researchers at PARC decided to include support for three sizes of devices. They speculate that "inch" sized devices will be attached to clothing or be handheld, "foot" sized notepads will be scattered around the office for all to use, and "yard" sized stationary whiteboard devices will be placed in offices and conference rooms. Experimental devices in each of these sizes were constructed at PARC. It is interesting to note that each device uses different communication mechanisms. The ParcTab uses infrared, the ParcPad uses near-field radio, and the Liveboard is connected by ethernet.

The ParcTab system is of particular interest because it addresses some of the more difficult Ubiquitous computing issues (size, display, power management, interconnectivity, software, and networking and cost). ParcTab is comprised of two major components - the mobile device and the networking infrastructure.

Design goals of the mobile device include small size (PDA like), useable display, ease of use, low power consumption, interconnectivity (context awareness), useful and intuitive software, and low cost of manufacture. The display of the Tab uses of a commercially available touch-sensitive screen (128 x 64 resolution). Control of the Tab is easily accomplished with two hands with one hand (ambidextrous) use possible. Power consumption was kept to a minimum by selecting a combination of processors that support low power modes and using infrared for communications. ParcTab used 850 nm IR for communications. Benefits of IR for this application include low power consumption, compactness, and commercial availability. It is also easier to design a cellular system that reduces interference between devices. As previously mentioned, software for a small mobile device should be easy to use (intuitive) and exhibit low latency between users and the system. The ParcTab offered a number of applications and input mechanisms. Applications were similar to ones that you find on the PDAs of today: calendars, email, paging, drawing, clocks, text processing, games and others. TCL/TK (Hypercard for Macs) was used for application development. The user interface incorporated a text-based display that includes icons, scrolling lists, pen based keyboard entry as well as unistroke hand recognition. Researchers found that, of the applications provided, applications for e-mail, weather, and file browsing (context sensitive) were among the most popular. Other popular applications included Web browsing, locator and paging, computer aided collaboration such as voting, remote control, and X10 remote control.

The network infrastructure for the ParcTab is a multi-layered architecture that consists of Tabs, IR transceivers [18] , IR gateways, Tab Agents and applications running on a UNIX workstation (SparcStation). ParcTabs are event driven and execute local functions in response to remote procedure calls that allow applications to control various resources. In addition to responding events ParcTabs can also generate events that are transmitted to IR transceivers and IR gateways that control them to UNIX processes called Tab Agents running on network machines. Tab Agents are responsible performing several tasks within the system. Agents are tasked with forwarding packets from applications to ParcTabs, forwarding messages from the Tab to an application, providing location information to applications for context awareness, and for starting the shell (interface used for managing applications). The overall architecture seemed to perform well.

In summary, the system was successful. The PARC system provided users with useful applications, reliable communication, and interconnectivity. Researchers learned that setting up the network was easy. They also learned that there were problems in highly utilized IR networks including corrupted packets, dropped packets caused by buffer overflow in the transceiver, and the interaction between corrupted packets and retransmission of those packets (led to increased traffic). The folks at PARC also found that the physical placement of IR transceivers was important. They experienced problems with transceivers located close to the ballast of fluorescent lights, transceivers located near direct sunlight, and obstructions. Similar care was required when positioning transceivers near doorways or windows in order to avoid interference with transceivers in other cells. It is interesting to note that user acceptance of the ParcTab system was based on factors that included size, applications supported, convenience and the look and feel of the device. Users also expressed the desire to use the tab device outside of the internal network. This would seem to imply that a secondary method of connectivity is needed; e.g. adding wireless capabilities. In time, we can expect more power efficient devices, better displays, smaller components and voice based systems. As these technologies evolve, many of the limitations in implementing Ubiquitous Systems will be removed.

Another system worth taking a quick look at is the Active Badges System [19] . Active Badges is a system that aids in locating a person within a building by finding the location of their Active Badge. Each individual is given a device that transmits an infrared signal at periodic intervals. This signal is detected by one of many networked sensors. Signals can then be mapped into the location of the badge within the office space. Two-way bi-directional communications is supported. In addition, each badge has a unique 48-bit address for identification. Olivetti indicates that there are about 1500 badges and 2000 sensors deployed across a number of European universities. Systems have also been used at Xerox PARC, DEC's research laboratory, MIT's Media Lab, and Bellcore.  Active Badges may be used in applications such as access control, environmental control, directing phone service, and others. The benefit of using Active Badges in these ways is that the user is required to take no active role to use the badge and to perform tasks. Olivetti is also researching a small desk top, low power RF version of Active Badges that will be used for inventory control.

