Chip Yeakey

Senior Seminar

Literature Survey

Location-Aware Computing

Location-aware computing is a part of a larger research field known as Context-aware computing. “Context-aware computing is a mobile computing paradigm in which applications can discover and take advantage of contextual `information (such as user location, time of day, nearby people and devices, and user activity)”¹. Although this field has been around for many years, it has recently begun to take off due to the smaller size and cost of the hardware needed to monitor the above information in a mobile environment. In order for this field to become a part of our everyday lives, it must become low cost and to happen without people’s attention. If people have to spend extra time to enter in their location or adjust the device, it will be much less useful. In the last few years items like a GPS receiver or wireless data access have become available in single chips, making it easier to integrate them into a device as small as a PDA. Having the ability to use these types of devices allows for a large amount of different data to be used in determining the location and context of a person and his/her work environment. There have been many ideas of how to use this information, from redirecting phone calls to modifying yellow pages searching to give results nearest your location. There are seemingly endless ways to use this information and to think of applications for which this is well suited.

The most commonly used technology for finding outdoor location is GPS. It is based on the easy to use longitude and latitude standard and can be done on increasingly smaller and cheaper hardware than ever before. It is very nice because the signals used require no payment. GPS allows developers to have a person’s location to within a few meters of accuracy. Despite the benefits, GPS is not without its limitations. GPS is based on receiving a time signal from various satellites and finding the position using those signals. This requires a clear view of the sky for the device to get the signals, limiting this technology to outdoor use. The GPS data can also not tell where a person is facing unless they are moving. These can be significant hang-ups for some applications.

Another technology used to determine location is IR. This application requires a beacon to be placed in each location that is to be identified. This, and IR’s relatively short range, can make it somewhat impractical. There are also some advantages to using IR. IR can provide for some communication between the beacon and the sensor. This ability to access location and data through one source gives IR a big edge. This opens the potential for crafting more computation or storage intensive applications where the majority of the work is done on a stationary server and the portable device simply displays the result. This also allows for interaction between devices so that a person’s context can be determined by not only their physical location but also by what other people or objects are in the area.

Some other technologies that make it possible to determine location and communicate data are the 802.11b wireless standard and cell phone communication networks. 802.11b can be used to determine location with a large granularity but can be used to transfer large amounts of data. It has some limitations since it uses radio waves and is subject to the same trouble that many other radio technologies encounter. It is difficult to get a large coverage area through many obstructions (walls, wire-mesh windows, etc.) and can have considerable trouble with interference from other devices. Cell phone communication can do much of the same type of communication, although bandwidth is much lower and would require leasing of towers. Wireless and cell phone technologies offer the same advantages that data communication over IR shows but have their share of problems as well.

In practice it is more sensible to incorporate two or more of these technologies to take advantage of all benefits and overcome limitations. Many organizations are researching and trying to make these possibilities part of our everyday lives. In both academic institutions and in industry, ideas from navigation systems to automated homes are becoming a reality.

At the Hewlett-Packard laboratories, the CoolTown research program is a look into what the world could be with these technologies. Their work includes “building ubiquitous computing systems that sense physical entities in the environment and map them to a Web browser.”² This is based on devices having network access to see information and services associated with physical objects. CoolTown also uses IR beacons and “Active Badges” to keep track of the location of individual people. Each person’s badge communicates their location and gathers information accordingly. This system is accurate but would not scale well with the need to install beacons in every room of every building you wish to access. In addition, badges must be worn and maintained at all times; they do no good if they are left in the car or the batteries go dead. Some large advantages to this are that people are able to choose when and how they will use this technology. If they have no use for it on a given day, they may simply take the badge off, and go about their day. CoolTown also uses context in everything. “In cooltown, every person, place, or thing can be connected to the web. Even a wrist watch has the capability of becoming an intelligent web appliance.”³ With all the things in a home interacting and adjusting without direct human control the technology becomes an asset instead of another thing to take up a person’s time.

At Purdue University, the PLACE Project (Pervasive Location Aware Computing Environments) works with off the shelf items combined to determine location and query a server that keeps track of the devices. Using the campus’ wireless network to communicate, and GPS devices for location, they are able to “realistically observe and evaluate relevant parameters.”⁴ This includes information about where people go and what patterns they make in their daily routine. Such a system is not only useful to the user of the device, but also to the moniter of the system since they can improve the campus based on people’s trends of activity.

At MIT, Nickolaos Michalakis is working on “PAC: Location Aware Control for Pervasive Computing Environments.”⁵ This has a specific focus on security without specific identity. “In a pervasive computing environment it is not always desirable for a user to authenticate his identity in order to access services. Some applications are interested in the user’s context (such as location, orientation, etc.) rather than his identity in order to interact with him.” ⁶ This is the basis for having a way to authenticate based on a specific aspect of the user’s context, not his/her identity. This also raises issues about the system’s security and how hard it is to fake an attribute of context in relation to how difficult it is to pose as a particular individual. Michalakis chooses to use a code from a beacon in a physical location to allow a client to access a server. The server in turn verifies that it was a correct code, and gives an access ticket to the services requested. This type of research will be very valuable as systems begin being used on a wide basis and providers wish to have only authorized users accessing their resources.

