4. DESCRIPTION: References:
[1] C. de Paula Bianchini, " Intelligent management of network devices aided by a strategy and a tool," Proceedings of IFIP/ACM Latin America conference on Towards a Latin American agenda for network research, 2003
[2] L. Zhong.M. Sinclair,N. K. Jha, "A personal-area network of low-power wireless interfacing devicesfor handhelds: system and
hardware design," Proceedings of the 7th international conference on Human computer interaction with mobile devices &
services,2005 [3] M. Rob, Xu Wenyuan, K.. Pandurang, and T. Wade, "Service Discovery and Device Identification inCognitive Radio Networks,"
Networking Technologies forSoftware Define Radio Networks,2nd IEEE Workshop onDigital Object Identifier, pp. 40 - 47,2007 [4] Neyre Tekbiyik, Elif Uysal-Biyikoglu, "Energy efficient wireless unicast routing alternatives for machine-to-machine networks" ,
Journal of Network and Computer Applications, Volume 34, Issue 5, September 2011, pp. 1587-1614, ISSN 1084-8045 [5] Jennifer Yick, Biswanath Mukherjee, Dipak Ghosal, "Wireless sensor network survey" , Computer Networks, Volume 52, issue 12, 22
August 2008, pp. 2292-2330 [6] R Baheti and H. Gill. Cyber-physical systems. The Impact of Control Technology, T. Samad and A.M. Annaswamy (eds), 2011 [7] Luigi Atzori, Antonio lera, Giacomo Morabito, The Internet of Things: A survey, Computer Networks, Volume 54, Issue 15,28
October 2010, Pages 2787-2805 [8] J. Choi, and C.Yoo, "Connect with Things through Instant Messaging", Proceedings of the First International Conference on The
Internet of Things, Zurich, Switzerland, 2008 . Springer, Lecture Notes in Computer Science Vol. 4952, pp. 1-18,2008 [9] M. Rodriguez, and). Favela, "A Framework for Supporting Autonomous Agents in Ubiquitous Computing Environments". [10] A.Wang, C Sorensen, and E. Indal, "A Mobile Agent Architecture for Heterogeneous Devices," Proceedings of the International
Conference on Wireless and Optical Communications, pp. 14-16,2003
[11] US Patent: US 2011/0136510 Al, "System and method for migrating agents between mobile devices" [12] US Patent: US 200 5/02 56876 A1, "Distributed applications using mobile agents"
[13] Drakontas LLC, "Wireless Collaboration for First Responders," A White Paper on Wireless Networking and Intelligent Mobile
Agents, 2005 [14] W W. Godfrey, and S.B. Nair, "Bio-inspired technique for servicing networked Robots (Accepted)," International Journal of Rapid
Manufacturing, Inderscience,2011
Description of drawings
Figure 1: Some typical ways of connecting devices within IoT
Figure 2: Inter-connecting IoTs via the Internet
Figure 3: The components of the off-board Mobile Agent Framework (MAF) running on a dedicated node or a computer; the device communicates with it via a communication link
Figure 4: The on-board version of the MAF embedded on a device/appliance
Figure 5: Flowchart of a SƒR agent to find a "Free" device
Figure 6: Flowchart of a PaS agent to find an "Allocated" device
Figure 7: Change of State for a device after "Execute" state
Background of Invention
Research communities have made several efforts to create networks of communicating and uniquely identifiable devices so as to develop sophisticated applications [1, 2, 3]. Device to device communication has evolved into different areas of research viz. M2M communication [5], sensor networks [5], Cyber-physical systems [6] and the Internet of Things (IoT) [7]. Atzori et al. [7] have highlighted different visions for realizing IoTs, the key issues involved and the inherent challenges and their real-world applications. The current state of an Internet of Things deals more on the identification of different devices using RFID techniques, Object Naming Service and Electronic Product Code global discovery, to name a few. Research on the IoT also deals with defining architectures to support various real-time scenarios which include supply chain management, shop-floor integration, facility management and logistics. Choi and
Yoo [8] address the issue of communication among devices over the Internet using message PaSsing techniques. Rodriguez and Favela [9] describe the use of autonomous agents that proxy for users and provide the required services in their absence. A prototype that employs mobile agents to be used in conjunction with mobile devices has been cited in [10, 11]. Mobile agents have been deployed to realize distributed applications [12]. Drakontas [13] has suggested the use of mobile agents to achieve communication and non-disruptive interaction between devices. The paper endorses the use of mobile intelligent agents that can adapt based on the knowledge gained from the environment which in turn alters the state of the system. Drakontas discusses the use of mobile agents on mobile ad-hoc networks and suggests possible solutions for problems faced within, such as security, congestion and prioritizing information flow.