There has been much thought given to the idea of ubiquitous computing but few systems developed. This is due in large part to the current state of technology. As technology progresses we will see more systems like the two already discussed and may yet see complex Ubiquitous computing systems.

See Appendix B for links to papers and sites discussing Ubiquitous Computing.

4. Additional technologies

4.1. Infrared

In 1993, the Infrared Data Association (IrDA) was formed to devise a standard [20] for low cost, low power, and interoperable infrared devices that supported a point-to-point model and provided for efficient and reliable data transfer. The standard required that IR devices be viable for use in a wide variety of computing and communication devices and appliances. In a very short period of time, better than 75 worldwide members, consisting of users and manufacturers, devised standard IrDA 1.0 (SIR). IrDA 1.0 supports data rates of 115Kb/s. Less than a year later, the standard extension IrDA 1.1 (FIR) was specified and supports a data rate of 4 MB/s. By 1995, IrDA compliant devices were in IR equipped PDAs, notebooks, printers and adapters. This is an impressive accomplishment.

A key component of the IrDA specification is the simplicity of the hardware. IR hardware includes the encoder/decoder, which interfaces with the UART, and the IR transceiver. The transceiver is itself comprised of the output driver, the IR emitter and the receiver detector. The hardware is produced by a variety of vendors, is inexpensive and has low power requirements. In addition, the IrDA standard defines requirements of the physical layer. Physical layer specifications include IR with a wavelength of 0.85 to 0.90 micrometers and a transmitter intensity of a minimum of 40 mw/Sr and a maximum of 500 mw/Sr within a 30-degree cone. Parameters for required minimum transfer rate, formats for data, error detection are also specified.

Prior to IrDA 1.0, implementations of IR used proprietary protocols. Interoperability was difficult if not non-existent. New implementations of IrDA compliant equipment resolves this issue and ensures that IR will work for all applications and hardware platforms produced by manufacturers. IrDA has defined a mandatory link protocol [21] called the Infrared Link Access Protocol (IrLAP). IrLAP is derived from an existing asynchronous data communication standard, namely HDLC. Like HDLC, a data link has at least two IR stations called the Primary or commanding station and the Secondary or responding station. The Primary station is responsible for the all transmission over the data link and can support point-to point or point-to-multipoint communication. IrLAP also uses many of the framing types used in HDLC. Frame types are U (unnumbered) frames, S (supervisory) frames, and I (information) frames. Similarly, IrLAP supports procedures found in HDLC such as link initialization and shutdown, connection startup, disconnection, and information exchange. Device address conflict resolution and device address discovery is supported as well.

The Infrared Link Management Protocol (IrLMP) is another mandatory element of the IrDA specification and helps establish connections between devices. IrLMP is comprised of two components. Theses are the Link Management Information Access Service (LM-IAS) and the Link Management Multiplexor (LM-MUX). Essentially the LM-ISA maintains an information base containing available services of the station. This information can then be presented to inquiring stations. The LM-MUX component of IrLMP provides for multiple data link connections over the single connection provided by IrLAP. Multiple connections can be supported since each IR station is free to define multiple Link Service Access Points (LSAP). If the LM-MUX is in the exclusive mode, only one LSAP connection may be active. In this instance, flow control is provided by IrLAP. If the LM-MUX is in multiplexed mode, many LSAP connections can occur on the base IrLAP connection. Here, flow control is provided by the application or upper layer protocols. It is important to note that although IrLMP is mandatory, not all aspects of it need be implemented.

Finally, additional optional protocols have been specified. The optional protocols, IrTP and TinyTP, provide flow control functions for a LM-MUX when it is in the multiplexed mode. IrTP and TinyTP are also used to segment or reassemble data when matching buffer size and data size is needed. Another example of an optional protocol provided by IrDA specifications is IrCOMM. IrCOMM is used to emulate existing serial and parallel ports and supports four service types. Additional optional protocols such as Obex and Plug-and-Play are included in the IrDA specifications and design.

