Trust-based Service Management for Social Internet
of Things
Systems
Abstract:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted “things” that can provide services needed when they come into
contact with each other opportunistically. We propose and analyze the design
notion of adaptive trust management for social IoT systems in which
social relationships evolve dynamically among the owners of IoT devices. We
reveal the design tradeoff between trust convergence vs. trust fluctuation in
our adaptive trust management protocol design. With our adaptive trust
management protocol, a social IoT application can adaptively choose the best
trust parameter settings in response to changing IoT social conditions such
that not only trust assessment is accurate but also the application performance
is maximized. We propose a table-lookup method to apply the analysis results
dynamically and demonstrate the feasibility of our proposed adaptive trust
management scheme with two real-world social IoT service composition
applications.
Architecture Diagram:
Existing System:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer (P2P) networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted things that can provide services needed when they come into
contact with each other opportunistically in both the physical world and
cyberspace. It is envisioned that the future social IoT will connect a great
amount of smart objects in the physical world, including radio frequency
identification (RFID) tags, sensors [40], actuators, PDAs, and smart phones, as
well as virtual objects in cyberspace such as data and virtual desktops on the
cloud.
Disadvantages:
The emerging
paradigm of the social Internet of Things (IoT) has attracted a wide variety of
applications running on top of it, including e-health, smart-home, smart-city,
and smart-community .
Proposed System:
A social IoT applications are likely oriented
toward a service oriented architecture where each thing plays the role of
either a service provider or a service requester, or both, according to the
rules set by the owners. Unlike a traditional service-oriented P2P network,
social networking and social relationship play an important role in a social
IoT, since things (real or virtual) are essentially operated by and work for
humans. Therefore, social relationships among the users/owners must be taken
into account during the design phase of social IoT applications. A social IoT
system thus can be viewed as a P2P owner-centric community with devices (owned
by humans) requesting and providing services on behalf of the owners. IoT
devices establish social relationships autonomously with other devices based on
social rules set by their owners, and interact with each other
opportunistically as they come into contact.
Advantages:
1. To best satisfy the service requester
and maximize application performance, it is crucial to evaluate the
trustworthiness of service providers in social IoT environments.
2.
The motivation of providing a
trust management system for a social IoT system is clear: There are misbehaving
owners and consequently misbehaving devices that may perform discriminatory
attacks based on their social relationships with others for their own gain at
the expense of other IoT devices which provide similar services.
Implementation modules:
1.
User-Centric Social IoT Environments
2.
Adaptive Trust Management
3.
Trust Composition
4.
Protocol performance evaluation
User-Centric
Social IoT Environments:
We consider a
user-centric social IoT environment with no centralized trusted authority. Each
IoT device has its unique identity which can be achieved through standard
techniques such as PKI. A device communicates with other devices through the
overlay social network protocols, or the underlying standard communication
network protocols (wired or wireless). Every device has an owner who could have
many devices. Social relationships between owners is translated into social
relationships between IoT devices as follows: Each owner has a list of friends
(i.e., other owners), representing its social relationships. This friendship
list varies dynamically as an owner makes or denies other owners as friends. If
the owners of two nodes are friends, then it is likely they will be cooperative
with each other. A device may be carried or operated by its owner in certain
community-interest environments (e.g., work vs. home or a social club). Nodes
belonging to a similar set of communities likely share similar interests or
capabilities. Our social IoT model is based on social relationships among
humans who are owners of IoT devices. We note that the device-to-device
autonomous social relationship is also a potential for the social IoT paradigm.
Adaptive
Trust Management:
A design
parameter is one that adaptive trust management can control to optimize
performance. A derived parameter is one that is generated during runtime as a
result of running the trust protocol. An input parameter is one that the
operating environment dictates. addresses all aspects of trust management: the
trust composition component addresses the issue of how to select
multiple trust properties according to social IoT application requirements. The
trust propagation and aggregation component addresses the issue
of how to disseminate and combine trust information such that the trust
assessment converges and is accurate. The trust formation component
addresses the issue of how to form the overall trust out of individual trust
properties and how to make use of trust in order to maximize application
performance. Essentially adaptive trust management is achieved by (1) selecting
the best trust propagation and aggregation parameter setting to achieve trust
accuracy and convergence and (2) selecting the best trust formation parameter
setting to maximize application performance, in response to an evolving IoT
environment.
