In an Artificial Intelligence world, Agent-based technology is one of the most vibrant and important areas of R&D, emerging in the Information Technology in industry recent years. In Agent Based Cloud Computing, Intelligent Agent does Coordination, Integration, Mobility, Believable Agent and Assistance in achieving its expectancy. Cloud computing frameworks give huge scale foundations to elite registering that are "flexible" since they can adjust to client and application needs. The Internet of Things is a progressive idea, inside cyberphysical frameworks, wealthy in potential just as in multifacet prerequisites and advancement issues. To appropriately deliver them and to completely bolster IoT frameworks advancement, Agent-Based Computing speaks to a reasonable and compelling demonstrating, programming, reproduction worldview. In this paper, we discuss about agent based cloud computing in a IoT system.

Keywords: Artificial Intelligence, agent, cloud computing, internet of things.

1. INTRODUCTION

Intelligent Agents [1] – [2] initially originated from Artificial Intelligence and Distributed Programming. These two areas have been combined together, to form Distributed Artificial Intelligence (DAI), from which the idea of Intelligent Agents (IA) emerges. The IA concept plays a major role in Artificial Intelligence (AI) as well as Mainstream Computer Science industry. This thesis mainly focuses on the importance of learning the critical factors in theoretical as well as practical issues related to the design and construction of Intelligent Agents. An IA concept emphasizes more on what an IA is, the mathematical formulations of agents in representations, and the reasoning properties of agents. Architecture of an IA is anticipated in the form of software engineering models of agents. The major role of researchers is to design software and hardware systems that accept the properties given by an IA theorist. It also focuses on IA language used in software systems in experimenting and programming. An IA language has all the potential proposed by theorists. 

Cloud computing [3] is a worldview that spotlights on sharing information and calculations over an adaptable system of hubs, crossing crosswise over end client PCs, server farms, and web administrations. An adaptable system of such hubs shapes a cloud. An application in light of these mists is taken as a cloud application. As of late, a large portion of the created programming depends on circulated engineering, for example, administration situated, P2P and distributed computing. With the advancement of PC equipment and systems administration, circulated designs have additionally developed, particularly benefit based distributed computing has changed the customary PC and unified stockpiling approach. It encourages preparing and capacity abilities according to the necessity. 

Infrastructure as-a-Service [9] – [17] (IaaS), gives Virtual Machines (VMs) completely fulfilling the client demands as far as assets. The assets of the suppliers are typically facilitated as a server farm. The server farm is an arrangement of physical machines which are interconnected, virtualized, and topographically appropriated. Since the client may have diverse geographic area, an administration supplier ought to have disseminated server farms all through the world in order to give administrations to the clients. In the distributed computing, the separation between datacenters prompts undesirable system idleness, which thusly prompts delay in administrations. For instance, a VM designated in a server farm, far from client area, the client will experience the ill effects of deferred reaction because of the confinement of system assets. What's more, if the VM is vigorously stacked it expands the reaction time to the client, contrasted with if VM is having less workload. Subsequently, a supplier is required to discover appropriate server farms for serving a client in light of the client area and workload of the server farms. Figure 1.1 given gives the engineering of a server farm. 

Figure 1.1 Architecture of server farm (data center)

As of late, the improvement of various Information and Communication Technologies (ICT) related to the production of ease little sensors have made it conceivable to screen numerous procedures. Remote sensor systems (WSN) [18] – [21] are an unmistakable precedent as they are frequently utilized for cultivating purposes. WSN have been utilized for checking the three force, nurseries and citrus crops. In addition, WSN are utilized to screen the condition of homestead creatures, for example, goats or dairy animals. A few frameworks have been proposed for observing fish ranches. They will be broke down separately in the related work segment. Most of them depend on observing water quality including only two or three water parameters to be checked. In addition, they for the most part utilize business tests. The business tests for submerged checking have a staggering expense. Along these lines, if a WSN were to be used to screen a few parameters utilizing business tests in all the generation tanks, the expense of the framework would be exorbitant for the fish ranches. Also, different creators propose frameworks for observing fish conduct [22] – [26]. In the related work segment, we will examine every proposition. In this manner, in the event that we seek to gauge diverse parameters in fish ranches offices with WSN, it is essential to diminish the expense of the sensors and incorporate a more extensive assortment of parameters in a similar framework.  IoT [28] has effectively demonstrated its enormous measure of uses areas in the most recent years. Notwithstanding, little are the fish cultivates today outfitted with clever gadgets with real-time and associated water observing abilities. There are numerous precedents where IoT could assist aquaculturalists with improving their working conditions. For instance, some fish ranches are far from the land and utilizing IoT to screen water at a separation could decrease their expenses. Another model is that adjustments in water quality can happen in all respects rapidly and whenever, so observing water continuously with cautions cannot miss a specific occasion. 