This change in how we interact with computers has been a long time coming according to some minds. Five years ago in the book, “Beyond Calculation, the next 50 years of computing,” many predictions were made that are already on their way to coming true. “In the long term, visual and special image input from sonar, radar, and global position sensing (GPS) with a worldwide exact time base coupled with radio data links will open up new portability and mobility applications. These include robots, robotic vehicles, autonomous appliances, and applications where exact location of objects is required.”⁷ If many of these underlying technologies are already being used and becoming more accurate, it may not be long until these types of services and devices are a reality. GPS and wireless network technologies allow for many new possibilities. “Radio networks and GPS open up more possibilities by having objects that know where they are and can report their state and that are not just adaptations of cellular phones. Nothing - from keys to cars to people - need be lost.”⁸ This makes for an interesting possibility, GPS receivers could be small and low power enough to be installed in a key chain. Having every object able to keep track of position and act according to its and other object’s positions and contexts can cause good or bad reactions, dependant on how well they operate. “The most potentially interesting, challenging, and profound change implied by the ubiquitous computing era is a focus on calm. If computers are everywhere, they had better stay out of the way, and that means designing them so that people being shared by computers remain serene and in control.” ⁹ It will prove to be very important for devices to serve without distracting the people they are serving. This will be the basis for many people’s decisions whether these products will be worth adding to their lives.

Along with this new amount of accessibility to information about devices and people comes the issue of security. An important question is how can this access be limited to only the devices and people one wishes to allow. Information about people’s work habits and activities could potentially be exploited in many ways from a burglar learning when you are likely to be away from home to someone taking control of your appliances. It should be possible to keep security under control with correct considerations, but this remains a continuing concern.

With this wealth of information about people and how they spend their time, many companies may be looking for ways to analyze this data to find how to better serve customers and how to better sell a product. This is the first of many situations where context aware computing will cause changes and opportunities in other fields. In this case, data mining through an enormous amount of data by today’s standards will be a booming field as context aware devices begin gathering and storing large amounts of information about location as well as other attributes of peoples lives. Having ways to gather context information will also assist areas such as robotics by providing ways for robots to do things never dreamed of without human interaction. Overall, context-aware computing, specifically location-aware computing, is creating a large number of possibilities for the present and the future.

Endnotes:

1. Chen, Guanling; Kotz, David "A Survey of Context-Aware Mobile Computing Research."Dartmouth Computer Science Technical Report TR2000-381. 16 September 2002. pg.1.

2. Pradhan, Salil, "Websign: hyperlinks from a physical location to the web" Hewlett-Packard Company. 11 June 2001. pg.2.

3. "Cooltown: the ecosystem explained." http://www.cooltown.hp.com/mpulse/backissues/0601/0601-cooltown.asp?print=yes. 16 September 2002. pg.1.

4. Hambrusch, Susanne E. "Pervasive Location Aware Computing Environments." http://www.cs.purdue.edu/homes/seh/PLACE.html. 23 September 2002. pg.1.

5. Michalakis, Nikolaos "PAC: Location Aware Access Control for Pervasive Computing Environments," MIT Laboratory of Computer Science, http://www.org.lcs.mit.edu/pubs/michalakis.pdf. 16 September 2002. pg.1.

6. Michalakis et al. pg.1.

7. Denning, Peter J., Metcalfe, Robert M., Beyond Calculation, The next 50 years of computing, Ch 1,2,6, Copernicus New York, New York. 1997. pg.15.

8. Denning et al. pg.29.

9. Denning et al. pg.79.
Bibliography:

Chen, Guanling; Kotz, David "A Survey of Context-Aware Mobile Computing Research."Dartmouth Computer Science Technical Report TR2000-381. 16 September 2002.

"Cooltown: the ecosystem explained." http://www.cooltown.hp.com/mpulse/backissues/0601/0601-cooltown.asp?print=yes. 16 September 2002.

Denning, Peter J., Metcalfe, Robert M.; Beyond Calculation, The next 50 years of computing. Ch 1,2,6, Copernicus New York, New York. 1997.

Hambrusch, Susanne E. "Pervasive Location Aware Computing Environments." http://www.cs.purdue.edu/homes/seh/PLACE.html. 23 September 2002.

Michalakis, Nikolaos "PAC: Location Aware Access Control for Pervasive Computing Environments." MIT Laboratory of Computer Science. http://www.org.lcs.mit.edu/pubs/michalakis.pdf. 16 September 2002.

Pradhan, Salil, "Websign: hyperlinks from a physical location to the web" Hewlett-Packard Company. 11 June 2001.