In our invention we present how mobile agents can be incorporated in an IoT using a Mobile Agent Framework hereafter referred to as MAF which facilitates communication and interaction between heterogeneous devices culminating in the search for a resource or a service required by a device.
Objective of the Invention:
The objective of this invention is to aid the realization of an IoT wherein devices forming the network can communicate and exchange data, information and programs (code) co-operatively and autonomously and eventually evolve and exhibit intelligence. The basic means for such operations viz. communication, exchange of data, information and programs or code, is done using mobile agents that can carry data, programs and intelligence. This invention provides a software framework which when installed within each of the heterogeneous devices populating an IoT will aid in their co-existence and also present itself as a generic mechanism for the development and deployment of applications to be embedded on these devices. The framework will thus also support flow of information, services, programs, etc. in a seamless and unobtrusive manner.
Novelty of this Invention:
i The main novelty of this invention is that it highlights the full application of all the vital features of a set of mobile agents viz. mobility, cloning, payload (data, programs, rules etc.), autonomy, co-operation, etc. Such agents can thus migrate from one networked device to another while the device is working or idling, non-intrusively and yet provide or take away information and other services. This invention can also cater to heterogeneous mobile agents. In addition the invention allows for scalability in terms of the number of devices populating the IoT and also allows for deployment of agents that can adapt and learn.
Summary of the Invention
The invention described herein is a mobile agent framework that can facilitate inter-device communication, interaction and learning within an IoT. The objective of the invention is to aid the realization of an IoT wherein devices forming the network can communicate and exchange data, information and programs (code) co-operatively and autonomously and eventually evolve and exhibit intelligence. The invention makes use of mobile agents to delegate the required services to the devices in order to co-exist symbiotically. Two types of mobile agents have been used in the invention viz. Search-for-Resource (SƒR) agent and Provide-a-Service (PaS) agent. The former is capable of searching and finding a resource or a free node or device capable of executing a required service. A mobile agent that carries the source code or a program or a set of rules or an application that can effect the desired service, as its payload, constitutes the PaS type. The invention considers a device to be in any of the following three states: (a) Free: The device in this state can engage itself in performing the services it is capable of, (b) Allocated: In this state the device has been allocated a service which it is capable of performing and (c) Execute: In this state the device executes the service related
program or code or rules and completes the service request. After execution of the same it returns to the free state. The invention also mentions how existing and new devices could host the MAF.
Detailed description of the preferred embodiment
This invention comprises a Mobile Agent Framework, hereinafter referred to as MAF, which is deployed on the devices constituting the IoT. The IoT could also be connected to a backbone network to interact with other such IoTs as shown in Figure 1 and 2 facilitating inter-IoT communication and interaction. The MAF acts as a hosting ground for mobile agents facilitating both entry and exit to/from a device or node. In addition, it also serves to interact with services and resources available or connected to the device. It also aids in agent-agent and inter-device communication and interaction. The MAF thus supports all mobile agent based functions such as agent creation, migration, cloning, termination, execution, entry, exit, inter-agent interactions, payload-carrying, etc. and allows the programmer to develop applications using these mobile agents.
MAF based Devices
Figures 3 and 4 show how the MAF can manifest itself in devices that constitute the nodes of an IoT. In Figure 3, the upper block comprises the MAF which runs on a dedicated node or computer, connected to the IoT. Existing devices with communication capabilities can connect to such nodes to gain the services carried by the mobile agents that populate the network.