Trust
Composition:
While there
is a wealth of social trust metrics available [38], we choose honesty, cooperativeness,
and community-interest as the most striking metrics for characterizing
social IoT systems, as illustrated in Figure 1 (2nd level). These trust
properties are considered orthogonal but complementary to each other to
characterize a node. Each trust property is evaluated separately as follows:
1. The honesty trust property represents whether or not a node is
honest. In IoT, a malicious node can be dishonest when providing services or
trust recommendations. We select honesty as a trust property because a
dishonest node can severely disrupt trust management and service continuity of
an IoT application. In an IoT application, a node relies on direct evidence
(upon interacting) and indirect evidence (upon hearing recommendations vs. own
assessment toward a third-party node) to evaluate the honesty trust property of
another node.
The cooperativeness trust property represents whether or not the
trustee node is socially cooperative [28] with the trust or node. A node may
follow a prescribed protocol only when interacting with its friends or nodes
with strong social ties (with many common friends), but become uncooperative
when interacting with other nodes. In an IoT application, a node can evaluate
the cooperativeness property of other nodes based on social ties and select
socially cooperative nodes in order to achieve high application performance.
3. The community-interest trust
represents whether or not the trustor and trustee nodes are in the same social
communities/groups (e.g. co-location or co-work relationships [3]) or have
similar capabilities (e.g., parental object relationships [3]). Two nodes with
a degree of high community-interest trust have more chances and experiences in
interacting with each other, and thus can result in better application
performance.
Protocol
performance evaluation:
Adaptive
trust management is a continuing process which iteratively aggregates past
information and new information. The new information includes both direct
observations (first-hand information) and indirect recommendations (second-hand
information). The trust assessment of node i towards node j at
time t is denoted by 𝑇𝑖𝑗𝑋
(𝑡) where X =
honesty, cooperativeness, or community-interest.
Configuration:-
H/W System Configuration:-
Processor - Pentium –III
Speed - 1.1 Ghz
RAM - 256
MB(min)
Hard Disk - 20 GB
Floppy Drive - 1.44 MB
Key Board - Standard Windows Keyboard
Mouse - Two or Three Button Mouse
Monitor - SVGA
S/W System
Configuration:-
v Operating System :Windows/XP/7.
v Application
Server : Tomcat5.0/6.X
v Front End : HTML, Java, Jsp
v Scripts : JavaScript.
v Server side Script :
Java Server Pages.
v Database : Mysql 5.0
v
Database
Connectivity : JDBC.
Trust-based Service Management for Social Internet
of Things
Systems
Abstract:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted “things” that can provide services needed when they come into
contact with each other opportunistically. We propose and analyze the design
notion of adaptive trust management for social IoT systems in which
social relationships evolve dynamically among the owners of IoT devices. We
reveal the design tradeoff between trust convergence vs. trust fluctuation in
our adaptive trust management protocol design. With our adaptive trust
management protocol, a social IoT application can adaptively choose the best
trust parameter settings in response to changing IoT social conditions such
that not only trust assessment is accurate but also the application performance
is maximized. We propose a table-lookup method to apply the analysis results
dynamically and demonstrate the feasibility of our proposed adaptive trust
management scheme with two real-world social IoT service composition
applications.
Architecture Diagram:
Existing System:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer (P2P) networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted things that can provide services needed when they come into
contact with each other opportunistically in both the physical world and
cyberspace. It is envisioned that the future social IoT will connect a great
amount of smart objects in the physical world, including radio frequency
identification (RFID) tags, sensors [40], actuators, PDAs, and smart phones, as
well as virtual objects in cyberspace such as data and virtual desktops on the
cloud.
Disadvantages:
The emerging
paradigm of the social Internet of Things (IoT) has attracted a wide variety of
applications running on top of it, including e-health, smart-home, smart-city,
and smart-community .
