Limits...
An approach to model based testing of multiagent systems.

Ur Rehman S, Nadeem A - ScientificWorldJournal (2015)

Bottom Line: These interactions include percepts and actions along with messages between the agents which can be modeled in a protocol diagram.The protocol diagram is converted into a protocol graph, on which different coverage criteria are applied to generate test paths that cover interactions between the agents.A prototype tool has been developed to generate test paths from protocol graph according to the specified coverage criterion.

View Article: PubMed Central - PubMed

Affiliation: Center for Software Dependability, Mohammad Ali Jinnah University, Islamabad 44000, Pakistan.

ABSTRACT
Autonomous agents perform on behalf of the user to achieve defined goals or objectives. They are situated in dynamic environment and are able to operate autonomously to achieve their goals. In a multiagent system, agents cooperate with each other to achieve a common goal. Testing of multiagent systems is a challenging task due to the autonomous and proactive behavior of agents. However, testing is required to build confidence into the working of a multiagent system. Prometheus methodology is a commonly used approach to design multiagents systems. Systematic and thorough testing of each interaction is necessary. This paper proposes a novel approach to testing of multiagent systems based on Prometheus design artifacts. In the proposed approach, different interactions between the agent and actors are considered to test the multiagent system. These interactions include percepts and actions along with messages between the agents which can be modeled in a protocol diagram. The protocol diagram is converted into a protocol graph, on which different coverage criteria are applied to generate test paths that cover interactions between the agents. A prototype tool has been developed to generate test paths from protocol graph according to the specified coverage criterion.

No MeSH data available.


Converting protocol diagram into protocol graph.
© Copyright Policy - open-access
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4385681&req=5

alg1: Converting protocol diagram into protocol graph.

Mentions: Protocol diagram shows details of how messages, action, and percepts are involved in a protocol. In our work we convert the protocol diagram into protocol graph. Protocol graph has been introduced by Miller et al. [8]. They defined two sets of test coverage criteria for multiagent interaction testing. The first uses only the protocol specification, while the second considers also the plans that generate and receive the messages in the protocol. Miller et al. [8] do not cover the actions and percepts during the interaction. We have extended the protocol graph with actions and percepts as they are a very important part of interaction protocol. Algorithm 1 is used to convert protocol diagram into protocol graph. We take protocol diagram as input and protocol graph has been produced by following Algorithm 1. Protocol diagram is represented in AUML representation as well. Code 2 shows AUML description of protocol graph. Protocol graph represents interaction protocol in nodes and vertices form, on which different coverage criteria have been applied.


An approach to model based testing of multiagent systems.

Ur Rehman S, Nadeem A - ScientificWorldJournal (2015)

Converting protocol diagram into protocol graph.
© Copyright Policy - open-access
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4385681&req=5

alg1: Converting protocol diagram into protocol graph.
Mentions: Protocol diagram shows details of how messages, action, and percepts are involved in a protocol. In our work we convert the protocol diagram into protocol graph. Protocol graph has been introduced by Miller et al. [8]. They defined two sets of test coverage criteria for multiagent interaction testing. The first uses only the protocol specification, while the second considers also the plans that generate and receive the messages in the protocol. Miller et al. [8] do not cover the actions and percepts during the interaction. We have extended the protocol graph with actions and percepts as they are a very important part of interaction protocol. Algorithm 1 is used to convert protocol diagram into protocol graph. We take protocol diagram as input and protocol graph has been produced by following Algorithm 1. Protocol diagram is represented in AUML representation as well. Code 2 shows AUML description of protocol graph. Protocol graph represents interaction protocol in nodes and vertices form, on which different coverage criteria have been applied.

Bottom Line: These interactions include percepts and actions along with messages between the agents which can be modeled in a protocol diagram.The protocol diagram is converted into a protocol graph, on which different coverage criteria are applied to generate test paths that cover interactions between the agents.A prototype tool has been developed to generate test paths from protocol graph according to the specified coverage criterion.

View Article: PubMed Central - PubMed

Affiliation: Center for Software Dependability, Mohammad Ali Jinnah University, Islamabad 44000, Pakistan.

ABSTRACT
Autonomous agents perform on behalf of the user to achieve defined goals or objectives. They are situated in dynamic environment and are able to operate autonomously to achieve their goals. In a multiagent system, agents cooperate with each other to achieve a common goal. Testing of multiagent systems is a challenging task due to the autonomous and proactive behavior of agents. However, testing is required to build confidence into the working of a multiagent system. Prometheus methodology is a commonly used approach to design multiagents systems. Systematic and thorough testing of each interaction is necessary. This paper proposes a novel approach to testing of multiagent systems based on Prometheus design artifacts. In the proposed approach, different interactions between the agent and actors are considered to test the multiagent system. These interactions include percepts and actions along with messages between the agents which can be modeled in a protocol diagram. The protocol diagram is converted into a protocol graph, on which different coverage criteria are applied to generate test paths that cover interactions between the agents. A prototype tool has been developed to generate test paths from protocol graph according to the specified coverage criterion.

No MeSH data available.