Figure 4 shows a similar architecture for the manufacture of new devices that need to conform to the IoT that uses this invention. Here the MAF is embedded in the device itself. This aids such devices to directly connect to the IoT so as to gain the services of the mobile agents and avoids the need of another node or device that hosts the MAF.
Figure 1 shows the manner in which these devices connect to the network that forms the IoT. It can be seen that some of the devices connect via the nodes running the MAF while the others have the MAF embedded within and hence are capable of consuming the services offered by the mobile agents, directly. Devices without an MAF can also use the services of these agents via the devices with an embedded MAF provided they have the necessary communication capabilities. This figure also depicts that the communication channels used by the devices could be either wired or wireless. The wireless channel may include any of the available standards such as IEEE 802.11, ZigBee, Bluetooth, etc.
Role of Mobile Agents within the MAF
Mobile agents created within the MAF can migrate from one node or device to another in an intelligent manner by avoiding recently visited nodes or devices forming the IoT or because they are attracted in some manner towards the node or device requiring a mobile agent's service. The embedded migration strategy within a mobile agent makes it patrol the network. These agents may also clone based on need to effect a parallel search within the IoT. Mobile agents carrying services that may be required by the devices constituting the IoT could be released onto the IoT by third party developers or manufacturers of these devices. These agents allow the devices to PaSs information to other devices, retrieve information and also signal the requirement of a service available in some other device within the IoT. The context of a service may change based on the type of device or appliance requiring or assisting in a service. A device may require a service that can be rendered by another. Also, a device may be capable of physical actuation but may not possess the correct program or code to actually perform an assigned task which will result in satisfying the desired service. Mobile agents in this architecture facilitate a variety of such scenarios/applications to aid in searching and providing services. A few such applications are described later depicting the use of such mobile agents as service searchers and providers within an IoT. The Pheromone-Conscientious (P-C) based
mobile agent migration mechanism cited in [14] can be used in conjunction with the MAF for efficient patrolling and servicing of the devices forming the IoT.
Types of Mobile Agents within the MAF and their Functions
There are two types of agents populating the MAF based IoT architecture viz. Search-for-Resource agent and Provide-a-Service agent hereinafter referred to as SƒR and PaS agents respectively. The former is capable of searching and finding a Resource or a free node or device capable of executing a required service. A Resource thus can perform the requested or required service. A mobile agent that carries the source code or a program or a set of rules or an application that can effect the desired service, as their payloads, constitute the PaS type. Mobile agents within the MAF could also be a combination of these two types and be capable of performing all functions of both SƒR and PaS agents. Both SƒR and PaS agents move within the devices/nodes of an IoT in a manner as cited in [14].
(a) SƒR Agents:
A node or device in need of a service may spawn an SƒR agent to search for a specific Resource that can perform the requested service. Initially all Resources within the IoT are in the "free" state. This state indicates that the Resource is free and willing to engage itself in performing the services it is capable of. For instance a robotic arm within the IoT could be a Resource capable of picking and placing an object. A mobile phone could be a Resource that is capable of conveying a message to a user. Likewise, a vacuum cleaner is also a Resource that can perform a cleaning service.
The SƒR mobile agents spawned by a node or device requesting a specific service, migrate across devices forming the IoT, looking for such free devices or Resources. If they find any, depending on their capability and demand, the task to be performed is conveyed to the device or Resource.
Free Resources diffuse pheromones onto the neighbouring devices so as to indicate that they are currently free and willing to serve. These pheromones could further diffuse onto the devices of the IoT and thus tend to attract the SƒR type agents. When an otherwise conscientiously migrating and Resource-seeking SƒR agent hits the pheromone trail it travels to the relevant free Resource. The state of this Resource that can provide the desired service now changes to "Allocated". This completes the task of the SƒR agent but does not necessarily mean that the service for the requestor device/node is performed'. Figure 5 depicts the flow of the SƒR agents.
