J. Amudhavel ¹, Giri Sruthy ², D. Padmashree ², N. Pazhani Raja ³, M. S. Saleem Basha ⁴, B. Bhuvaneswari ⁵

Affiliations
1 Department of CSE,Mannadipet Commune, Madagadipet, Puducherry – 605107, Pondicherry, IN
2 Department of CSE, Mannadipet Commune, Madagadipet, Puducherry – 605107, Pondicherry, IN
3 Department of IT, SMVEC, Mannadipet Commune, Madagadipet, Puducherry – 605107, Pondicherry, IN
4 Department of CSE, Mazoon University College, Muscat
5 Department of CSE, Pondicherry University, Kalapet, Puducherry – 605014, Pondicherry, IN

Abstract

Background/Objectives: To analyze and find the decision making systems in multi agent capable of solving complex problems. Method/Statistical Analysis: Multi agent systems are the collection of many individual intelligent systems. Decision making is important because a multi agent system consists of many agents that may be homogeneous or heterogeneous. In heterogeneous network agent must trust another agents in the network for sharing of messages. Hence an agent must be capable to make decision towards trusting of neighbor agents. Decision making technique plays an important role to make decision in such a situation. This is one scenario. Multiple scenarios are discussed in this paper towards the decision making capability of multi agent systems. Findings: In this research, the study of multi agent system, problem solving and decision making are considered as the two important concepts. Multi agent systems are capable of interacting with different environments like virtual environment or real time environment. In this paper a survey is done towards the decision making capability of multi agent systems. Applications/Improvements: The results from this work serve as the motivation to apply the future implementation of multi agent decision making in the complex problem solving.

Keywords

Multi Agent Systems, Decision Making, Environment, Agents, Trust, Problem Solving

Full Text

   

J. Amudhavel ¹, C. Kodeeswari ¹, S. Jarina ¹, S. Jaiganesh ², B. Bhuvaneswari ³

Affiliations
1 Department of CSE, SMVEC, Pondicherry - 605107, Tamil Nadu, IN
2 Department of CSE, R&D Centre, Bharathiar University, Coimbatore - 641046, Tamil Nadu, IN
3 Department of CSE, Pondicherry University, Pondicherry - 605014, Tamil Nadu, IN

Abstract

Background/Objectives: To analyze the behaviors of Forager ants and to utilize some form of memory as a warning system of possible harm to all other members of a colony. Methods/Statistical Analysis: Ants acquire experiential memory within a generation. Experience is shared among members of the colony. For instance, knowledge about an unsuitable leaf is shared by means of avoidance, which is subsequently reinforced by autocatalysis. The loss of avoidance thus marks the loss of past knowledge of the unsuitability of a set of leaves. Findings: In this research, the highly social behavior of ants, a colony appears to act as a collective entity that is more than the sum of its parts. An individual ant to possess the level of intelligence that enables complex decision-making, the colony made up of such individuals appear to exhibit such complex task. The interactions between individual ants in the colony seem to synchronize with each other and align with group tasks. Ants interact with one another by means of basic signaling protocols, such as trophallaxis among leaf-cutting ants. Applications/Improvements: The system has capacity of the ant colony to exhibit transitions from one collective decision and colony memory to the next. The colony of leaf-cutting antscan switch from avoidance to acceptance of leaves which had been deemed unsuitable due to the presence of unwarranted chemical compounds. The transition is not as gradual as one would expect but rather abrupt as evident in the intake rates measured.

Keywords

Ant Colonies, Decision-Making, Leaf-Cutting

Full Text

   

J. Amudhavel ¹, R. Ilamathi ¹, D. Pradeepa ¹, Sasidevi Ganesan ¹, B. Bhuvaneswari ²

Affiliations
1 Department of CSE, SMVEC, Pondicherry - 605107, Tamil Nadu, IN
2 Department of CSE, Pondicherry University, Pondicherry - 605014, Tamil Nadu, ID

Abstract

Background/Objectives: To analyze the Nutrient diffusion through an ant colony is a complex process of transmitting information from one node to another. Methods/Statistical Analysis: Ants have a special mode of transmission known as trophallaxis, which involves the transfer of liquid-based food from one ant's mouth to another. The worker ants may be categorized into two types: forager and non-forager, both of which are involved in the food transfer. Findings: In this research, by considering trophallaxis, the diffusion of the food signal through the ant colony is modelled as a two-type branching process. Analytic results based on asymptotic estimates of the branching process are here shown to be consistent with empirical results gathered from mini-colony experiments. The division of labor in worker ants is an important feature that exhibits the sociality of the colony. Indeed, resource allocation becomes more efficient if each worker ant is able to focus on one particular task only throughout its lifetime. Applications/Improvements: The division of labor, forager ants tends to be the more “aggressive” type in terms of seeking ants from which food can be received and subsequently exchanged. Meanwhile, the non-forager ants are more specialized in nest-maintenance activities and because of their relatively smaller size are more likely to obtain nourishment by another means. The trophallactic transmission process is investigated by Moreira et al. by conducting an experiment on a mini-colony of ants consisting of a mixture of the two classes3.

Keywords

Leaf-Cutting, Nutrient, Trophallaxis

Full Text

   

J. Amudhavel ¹, V. Brindha ¹, B. Anantharaj ², P. Karthikeyan ¹, B. Bhuvaneswari ³, M. Vasanthi ¹, D. Nivetha ¹, D. Vinodha ⁴

Affiliations
1 Department of CSE, SMVEC, Pondicherry - 605107, Tamil Nadu, IN
2 Department of CSE, TCET, Vandavasi - 627425, Tamil Nadu, IN
3 Department of CSE, Pondicherry University, Pondicherry - 605014, Tamil Nadu, IN
4 Department of CSE, SMVEC, Pondicherry - 605107, Tamil Nadu

Abstract

Background/Objectives: To analyze and monitoring the activity of nodes using an IDS Intrusion Detection System which provide security to the system. Findings: In this research, a network building was made with simple node formation with a fitting technique. So that it is easy for a user or beginner to understand the asymptotic dominion factors, due to the wide growth of nodes on network. Intrusion detection system is made of several nodes. These nodes are grouped and combined as a network. Since each node in a network has unique characteristics they may have high chance of attack from malware during the communication between systems. Applications/Improvements: Network of topology is done by creating and connecting various sub networks. Sub networks traffic is cleared and monitored with the package of IDS. It works on synchronous message transfer mode which provides acknowledgement to sender and receiver. All sub networks is managed by its corresponding system and further provide update to the master node on the network. Due to this broadcast of messages all other nodes may be in safe state of avoidance from attacker.

Keywords

Ad-Hoc Network, Conversation Exchange Dynamics, Intrusion Detection System, Intrusion Prevention System, Network Security

Full Text