Proposed System:
A social IoT applications are likely oriented
toward a service oriented architecture where each thing plays the role of
either a service provider or a service requester, or both, according to the
rules set by the owners. Unlike a traditional service-oriented P2P network,
social networking and social relationship play an important role in a social
IoT, since things (real or virtual) are essentially operated by and work for
humans. Therefore, social relationships among the users/owners must be taken
into account during the design phase of social IoT applications. A social IoT
system thus can be viewed as a P2P owner-centric community with devices (owned
by humans) requesting and providing services on behalf of the owners. IoT
devices establish social relationships autonomously with other devices based on
social rules set by their owners, and interact with each other
opportunistically as they come into contact.
Advantages:
1. To best satisfy the service requester
and maximize application performance, it is crucial to evaluate the
trustworthiness of service providers in social IoT environments.
2.
The motivation of providing a
trust management system for a social IoT system is clear: There are misbehaving
owners and consequently misbehaving devices that may perform discriminatory
attacks based on their social relationships with others for their own gain at
the expense of other IoT devices which provide similar services.
Implementation modules:
1.
User-Centric Social IoT Environments
2.
Adaptive Trust Management
3.
Trust Composition
4.
Protocol performance evaluation
User-Centric
Social IoT Environments:
We consider a
user-centric social IoT environment with no centralized trusted authority. Each
IoT device has its unique identity which can be achieved through standard
techniques such as PKI. A device communicates with other devices through the
overlay social network protocols, or the underlying standard communication
network protocols (wired or wireless). Every device has an owner who could have
many devices. Social relationships between owners is translated into social
relationships between IoT devices as follows: Each owner has a list of friends
(i.e., other owners), representing its social relationships. This friendship
list varies dynamically as an owner makes or denies other owners as friends. If
the owners of two nodes are friends, then it is likely they will be cooperative
with each other. A device may be carried or operated by its owner in certain
community-interest environments (e.g., work vs. home or a social club). Nodes
belonging to a similar set of communities likely share similar interests or
capabilities. Our social IoT model is based on social relationships among
humans who are owners of IoT devices. We note that the device-to-device
autonomous social relationship is also a potential for the social IoT paradigm.
Adaptive
Trust Management:
A design
parameter is one that adaptive trust management can control to optimize
performance. A derived parameter is one that is generated during runtime as a
result of running the trust protocol. An input parameter is one that the
operating environment dictates. addresses all aspects of trust management: the
trust composition component addresses the issue of how to select
multiple trust properties according to social IoT application requirements. The
trust propagation and aggregation component addresses the issue
of how to disseminate and combine trust information such that the trust
assessment converges and is accurate. The trust formation component
addresses the issue of how to form the overall trust out of individual trust
properties and how to make use of trust in order to maximize application
performance. Essentially adaptive trust management is achieved by (1) selecting
the best trust propagation and aggregation parameter setting to achieve trust
accuracy and convergence and (2) selecting the best trust formation parameter
setting to maximize application performance, in response to an evolving IoT
environment.
Trust
Composition:
While there
is a wealth of social trust metrics available [38], we choose honesty, cooperativeness,
and community-interest as the most striking metrics for characterizing
social IoT systems, as illustrated in Figure 1 (2nd level). These trust
properties are considered orthogonal but complementary to each other to
characterize a node. Each trust property is evaluated separately as follows:
1. The honesty trust property represents whether or not a node is
honest. In IoT, a malicious node can be dishonest when providing services or
trust recommendations. We select honesty as a trust property because a
dishonest node can severely disrupt trust management and service continuity of
an IoT application. In an IoT application, a node relies on direct evidence
(upon interacting) and indirect evidence (upon hearing recommendations vs. own
assessment toward a third-party node) to evaluate the honesty trust property of
another node.
The cooperativeness trust property represents whether or not the
trustee node is socially cooperative [28] with the trust or node. A node may
follow a prescribed protocol only when interacting with its friends or nodes
with strong social ties (with many common friends), but become uncooperative
when interacting with other nodes. In an IoT application, a node can evaluate
the cooperativeness property of other nodes based on social ties and select
socially cooperative nodes in order to achieve high application performance.