(b) PaS Agents:
Now that the SƒR agent has discovered a Resource that can satiate the service request, it is possible that this Resource may not have the concerned program, code, information, rules or intelligence required to execute the desired service. If it does not possess the same, then the Resource diffuses pheromones across its neighbours and thus to other devices of the IoT so as to attract the relevant PaS agent. PaS agents populate the IoT and use the same migration and search mechanism as that of the SƒR agents.
Once the PaS agent carrying the relevant program, code, information, rules or intelligence required to execute the desired service, reaches the Allocated Resource, the latter executes the same, thereby effecting the service request and thus ending a request cycle. Before commencing the execution, the Resource changes its state from "Allocated" to "Execute" but after the service is rendered the Resource status jumps back to "Free". The flow of execution of a PaS agent is depicted in Figures 6 and 7.
With several SƒR and PaS agents released into the IoT to cater to service requests from either specific or generic devices, the development of new programs for older and newer services can be performed in a distributed manner. A human user need not go through the
arduous task of programming the concerned device or gadget since the same may be available within one of the mobile agents within the IoT. Instead the devices pro-actively pheromone as cited in [14] to attract PaS agents that carry the required program and make use of the same. Third party developers and manufacturers who provide the device can release PaS agents with programs that their respective devices can execute thus facilitating scalability and easy upgradation of available services within the loT. Devices then pheromone and attract the desired mobile agents to avail of these services. A device incapable of performing a service calls upon other devices within the IoT to service it, autonomously. The overall set-up thus forms a congregation of heterogeneous devices that can function in a distributed and autonomous manner with minimal user intervention,
Applications:
1. In the IoT, there could be a need for devices to co-operate and coordinate. For instance a parcel which is being couriered from its source to the destination may PaSs through several intermediate zones. A web service could facilitate the tracking of the same by the receiver or the sender. When a tracking request is received by the device hosting the web service, it could spawn a mobile agent that could search and discover a strategically located networked camera and actuate it so as to get a snapshot of the parcel. With the snapshot as its payload the agent could then go about searching for another device that could enable it to email this snapshot as an attachment to the intended receiver or sender so as to provide him/her with a view of the physical condition of the parcel along with other requested details. Cloning could occur when multiple paths (links) between the devices exist so as to hasten the search process. Cameras need not be static; they could be mounted on robots capable of wireless communication. In such a case one of the various robot programs carried as payloads by the agent could be executed by the robot so as to realize the desired task. The agent could provide the information on the parcel (tracking number, destination address, etc.) which could be used by the robot to locate the concerned parcel and then take the snapshot.
2 In real-world situations such mobile agents on an IoT can help provide emergency services too. If we imagine that all people possess a mobile phone (device) each of which contains information about its owner then in case of a medical emergency the patient (who could also be the user) can force his/her mobile phone to spawn an agent that would travel to other neighbouring devices (mobile phones) within the area and almost non-intrusively find a device carried by a medical doctor, alert him/her and guide him to the patient. If information such as symptoms, phone number, the GPS co-ordinates of the patient or his/her present address or location are carried by such an agent, it could greatly hasten the process of medical aid reaching the patient. It may also search for a device within a nearby hospital or hotel and inform the concerned staff about this contingency. Here the patient (or user) of the device (mobile phone) receives the services on demand. The MAF can thus be deployed and used to cater to several different scenarios/applications. The same scenario/application can be extended to reach out to security or anti disaster services including the police and fire stations.
3. Consider a case when a person arrives inside a new building or place and requires a service. For example he needs to print a document. The person unfortunately does not know the exact location of the printer within the building. He thus issues a print command on his MAF based device which in turn spawns a mobile agent. This agent now searches for devices in the neighbourhood or vicinity where a free printer (Resource) is available and transfers the job to that printer, after which it informs the person of the exact location where the same can be collected. The person can then go to the location and collect the printed document. Alternatively if a robot or a human being is manning the printer, the mobile agent can guide them to the person and hand over the same to him. In addition to this the MAF can also allow printers to request for paper autonomously from another device (such as a robot) which could eventually find a ream of paper and load the printed document onto the printer.
