200
Maria Luisa Damiani and Elisa Bertino
{Dis joint, T ouch, In, Contains, Equal, Cross, Overlap}. These relations are binary,
mutually exclusive (if one is true, the others are false), and they are a refinement
of the well-known set of topological relations [16]. Moreover, features have an
application-dependent semantics that is expressed through the concept of feature
type. A feature type captures the intensional meaning of the entity, for example,
Campus and Department. The extension of a feature type is a set of semantically
homogeneous features.
9.4.2 Basic Concepts
The Position Model
The notion of position is fundamental since it characterizes the mobile user. Unlike most proposals in mobile computing which describe the position uniquely
in geometric terms, for example a point, we introduce, for the sake of flexibility, the distinction between the real and the logical position. The real position
corresponds to the position of the user on Earth acquired through some
positioning technology. Real positions can thus be represented as geometric elements of different types since, depending on the chosen technology and accuracy
requirements, they may correspond to points or polygons. Conversely, the logical
position is defined at a higher level of abstraction to represent positions in a way
that is almost independent from the underlying positioning technology. Further, besides a geometry, the logical position has a semantics. For example, logical positions
can be a house, an address number, or a road, which are represented in terms of
spatial feature types. The logical position is computed from real positions by using a location mapping function. For example, a location mapping function can be
defined to map a position acquired through global positioning system (GPS) onto
the closer road segment. More formally, given a feature type f t, a position mapping
function for f t is a function m f t which, given a real position rp, returns a logical
position corresponding to an instance of f t having rp as real position. As we will
see, since the localization may respond to different application requirements, for example, with respect to accuracy, the meaning of location may vary depending on the
role. For example, the position of a generic campus member may be coarsely defined
in terms of campus sectors, assuming that the campus area is subdivided in sectors,
whereas the position of the teacher can be represented at a higher resolution by an
address.
Spatial Role
The concept of spatial role is the distinguishing aspect of our model. To account for
the spatial context, a spatial role is defined not only by a name as in RBAC but also
by a role extent. The extent of a role defines the boundaries of the region contained
in the reference space. A user, who has been conferred a role r, is thus recognized
to effectively play such a role only when logically located within the extent of r.
For example, CampusMember(UniMi) is a spatial role: CampusMember is the role
Maria Luisa Damiani and Elisa Bertino
{Dis joint, T ouch, In, Contains, Equal, Cross, Overlap}. These relations are binary,
mutually exclusive (if one is true, the others are false), and they are a refinement
of the well-known set of topological relations [16]. Moreover, features have an
application-dependent semantics that is expressed through the concept of feature
type. A feature type captures the intensional meaning of the entity, for example,
Campus and Department. The extension of a feature type is a set of semantically
homogeneous features.
9.4.2 Basic Concepts
The Position Model
The notion of position is fundamental since it characterizes the mobile user. Unlike most proposals in mobile computing which describe the position uniquely
in geometric terms, for example a point, we introduce, for the sake of flexibility, the distinction between the real and the logical position. The real position
corresponds to the position of the user on Earth acquired through some
positioning technology. Real positions can thus be represented as geometric elements of different types since, depending on the chosen technology and accuracy
requirements, they may correspond to points or polygons. Conversely, the logical
position is defined at a higher level of abstraction to represent positions in a way
that is almost independent from the underlying positioning technology. Further, besides a geometry, the logical position has a semantics. For example, logical positions
can be a house, an address number, or a road, which are represented in terms of
spatial feature types. The logical position is computed from real positions by using a location mapping function. For example, a location mapping function can be
defined to map a position acquired through global positioning system (GPS) onto
the closer road segment. More formally, given a feature type f t, a position mapping
function for f t is a function m f t which, given a real position rp, returns a logical
position corresponding to an instance of f t having rp as real position. As we will
see, since the localization may respond to different application requirements, for example, with respect to accuracy, the meaning of location may vary depending on the
role. For example, the position of a generic campus member may be coarsely defined
in terms of campus sectors, assuming that the campus area is subdivided in sectors,
whereas the position of the teacher can be represented at a higher resolution by an
address.
Spatial Role
The concept of spatial role is the distinguishing aspect of our model. To account for
the spatial context, a spatial role is defined not only by a name as in RBAC but also
by a role extent. The extent of a role defines the boundaries of the region contained
in the reference space. A user, who has been conferred a role r, is thus recognized
to effectively play such a role only when logically located within the extent of r.
For example, CampusMember(UniMi) is a spatial role: CampusMember is the role