Method and apparatus for organizing and detecting swarms in a network
A method and an apparatus for organization and detection of homogeneous and heterogeneous swarms of devices and application of swarm intelligence using swarm intelligence framework are provided. The Swarm Intelligence Framework provides a generic platform for realizing solutions involving Swarm Intelligence Technology via flexible container-based Algorithm Plug-in Architecture which is essential to utilize Swarm Intelligence Framework for various scenarios and use cases, including dynamically loading and using the Swarm Detection Algorithm.
1. A method of detecting swarms in a network, the method comprising:
receiving input characteristics of nodes in the network;
generating a matrix representing a connectivity graph of devices based on the input characteristics comprising proximity among the nodes and homogeneity of the nodes;
detecting swarms based on an analysis of the input characteristics and the matrix; and
selecting at least one algorithm based on at least one predefined parameter, type of nodes, or network characteristics,
wherein the analysis of the input characteristics and the matrix comprises:
comparing the input characteristics against validation thresholds, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are multi-dimensional parameters; and
comparing the input characteristics against at least one validation threshold, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are uni-dimensional parameters,
wherein the at least one algorithm comprises a pluggable algorithm which is plugged-in using a container-based algorithm plug-in architecture, wherein the container-based algorithm plug-in architecture enables binary level reuse of the at least one algorithm in different applications, and wherein the at least one algorithm comprises a generic swarm detection algorithm using a set inclusion and a node degree representing a number of edges connected to at least one node among the nodes.
2. The method of claim 1 , further comprising:
detecting connection and disconnection among the nodes.
3. The method of claim 2 , wherein the input characteristics of the nodes further comprise at least one of network parameters, or information on connection or disconnection among the nodes.
4. The method of claim 3 , wherein the proximity among the nodes indicates signal strength among the nodes.
5. The method of claim 1 , further comprising:
detecting an anomalous node or a malicious node among the nodes based on consensus algorithm or an agreement by neighboring nodes in the network,
wherein the neighboring nodes are nodes located close to each other, and
wherein the consensus algorithm is plugged in a plug-in architecture.
6. The method of claim 5 , wherein the detecting of the anomalous node or the malicious node comprises transmitting a detection status to an electronic device.
7. The method of claim 1 , further comprising:
notifying the detected swarm to a monitoring unit.
8. The method of claim 1 , further comprising:
initializing connectivity among the nodes in the detected swarm for communication; and
sharing data among the nodes in the detected swarm.
9. An apparatus for detecting swarms in a network,
the apparatus comprising:
a memory to store instructions; and
at least one processor, by executing the instructions, configured to:
receive input characteristics of nodes in the network,
generate a matrix representing a connectivity graph of devices based on the input characteristics comprising proximity among the nodes and homogeneity of the nodes,
detect swarms based on an analysis of the input characteristics and the matrix, and
select at least one algorithm based on at least one predefined parameter, type of nodes, or network characteristics,
wherein the analysis of the input characteristics and the matrix comprises:
comparing the input characteristics against validation thresholds, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are multi-dimensional parameters, and
comparing the input characteristics against at least one validation threshold, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are uni-dimensional parameters,
wherein the at least one algorithm comprises a pluggable algorithm which is plugged-in using a container-based algorithm plug-in architecture,
wherein the container-based algorithm plug-in architecture enables binary level reuse of the at least one algorithm in different applications, and
wherein the at least one algorithm comprises a generic swarm detection algorithm using a set inclusion and a node degree representing a number of edges connected to at least one node among the nodes.
10. The apparatus of claim 9 , wherein the at least one processor is further configured to detect connection and disconnection among the nodes.
11. The apparatus of claim 9 , wherein the input characteristics comprise at least one of proximity among the nodes, homogeneity of the nodes, network parameters, or information on connection or disconnection among the nodes.
12. The apparatus of claim 9 ,
wherein the at least one processor is further configured to detect an anomalous node or a malicious node among the nodes based on consensus algorithm or an agreement by neighboring nodes in the network, and
wherein the neighboring nodes are nodes located close to each other and the consensus algorithm is plugged in a plug-in architecture.
13. A non-transitory computer readable medium embodying a computer program for operating an electronic device including a memory and at least one processor, the computer program comprising computer readable program code that, when executed by the at least one processor, causes the electronic device to:
receive input characteristics of nodes in a network;
generate a matrix representing a connectivity graph of devices based on the input characteristics comprising proximity among the nodes and homogeneity of the nodes;
detect swarms based on an analysis of the input characteristics and the matrix; and
select at least one algorithm based on at least one predefined parameter, type of nodes, or network characteristics,
wherein the analysis of the input characteristics and the matrix comprises:
comparing the input characteristics against validation thresholds, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are
multi-dimensional parameters; and
comparing the input characteristics against at least one validation
threshold, when the input characteristics are value pairs, and representing and processing the input characteristics as a neural network, when the input characteristics are uni-dimensional parameters,
wherein the at least one algorithm comprises a pluggable algorithm which is plugged-in using a container-based algorithm plug-in architecture,
wherein the container-based algorithm plug-in architecture enables binary level reuse of the at least one algorithm in different applications, and
wherein the at least one algorithm comprises a generic swarm detection algorithm using a set inclusion and a node degree representing a number of edges connected to at least one node among the nodes.