3. The community-interest trust
represents whether or not the trustor and trustee nodes are in the same social
communities/groups (e.g. co-location or co-work relationships [3]) or have
similar capabilities (e.g., parental object relationships [3]). Two nodes with
a degree of high community-interest trust have more chances and experiences in
interacting with each other, and thus can result in better application
performance.
Protocol
performance evaluation:
Adaptive
trust management is a continuing process which iteratively aggregates past
information and new information. The new information includes both direct
observations (first-hand information) and indirect recommendations (second-hand
information). The trust assessment of node i towards node j at
time t is denoted by 𝑇𝑖𝑗𝑋
(𝑡) where X =
honesty, cooperativeness, or community-interest.
Configuration:-
H/W System Configuration:-
Processor - Pentium –III
Speed - 1.1 Ghz
RAM - 256
MB(min)
Hard Disk - 20 GB
Floppy Drive - 1.44 MB
Key Board - Standard Windows Keyboard
Mouse - Two or Three Button Mouse
Monitor - SVGA
S/W System
Configuration:-
v Operating System :Windows/XP/7.
v Application
Server : Tomcat5.0/6.X
v Front End : HTML, Java, Jsp
v Scripts : JavaScript.
v Server side Script :
Java Server Pages.
v Database : Mysql 5.0
v
Database
Connectivity : JDBC.
Trust-based Service Management for Social Internet
of Things
Systems
Abstract:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted “things” that can provide services needed when they come into
contact with each other opportunistically. We propose and analyze the design
notion of adaptive trust management for social IoT systems in which
social relationships evolve dynamically among the owners of IoT devices. We
reveal the design tradeoff between trust convergence vs. trust fluctuation in
our adaptive trust management protocol design. With our adaptive trust
management protocol, a social IoT application can adaptively choose the best
trust parameter settings in response to changing IoT social conditions such
that not only trust assessment is accurate but also the application performance
is maximized. We propose a table-lookup method to apply the analysis results
dynamically and demonstrate the feasibility of our proposed adaptive trust
management scheme with two real-world social IoT service composition
applications.
Architecture Diagram:
Existing System:
A social
Internet of Things (IoT) system can be viewed as a mix of traditional
peer-to-peer (P2P) networks and social networks, where “things” autonomously
establish social relationships according to the owners’ social networks, and
seek trusted things that can provide services needed when they come into
contact with each other opportunistically in both the physical world and
cyberspace. It is envisioned that the future social IoT will connect a great
amount of smart objects in the physical world, including radio frequency
identification (RFID) tags, sensors [40], actuators, PDAs, and smart phones, as
well as virtual objects in cyberspace such as data and virtual desktops on the
cloud.
Disadvantages:
The emerging
paradigm of the social Internet of Things (IoT) has attracted a wide variety of
applications running on top of it, including e-health, smart-home, smart-city,
and smart-community .
Proposed System:
A social IoT applications are likely oriented
toward a service oriented architecture where each thing plays the role of
either a service provider or a service requester, or both, according to the
rules set by the owners. Unlike a traditional service-oriented P2P network,
social networking and social relationship play an important role in a social
IoT, since things (real or virtual) are essentially operated by and work for
humans. Therefore, social relationships among the users/owners must be taken
into account during the design phase of social IoT applications. A social IoT
system thus can be viewed as a P2P owner-centric community with devices (owned
by humans) requesting and providing services on behalf of the owners. IoT
devices establish social relationships autonomously with other devices based on
social rules set by their owners, and interact with each other
opportunistically as they come into contact.
Advantages:
1. To best satisfy the service requester
and maximize application performance, it is crucial to evaluate the
trustworthiness of service providers in social IoT environments.
2.
The motivation of providing a
trust management system for a social IoT system is clear: There are misbehaving
owners and consequently misbehaving devices that may perform discriminatory
attacks based on their social relationships with others for their own gain at
the expense of other IoT devices which provide similar services.
Implementation modules:
1.
User-Centric Social IoT Environments
2.
Adaptive Trust Management
3.
Trust Composition
4.
