Context is a very important concept in mobile computing. Generally speaking, context indicates "the words that are used with a certain word or phrase and that help to explain its meaning". Alternatively, context describes "the situation in which something happens: the [...] conditions that exist where and when something happens". The latter definition better describes the term in relation with mobile computing. According to Chen and Kotz, context is "the set of environmental states and settings that either determines an application's behavior or in which an application event occurs and is interesting to the user". Context-awareness in mobile computing leverages some form of context to modify or adapt, for instance, application behavior.

Different forms of context

Location (or equivalently, position) indicates the place where an object or person is situated. Location is expressed in terms of coordinate systems; the most useful one is the ellipsoidal coordinate system for the earth, wherein location is measured on its surface in terms of latitude and longitude. Positioning systems allow to determine location by using certain technologies and algorithms. They have two main aspects: sensing obtains a physical quantity (e.g., through radio, ultrasound, or infrared) that is leveraged to measure distance in a process called ranging; estimation involves techniques (algorithms) to derive actual location from sensed data. One example is triangulation, which takes angular measurements from known reference points, a technique widely used for land surveying. A more practical technique is trilateration, which instead takes the distance between one target and a few reference points (i.e., sources). When distance is precisely measured, only three sources are needed for accurate positioning – this follows from the geometric properties of circles and spheres. Location can be leveraged to offer different types of services. One is navigation, which allows to find our way to a certain destination with different means of transportation (e.g., by walk, car or public transport). Another is geo-fencing that triggers some action based on entering or leaving certain geographic boundaries (e.g., automatically open a shopping list when entering a grocery shop).

Proximity is the condition of being close to something or a place (i.e., a target). Clearly, proximity can be derived from location when the position of both the target and the device are known (for instance, through GPS or WiFi-based fingerprinting). However, there are also other options to detect proximity without relying on location information.

  • Radio Frequency Identification (RFID). It employs radio tags – namely, passive (i.e., batteryless) or semi-passive means of identification – as a machine-readable description of an object. The mobile device acts as a tag reader by sending a wireless signal that powers the tag so as to obtain some small amount of data from it. The Near Field Communication RFID technology has a range of a few centimeters. Reading a tag successfully implies physical proximity to it.
  • Beacons. They are active (battery-powered) devices periodically sending advertising messages through a low-power and short-range radio technology such as Bluetooth Low Energy. The mobile device listens for beacons and associates them with physical objects. Examples of beacons include Apple iBeacon and Google Eddystone. Beacons may provide coarse-grained distance information which is accurate only in very short proximity (i.e., one or two meters).
  • Bi-dimensional barcodes. They are a 2D arrangement of modules (usually black and white blocks) that encodes machine-readable data. They are scanned by using a camera; Quick Response (QR) codes are widely used in mobile computing.

Activity is broadly defined as "the condition in which things are happening or being done". Different from time and location, activity is a high-level description of state: working, running and sleeping are some possible activities. They can be detected from sensors and pattern recognition, for instance, from accelerometer readings. Activity can also be leveraged in the context of mobile computing. For instance, phone calls can be automatically rejected when a person is in a meeting or jogging.

Viimeksi muutettu: lauantaina 19. syyskuuta 2020, 23.10