The value of the IrDA specification and its flexibility are evident in the number of devices in which infrared is now deployed. Infrared is currently being used in devices that include cellular phones, printers, pagers, PDAs, digital cameras, and medical devices.  Applications have been written that use infrared technology to transfer data, files, graphics, audio and video. Even applications for infrared LAN connectivity have been created. In summary, infrared technology as defined by IrDA presents new opportunities to a wide variety of equipment manufacturers. Infrared can be used to connect existing infrastructures such as ATM, Internet and LANs. Infrared also facilitates interconnectivity of many devices and leverages existing investments in hardware and software.

More and more devices are supporting the IrDA specification as time goes on. IR ports are common place on notebooks, cellular phones, printers, etc. Just recently, IrDA announced that it has released a draft IR connector standard. This standard will enable new IR connectivity between desktop computers and devices such as cellular phones, notebooks, PDAs and other peripheral devices. In addition, two other notable standard initiatives have been undertaken. The reader is probably familiar with the lack of standards for infrared remotes for consumer appliances (multiple remotes for TVs, VCRs and stereos). In February of 1996, a consumer IR charter meeting took place and established the Consumer Infrared Bi-Directional Standards Initiative. This initiative will bring consumer infrared products into a common standard. Similarly, in January 1997, a Telephony Charter Meeting was held to draft a proposal for standards to enable various wireless communication devices to communicate and transfer data using bi-directional infrared. These initiatives will enhance connectivity among many types of devices and create new opportunities for developing new networks and software applications.

See Appendix C for links to IrDA resources.

4.2. Low Power RF

Using RF for Wireless LANs is a developing technology. Wireless RF LANs [22] use the ISM (Industrial, Scientific, and Medical) bands in frequency ranges that include 900 MHz, 2.4 GHz, and 5.7-5.8 GHz and the Data Termination Services bands that include 18 GHz and 19 GHz frequencies. The ISM band requires no licensing while the DTS bands do. ISM bands are susceptible to noise or interference. Interference can be generated in ISM bands by everyday appliances that use these bands; e.g. microwave ovens. However, these undesirable effects can be mitigated with Spread Spectrum technology. Spread spectrum technology is either of two types: Frequency Hopping Spread Spectrum or Direct-Sequence Spread Spectrum. ISM frequencies can penetrate walls and building structures and can be used to create larger networks. DTS frequencies, on the other hand, do not penetrate building structures very well and work over short ranges - making them suitable to creating smaller localized networks [23] . Standards have been proposed to insure interoperability. The IEEE 802.11 specification is one such standard and specifies the protocol layers for Media Access Control (MAC) and the physical layer (PHY). An additional standard, Internet Access Point Protocol (IAPP), is being proposed by companies including Lucent Technologies and will support OSI upper layers. An example of an existing wireless network is Lucent's WaveLAN [24] . Similar efforts are underway by a group of researchers at the Intel Corporation.

Researchers at Intel are proposing a usage model of networking for notebooks that is enabled by ubiquitous wireless data infrastructures. They propose [25] an infrastructure that appears to be modeled after the infrastructure proposed and specified by the IrDA organization for infrared networking. Their model is referred to "Always On, Always Connected" (AOAC) and specifies APIs similar to those in the IrDA specifications. In order to provide transport layer services for their narrowband messaging network, Intel entered into a joint effort with Nokia and developed Narrow Band Sockets (NBS). Narrow Band Sockets provide both a simple connection orientated protocol (SCP) and a simple connectionless protocol (SDP). An added benefit to NBS is the ability to implement application level addressing using standard APIs.

Other RF technologies may be used to create wireless networks. Cordless phone technologies that use the 46-49 MHz and the 900 MHz range are an example. Over time, we will see more of these localized wireless networks being developed and deployed in homes, offices and hospitals [26] . They will provide an integrated information, control, and data gathering mobile computing environment in many domains.

See Appendix D for links to low power RF resources.

4.3. EMF Transceivers and Connectivity (Aether Wire)

Aether Wire and Location Incorporated is a company that worked on a project for ARPA called "Low-power, Miniature, Distributed Position Location and Communication Devices Using Ultra-Wideband Nonsinusoidal Communication Technology" [27] . The interesting aspect of this technology is that it uses electromagnetic impulses to transmit and receive information with relative bandwidths approaching 100 percent. Applications using electromagnetic radiation are not new. In fact, electromagnetic radiation (nonsinusoidal) has been used in both ground probing radar and stealth technologies.  It is anticipated that devices using electromagnetic radiation will be very small and have low power consumption. Devices of this type seem particularly suited to being imbedded in a variety of hardware including cell phones. One can envision using these low power transceivers in commercial applications such as personal location, Inventory and in smart homes where contain context aware appliances may exist.  For details the reader is referred to the report "Low-power, Miniature, Distributed Position Location and Communication Devices Using Ultra-Wideband Nonsinusoidal Communication Technology" found in the original ARPA report.