In this paper, we will probably indicate how Agent Based Computing has been adequately abused for displaying, programming and reproducing IoT frameworks.

2. AGENT BASED COMPUTING

An Intelligent Agent is considered to be a software entity located in an environment. IA [4] can be: Autonomous; respond to changes in the environment; be proactive in attaining its goals; and also Sociable. 

IA is shown in the figure 3.1 [5]. For the purpose of attaining the goal, an IA learns by itself and makes use of its internal knowledge base. Thus it is seen as natural metaphor for human acts. It has an elevated performance behavior in data distribution and control of self-imposed expertise.

Figure 3.1 Basic Agent Diagram

The core utilization of an IA [6] Model is in the area of documentation, where various IA types will be supported in the system which is in the development stage. IA instances will understand these agents clearly during its execution. Characteristics of an IA play a critical role in the implementation of any IA-based Applications. It may be one-to-one link between roles and agent types. A simple IA-type tree defines an IA model, where leaf nodes of the tree respond to roles whereas other intelligent agent types are referred by other nodes.  

3.1 Attributes of Intelligent Agents

The three attributes [7]: agency, intelligence and mobility are used in intelligent agent systems, to measure system properties. Figure 3.2 illustrates the relationship between agency and intelligence:

• Agency - The degree and extent to which independence is exhibited by an agent. For example, given that an agent operates in an Internet environment, it must at the least, be able to go on working while the user might not be connected or might not be connected to the Web. 
• Intelligence - The ability of an agent to learn and adapt to an environment, in terms of user requests and available resources to the agent. 

Figure 3.2 Intelligent agent scope

Systems above the threshold lines are recognized as intelligent agents as shown in Figure 3.2. Those falling below the threshold line, for example expert systems i.e. “systems representing some knowledge they gathered by means of elicitation or knowledge-acquisition into a computer program to perform specific tasks”.

3.2       Agent Function

The agent function which maps a series of observations into action is a mathematical function. The agent function is applied as agent program. The part in which the agent takes an action is called an actuator.

3.2.1 Agent versus Program

• Size: The size of an agent is usually less than that of a program.
• Purpose: The purpose of an agent is limited and specific whereas the purpose of programs is multi-functional.
• Persistence: an agent's life span does not depend entirely on a user launching and quitting it.
• Autonomy: an agent is autonomous and is not dependent on user's input for functioning.

An agent is in charge of fulfilling particular objectives. There can be distinctive sorts of objectives, for example, accomplishing a particular status, augmenting a given capacity (e.g., utility), and so on. The state of an agent incorporates state of its internal environment and state of knowledge and beliefs about its external environment.

Figure 3.3 Basic structure of agent

3.3       Multi-Agent Systems (MAS)

A Multi-Agent System (MAS) is a product framework that utilizes various intelligent operators to take care of an issue in open and decentralized unsure situations. A focal element of MAS is that there is no brought together control instrument; operators are required to team up to accomplish the structure goal of a given MAS. A MAS has aggregate capacities that an individual operator does not have.

In MAS computational assets and abilities are circled over a system of interconnected specialists to take care of issues that are unreasonably vast for an individual operator. A unified framework might be tormented by asset restrictions, execution bottlenecks, or basic disappointments.

A MAS considers the interconnection and collaboration of different existing heritage frameworks; by structure a specialist wrapper around such frameworks, with the goal that they can be fused into an operator society. In MAS issues are displayed regarding self-ruling collaborating part specialists that enable these operators to work in self-coordinated way.

In a MAS data from sources that are spatially appropriated is effectively recovered, separated, and internationally planned. Utilization of MAS gives arrangements in circumstances where mastery is spatially and transiently appropriated. As for the social capacity of specialists, mastery and assets can be shared. Utilization of MAS improves by and large framework execution, particularly along the elements of: computational effectiveness, unwavering quality, extensibility, vigor, viability, responsiveness, adaptability and reuse because of its conveyed nature.

3. CONTRIBUTION OF AGENTS’ IN DEVELOPING IOT SYSTEMS

The agent-oriented perspective on the world is maybe the most regular method for moving toward a few kinds of (normal and fake) frameworks, included by an important multifaceted nature, dynamicity, situatedness and self-sufficiency [7]. Specifically, solid theoretical connection exists among specialists and SOs, just as among MAS and IoT frameworks [2]. Along these lines, considering the whole arrangement of necessities and issues identified with the improvement of IoT frameworks, ABC has been misused for displaying, programming and recreating IoT applications and frameworks, and accordingly deliberately driving and accelerating their advancement.

