For applications to become location-aware, the wireless network must acquire the current location of the device. This can be done either automatically or manually. Automatic location acquisition uses a positioning network to location a fix of a given device using technology such as:
- Global Positioning System (GPS) satellites
- Cellular base stations
Each of these automatic-positioning technologies has benefits and drawbacks: Cellular base stations are ubiquitous across most urbanized areas, but their positioning is not precise enough to accurately locate a user within FCC mandates. In contrast, GPS satellite positioning can be extremely precise (accuracy within meters), but the signal is easily obscured inside buildings, urban canyons, and forested areas. While these automatic positioning technologies will become commonplace over the next few years, it is possible to deploy location services now – using the existing wireless positioning infrastructure.
Fig. 3: Getting a Map Delivered to a Handheld Device
Instead of waiting for devices with automatic location detection (either by “triangulation” of the three nearest base stations or by built-in GPS receivers), applications can be designed to enable quick manual inputs for location acquisition. These include using:
- Landmarks (Users save locations such as home, office, and hotels, and then choose those to serve as reference points.)
- Previous locations (Applications can store the last few locations that were used.)
- Address, city or zip code (depending upon the level of granularity required)
- Cell-id boundaries (for defining and delivering LBS services)
In addition, it is possible for wireless carriers in many parts of the world to use existing cell-id boundaries for locating devices and delivering LBS services. For example, GSM phones (the dominant mobile network in much of the world) must know their cell identity if they are switched on and have radio coverage (the identity is carried by a broadcast signal from each base station). It is also possible to find out roughly how far a GSM phone is from the current base station. This is achievable because the devices share radio channels by using different allocated time slots. Therefore, they have to send their data either sooner or later, depending upon how far from the base station they are. This is called timing advance. All mobile phones must know their timing advance if they are on a call, so they know roughly how far they are from the base station (within about 500 meters).
Fig. 4: Entering ZIP Code for Yellow Pages Search
Key Technology Enablers for Mobile Location-Based Services
The performance and capability requirements expected for wireless location-based service can easily approach that of a top Internet portal – millions of queries on a daily basis, hundreds of concurrent transactions, and millisecond query response times. Thus, the required system must support all the unique CPU-intensive location queries, and provide scalability, storage, and interoperability.
Real-time, transaction-based location services have the kinds of feature and performance requirements listed in Table 1.
|
Feature Requirements
|
Performance Requirements
|
Address verification and matching
Map rendering
Yellow page directory query
Driving directions
Personalization by location
Proximity analysis
Standards-based location service APIs
Personal/in-car navigation capability
Voice
(VoiceXML) capability
XML integration with e-business apps
Web Services
Directories
|
Scalable architecture
Gigabytes to terabytes of data
Multiple CPU processing
DBMS table partitioning
Distributed processing
Native spatial data management
Online services interoperability
Millisecond location query
Million + daily queries
25,000+ user sessions per hour
Portal caching
|
Java, XML, and the use of Spatial databases have emerged as enabling technologies that provide fundamental infrastructure for the delivery of mobile location services.