4.4. X10 networks and the Smart Home

X10 networks are a good example of integrating computers and computer controlled devices into the home. X10 is a communications protocol for the remote control of electrical devices. The X10 network of a home consists of X10 transmitters and X10 receivers. Overall control and system programming may be performed using a PC. The benefit of an X10 network is that it uses the homes electrical system for the network infrastructure. Installation of the network is easy - simply plug the transmitters and receivers into wall sockets.

An X10 controller issues commands to one or more receivers. A controller may tell a lamp to turn on, a thermostat to change settings (HVAC applications), turn on sprinklers or even place a call to emergency service (security applications). Transmitters can also send commands upon detecting certain conditions (unauthorized entry, freezing temperatures, flooding) and even be remotely commanded from a telephone. Since PCs can be used to program controllers or be used as a controller, overall system configuration and management is quite easy.

X10 receivers accept commands from controllers but do not issue any commands. Each receiver has a unique identifier code and responds only commands directed to its identifier. Up to 256 codes can be used in a single home. Receivers may be given the same identifier for instances when it is desirable to have multiple events occur on a single command; e.g. to turn on four lamps.  The most common use of a receiver is turn on an appliance or to respond to an event such as notification from a motion detector or temperature gauge.

There are many devices that can be used to create an integrated and controlled home. There are several companies offering X10 devices. The reader is encouraged to look at the Smart-Home link listed in Appendix D. It is likely that, in the near future, we will to see many more devices and appliances that may be integrated into a X10 home network right out of the box. This will present new opportunities for moving toward Ubiquitous computing in the home and office.

See Appendix E for links to X10 resources.

5. Conclusions

We have attempted to present an overview of both existing and emerging technologies for wireless/mobile computing and ubiquitous computing. There is a great deal of work being pursued in these areas. For the first time, various industry leaders have agreed (to agree) upon future standards and implementations of technology in the areas of wireless systems, interface languages, and protocols. In the short term, disparate technologies will pose interesting problems when developing new hardware, software and for system integration.

Ubiquitous computing will happen gradually. Ubiquitous computing needs ideas, networks and hardware that do not yet exist. The researchers at Xerox PARC have made great headway by designing and implementing their ParcTab (Pad and LiveBoard) system(s). ParcTab pioneered infrared technologies, context aware applications and devices, interface issues and a variety of hardware issues. Similarly, significant efforts and contributions were made in ubiquitous computing by the researchers at Olivetti and their the Active Badges project. The work continues as others make contributions to the field by exploring technologies like X10 and AetherWire.

Additionally, overviews of technological developments if the areas of infrared and X10 and others were presented.

A great deal of information is contained in the links referenced in Appendices A through E. The links contain everything from tutorials on protocols to examples of the various technologies. The reader is encouraged to explore them. Once again, this document will be modified as new technologies arrive and more information is found on the ones described in this paper. Please feel free to forward your comments, insights, and useful information to us.


[1] J.C.S. Cheung, M.A. Beach and J.P. McGeehan, "Network planning for third-generation mobile radio systems", IEEE Communications Magizine, November 1994.

[2] Internet Engineering Task Force, "IP Mobility Support", draft-ietf-mobileip-protocol-08.txt, 4 January, 1995. Available online at

[3] G. Korteum, S. Fickas, Z. Segall, "On-Demand Delivery of Software in Mobile Environments", Department of Computer Science, University of Oregon, NSF grant MIP 9403573 .