There are four technical communications implementation models for internet of Everything (IoE) [8] as defined by the Internet Architecture Board. These models are: Device-to-Device (D2D), Device-to-Cloud, Device-to-Gateway/Server (D2S) and Back-End Data Sharing. The models clearly show how flexible devices can connect and communicate to provide the necessary value-added services for users. This protocol design specification is aimed at the device-to-device (D2D) communication model.

4. CONCLUSION

IoT full acknowledgment isn't thwarted by equipment imperatives or computational/stockpiling/correspondence constraints, however by certain necessities that have not been absolutely or all the while tended to. Utilizing specialists key highlights of independence, proactiveness, insight and friendliness and, as indicated by the various commitments overviewed in this work, we trust that ABC can be adequately misused as displaying, programming, and recreation worldview for creating IoT environments. In fact, superior to anything other figuring ideal models (object-situated, administration arranged, segment arranged) and both at things and at framework levels, ABC permits demonstrating at various degrees of subtleties, encouraging (specialized, linguistic and semantic) interoperability, autonomicity and appropriated knowledge, and approving different plan decisions, before their real organization.

5. REFERENCES

1. Balakrishnan. S and K L Shunmuganathan. Article: A JADE Implementation of Integrated Agent System for E-Mail Coordination (IASEC). International Journal of Computer Applications 58(5): 5-9, November 2012.
2. S.Balakrishnan, “An Overview of Agent Based Intelligent Systems and Its Tools”, CSI Communications magazine, Volume No. 42, Issue No. 10, January 2019, pp. 15-17.
3. S. Balakrishnan, K.N. Sivabalan and J. Janet “MASFE - Mutliagent System for Filtering E-Mails Using JADE”, Advanced Engineering Research and Applications (AERA), Research India Publications, ISBN- 978-93-84443-42-9, pp. 148-167, 2017.
4. P.Arivazhagan, Balakrishnan. S and K L Shunmuganathan. “An Agent Based Centralized Router with Dynamic Connection Management Scheme Using JADE”, International Journal of Applied   Engineering Research, ISSN 0973-4562, Volume 11, Number 3 (2016) pp 2036-2041.
5. Balakrishnan. S and K L Shunmuganathan, R. Sreenevasan, “Amelioration of Artificial Intelligence using Game Techniques for an Imperfect Information Board Game Geister” International Journal of Applied Engineering Research (IJAER). ISSN 0973-4562. Vol 9, Number 22 (2014) pp. 11849-11860.
6. Balakrishnan. S and K L Shunmuganathan, An Agent Based Collaborative Spam Filtering Assistance Using JADE”, International Journal of Applied   Engineering Research, ISSN 0973-4562, Volume 10, Number 21 (2015) pp 42476-42479.
7. A.Jebaraj Rathnakumar, S.Balakrishnan, Design Of Multi-Agent Based Systems For Entrusted Communication Using JADE”, Taga Journal of Graphic Technology, Vol. 14, pp. 766-774, 2018. 
8. S.Balakrishnan, “Peer-To-Peer Central Registry Based Internet of Everything (IoE) Protocol”, CSI Communications magazine, Vol. 41, issue 4, July 2017, pp. 26-29.
9. S. Balakrishnan, J. Janet, K.N. Sivabalan, “Secure Data Sharing in a Cloud Environment by Using Biometric Leakage resilient Authenticated Key Exchange”, Pak. J. Biotechnol. Vol. 15 (2) 293-297 (2018).
10. J. Janet, S. Balakrishnan and E. Murali, "Improved data transfer scheduling and optimization as a service in cloud," 2016 International Conference on Information Communication and Embedded Systems (ICICES), Chennai, 2016, pp. 1-3.
    doi: 10.1109/ICICES.2016.7518895.
11. Balakrishnan S., Janet J., Spandana S. ”Extensibility of File Set Over Encoded Cloud Data Through Empowered Fine Grained Multi Keyword Search”. In: Deiva Sundari P., Dash S., Das S., Panigrahi B. (eds) Proceedings of 2nd International Conference on Intelligent Computing and Applications. Advances in Intelligent Systems and Computing, vol 467. 2017. Springer, Singapore.