Protocol performance evaluation
User-Centric
Social IoT Environments:
We consider a
user-centric social IoT environment with no centralized trusted authority. Each
IoT device has its unique identity which can be achieved through standard
techniques such as PKI. A device communicates with other devices through the
overlay social network protocols, or the underlying standard communication
network protocols (wired or wireless). Every device has an owner who could have
many devices. Social relationships between owners is translated into social
relationships between IoT devices as follows: Each owner has a list of friends
(i.e., other owners), representing its social relationships. This friendship
list varies dynamically as an owner makes or denies other owners as friends. If
the owners of two nodes are friends, then it is likely they will be cooperative
with each other. A device may be carried or operated by its owner in certain
community-interest environments (e.g., work vs. home or a social club). Nodes
belonging to a similar set of communities likely share similar interests or
capabilities. Our social IoT model is based on social relationships among
humans who are owners of IoT devices. We note that the device-to-device
autonomous social relationship is also a potential for the social IoT paradigm.
Adaptive
Trust Management:
A design
parameter is one that adaptive trust management can control to optimize
performance. A derived parameter is one that is generated during runtime as a
result of running the trust protocol. An input parameter is one that the
operating environment dictates. addresses all aspects of trust management: the
trust composition component addresses the issue of how to select
multiple trust properties according to social IoT application requirements. The
trust propagation and aggregation component addresses the issue
of how to disseminate and combine trust information such that the trust
assessment converges and is accurate. The trust formation component
addresses the issue of how to form the overall trust out of individual trust
properties and how to make use of trust in order to maximize application
performance. Essentially adaptive trust management is achieved by (1) selecting
the best trust propagation and aggregation parameter setting to achieve trust
accuracy and convergence and (2) selecting the best trust formation parameter
setting to maximize application performance, in response to an evolving IoT
environment.
Trust
Composition:
While there
is a wealth of social trust metrics available [38], we choose honesty, cooperativeness,
and community-interest as the most striking metrics for characterizing
social IoT systems, as illustrated in Figure 1 (2nd level). These trust
properties are considered orthogonal but complementary to each other to
characterize a node. Each trust property is evaluated separately as follows:
1. The honesty trust property represents whether or not a node is
honest. In IoT, a malicious node can be dishonest when providing services or
trust recommendations. We select honesty as a trust property because a
dishonest node can severely disrupt trust management and service continuity of
an IoT application. In an IoT application, a node relies on direct evidence
(upon interacting) and indirect evidence (upon hearing recommendations vs. own
assessment toward a third-party node) to evaluate the honesty trust property of
another node.
The cooperativeness trust property represents whether or not the
trustee node is socially cooperative [28] with the trust or node. A node may
follow a prescribed protocol only when interacting with its friends or nodes
with strong social ties (with many common friends), but become uncooperative
when interacting with other nodes. In an IoT application, a node can evaluate
the cooperativeness property of other nodes based on social ties and select
socially cooperative nodes in order to achieve high application performance.
3. The community-interest trust
represents whether or not the trustor and trustee nodes are in the same social
communities/groups (e.g. co-location or co-work relationships [3]) or have
similar capabilities (e.g., parental object relationships [3]). Two nodes with
a degree of high community-interest trust have more chances and experiences in
interacting with each other, and thus can result in better application
performance.
Protocol
performance evaluation:
Adaptive
trust management is a continuing process which iteratively aggregates past
information and new information. The new information includes both direct
observations (first-hand information) and indirect recommendations (second-hand
information). The trust assessment of node i towards node j at
time t is denoted by 𝑇𝑖𝑗𝑋
(𝑡) where X =
honesty, cooperativeness, or community-interest.
Configuration:-
H/W System Configuration:-
Processor - Pentium –III
Speed - 1.1 Ghz
RAM - 256
MB(min)
Hard Disk - 20 GB
Floppy Drive - 1.44 MB
Key Board - Standard Windows Keyboard
Mouse - Two or Three Button Mouse
Monitor - SVGA
S/W System
Configuration:-
v Operating System :Windows/XP/7.
v Application
Server : Tomcat5.0/6.X
v Front End : HTML, Java, Jsp
v Scripts : JavaScript.
v Server side Script :
Java Server Pages.
v Database : Mysql 5.0
v
Database
Connectivity : JDBC.
Comments
Post a Comment