[4] European Commission, Communication to the European Parliament. The Council, The Economic And Social Committee and The Committee of The Regions On The Further Development of Mobile and Wireless Communications, "Challenges and choices for the European Union", COM (97) 217 final, Brussels, 29.05.97. Available online

[5] CDPD Forum, Inc., "Wireless Mobile Data Networking The CDPD Approach". Available online at

[6] John Scourias, "Overview of the Global System for Mobile Communications". Available online at

[7] "GSM Markets, Users, Case Studies". Available online at

[8] B.R. Bandrinath, Anup Talukdar, "IPv6 + MOBILE-IP + MRSVP = Internet Cellular Phone?", Rutgers University, New Brunswick, NJ, DARPA DAAH04-95-1-0595, NSF CCR95-09620 and IRIS 95-09816 . Available online at

[9] Tomasz Imielinski, B.R. Badrinath, "Mobile Wireless Computing: Solutions and Challenges in Data Management", Department of Computer Science, Wireless Information Network Laboratory, Rutgers University, New Brunswick, NJ.DCS-TR-296/ WINLAB-TR-49. Available online at

[10] G. Welling, B.R. Bandrinath, "A Framework for Environment Aware Mobile Applications", Department of Computer Science, Wireless Information Network Laboratory, Rutgers University, New Brunswick, NJ. Available online at

[11] WAP Architecture, "Wireless Application Protocol Architecture Specification", Draft Version 0.9 (1997-09),

[12] B. Lancki, A. Dixit, V. Gupta, "Mobile-IP: Supporting Transparent Host Migrationon the Internet", Linux journal, August 1996.

[13] Bill Schilit, Norman Adams, Roy Want, "Context-Aware Computing Applications", Workshop on Mobile Computing Systems and Applications, Pages 85-90, Santa Cruz, CA, December 1994. IEEE Computer Society. Available onlin at

[14] Geoffrey M. Voelker, Brian N. Bershad, "Mobisaic: An Information System for a Mobile Wireless Computing Environment", Department of Computer Science and Engineering, University of Washington, September 19, 1994. Available online at

[15] B. Schilit, F. Douglas, D. Kristol, P. Krzyzanowski, J. Sienicki, J. Trotter,
"TeleWeb: Loosely Connected Access to the World Wide Web",
Fifth International World Wide Web Conference, May 6-10, 1996, Paris, France. available online at

[16] Ora Lassila, "Mobile Communication Services of the Next Century: The Internet Approach", International Business Communications, Mobile Internet, February 24-26, 1997, Orlando, Florida.

[17] R. Want, B. Schilit, N. Adams, R. Gold, K. Peterson, D Goldberg, J. Ellis, M. Weiser, "The ParTab Ubiquitous Computing Experiment", Xerox PARC, ARPA project DABT63-91-C-0027. Available online at .

[18] N. Adams, R. Gold, et al, "An Infrared Network for Mobile Computing", Xerox PARC, ARPA project DABT63-91-C-0027. Available online at

[19] The Olivetti and Oracle Research Laboratory, "The Olivetti Active Badge System", Available online at see also and

[20] IrDA Serial Infrared (SIR) Physical Layer Specification Version 1.1g, April 29, 1996, and  IrDA Serial Infrared Link Access Protocol Version 1.1, June 16, 1996. Available online at

[21] IrDA Link Management Protocol Version 1.1, January 23, 1996, IrDA TinyTP: A Flow-Control Mechanism for use with IrLMP Version 1.1, October 20, 1996, IrDA Plug and Play Extensions to Link Management Protocol Version 1.0, September 30, 1994. Available online at

[22] Phong Quang Ta, Dimitrios G. Soulios, "Wireless LANs", Telecommunications Network Project, April 15, 1997. Available online at

[23] National Information Infrastructure, "NTIA Special Publication 95-33: Survey of Rural Information Infrastructure Technologies", Section 4. Available online at

[24] Lucent Technologies, "Systimax SCS WaweLAN Wireless Solution". Available online at

[25] Michael M. Tso, Daniel J. Gillespie, David A. Romerell, "Always On, Always Connected Mobile Computing", IEEE ICUPC, Session 11D - Emerging Technologies, 1996.

[26] John Mahr, "Hospitals Streamline Communication Through Wireless LAN", International Business Communications, Mobile Internet, February 24-26, 1997, Orlando, Florida.

[27] Robert Fleming, Cherie Kushner, "Low-power, Miniature, Distributed Position Location and Communication Devices Using Ultra-Wideband Nonsinusoidal Communication Technology", ARPA Semi-Annual Technical Report, July, 1995.


Appendix A: Wireless, Mobile and Nomadic Resources

Appendix E: X10 Resources

Brought to you
The Cyberpunk Project