12. J. Janet, S. Balakrishnan and K. Somasekhara, "Fountain code based cloud storage mechanism for optimal file retrieval delay," 2016 International Conference on Information Communication and Embedded Systems (ICICES), Chennai, 2016, pp. 1-4.
    doi: 10.1109/ICICES.2016.7518901.
13. J. Janet, S. Balakrishnan and E. R. Prasad, "Optimizing data movement within cloud environment using efficient compression techniques," 2016 International Conference on Information Communication and Embedded Systems (ICICES), Chennai, 2016, pp. 1-5.
    doi: 10.1109/ICICES.2016.7518896.
14. M. Balasubramaniyan, M. Balasubramanian, S. Balakrishnan, “Data Movement Optimization In A Cloud Environment Using Capacity Optimization Technique”, Jour of Adv Research in Dynamical & Control Systems. Vol. 10, 11-Special Issue, 2018, pp. 740- 743.
15. Sruthi Anand, N.Susila, S.Balakrishnan, Challenges and Issues in Ensuring Safe Cloud Based Password Management to Enhance Security”, International Journal of Pure and Applied Mathematics, Volume 119, No. 12, 2018, pp.1207-1215.
16. Dipon Kumar Ghosh , Prithwika Banik , Dr. S. Balakrishnan (2018), “Review-Guppy: A Decision-Making Engine for Ecommerce Products Based on Sentiments of Consumer Reviews”, International Journal of Pure and Applied Mathematics, Volume 119, No. 12, 2018, pp.1135-1141.
17. K. Aravind, J. Granty Regina Elwin, T. Sujatha and S. Balakrishnan, (2018), “A Novel And Efficient Mobile Cloud Service For Searching Encrypted Data”, ARPN Journal of Engineering and Applied Sciences, Vol.13, No.16, pp. 4683- 4686, 2018.
18. Espinosa-Faller, F.J.; Redón-Rpdríguez, G.E. 2012, “A ZigBee Wireless Sensor Network for Monitoring an Aquaculture Recirculating System”. J. Appl. Res. Technol. 2012, 10, 380–387.
19. S. Balakrishnan, B. Persis Urbana Ivy and S. Sudhakar Ilango, 2018, “A Novel And Secured Intrusion Detection System For Wireless Sensor Networks Using Identity Based Online/Offline Signature”, ARPN Journal of Engineering and Applied Sciences. November 2018, Vol. 13  No. 21, pp. 8544-8547.
20. S.Balakrishnan, Vinod K, B. Shaji. 2018, “Secured and Energy Efficient AODV Routing Protocol For Wireless Sensor Network”, International Journal of Pure and Applied Mathematics, Vol. 119, No. 10c, 2018, pp. 563-570.
21. S.Balakrishnan, J.P.Ananth, L.Ramanathan, S.P.Premnath, 2018, “An Adaptive Energy Efficient Data Gathering In Wireless Sensor Networks”, International Journal of Pure and Applied Mathematics, Volume 118 No. 21, 2018, pp. 2501-2510.
22. S.Balakrishnan, S.Sheeba Rani, K.C.Ramya, “Design and Development of IoT Based Smart Aquaculture System in a Cloud Environment”, International Journal of Oceans and Oceanography, ISSN 0973-2667, Volume 13, Number 1 (2019), pp. 121-127.
23. J.Janet, S.Balakrishnan, S.Sheeba Rani, “IOT Based Fishery Management System”, International Journal of Oceans and Oceanography, ISSN 0973-2667, Volume 13, Number 1 (2019), pp. 147-152.
24. J.Janet, S.Balakrishnan, S.Sheeba Rani, “IoT based lake and reservoir management system”, International Journal of Lakes and Rivers (IJLR).
25. S.Sheeba Rani, S.Balakrishnan, V.Kamatchi Sundari, K.C.Ramya, IoT Based Water Level Monitoring System for Lake in a Cloud Environment, International Journal of Lakes and Rivers (IJLR).
26. Ranjeethapriya K, Susila N, Granty Regina Elwin, Balakrishnan S, “Raspberry Pi Based Intrusion Detection System”, International Journal of Pure and Applied Mathematics, Volume 119, No. 12, 2018, pp.1197-1205.
27. K. Dasaradharami Reddy, S. Mohanraju, Dr.A. Jebaraj Ratnakumar, Dr.S. Balakrishnan, “Querying and Searching of Friendship Selection in the Social IoT, Jour of Adv Research in Dynamical & Control Systems. Vol.10, 11-Special issue, 2018, pp. 910- 914.
28. V. Anandkumar, Kalaiarasan T R, S.Balakrishnan, “IoT Based Soil Analysis and Irrigation System”, International Journal of Pure and Applied Mathematics, Volume 119, No. 12, 2018, pp.1127-1134.

Dr.S. Balakrishnan

Professor and Head, Department of Computer Science and Business Systems, Sri Krishna College of Engineering and Technology, Coimbatore, Tamilnadu, India.