4 In case an ambulance is required to be summoned to carry an accident victim on the road, a by-stander could make the MAF on his/her mobile device to spawn a mobile agent which could route through the various other such mobile devices possessed by people or vehicle drivers on the road to eventually discover a device situated within an ambulance (such as the ambulance driver's mobile phone) and then notify its driver of the casualty and the location of the same. As the ambulance starts to move towards the area where the victim is, the same agent could knit back much faster than the ambulance through MAF based devices possessed by drivers of the vehicles moving along the return path and inform them, in advance, to make way for the ambulance. In this way such mobile agents moving through MAF based devices can secure a fast and open path for emergency services.
Glossary of Terms :
1. Device : Any electronic equipment such as Laptops, Personal computers, smart phones, robots, home or office or industrial appliances, etc. capable of connecting to an electronic network. A device needs to host the Mobile Agent Framework to act as a node of the IoT.
2. Service: Any program or code or rule or set of rules or application program required by a device to execute a task.
3. Resource: A Resource is a device connected to an IoT which can accomplish tasks for which it has been designed. E.g. a printer which can print, a robot could pick and place an object, etc.
4. Program: Any software or code or a rule or a set of rules written to perform some particular operation. These programs are generally carried as payloads by the mobile agents populating the IoT.
5. Pheromones: Messages that a device which hosts the MAF and forms a node of the IoT, emanates to attract mobile agents towards it.
5. CLAIMS
13. We claim a system allowing diverse heterogeneous devices to co-exist, communicate, interact and work co-operatively as an Internet of Things (IoT) comprising -
a. A Mobile Agent Framework or MAF running on every device, which can host static and mobile agents and is capable of sending and receiving mobile programs and messages including mobile agents.
b. Devices which need to either collaborate or co-operate or on top of whom, sophisticated composite applications could be written.
c. Two types of mobile agents termed the Provide a Service (PaS) Agent and the Search for Resource (SƒR) Agent.
d. Static agents which aid in book keeping information and augment the device to device interaction mechanism.
e. Background communication network such as wireless or wired network, LAN or WAN.
14. We claim that the MAF based IoT mentioned in claim 1 to be scalable by allowing new devices to be added to it.
15. We claim that the MAF based IoT mentioned in claim 1 could either be embedded on the device by the manufacturer or facilitated by running it on another device.
16. We claim that the MAF mentioned in claim 1 can support the deployment of multiple types of mobile agents having various capabilities.
17. We claim the use of all the vital features of a set of mobile agents viz. mobility, cloning, payload (data, programs, rules, etc.), autonomy, co-operation, etc. in realizing an MAF based IoT mentioned in claim 1.
18. We claim that the MAF based IoT mentioned in claim 1 allows the devices to autonomously request a resource and also provide services.
19. We claim that the MAF based IoT mentioned in claim 1 allows the programs carried as the mobile agent payload, to adapt and evolve, thus facilitating the realization of an evolving and intelligent IoT.
20. We claim that the MAF based IoT mentioned in claim 1 can make use of the PherCon mobile agent migration mechanism [14] to fulfill device requests and to search for Resources.
21. We claim that the MAF based IoT mentioned in claim 1 facilitates device to device cooperation and coordination in an asynchronous and autonomous manner.
22. We claim that the mobile agents in the MAF based IoT mentioned in claim 1 can communicate and interact with each other and perform the task of searching for Resources and also providing services to the various devices that form the IoT.
23. We claim that the devices forming an IoT using the MAF mentioned in claim 1 can spawn a mobile agent which can clone if required to fulfill the activities mentioned in claim 6.
24. We claim that the MAF based IoT mentioned in claim 1 facilitates a set of devices forming an IoT to interact and also perform activities mentioned in previous claims, with those forming another IoT via the Internet using mobile agents.

ABSTRACT

The invention described herein is a mobile agent framework that can facilitate inter-device communication, interaction and learning within an IoT. The objective of the invention is to aid the realization of an IoT wherein devices forming the network can communicate and exchange data, information and programs (code) co-operatively and autonomously and eventually evolve and exhibit intelligence. The invention makes use of mobile agents to delegate the required services to the devices in order to co-exist symbiotically. Two types of mobile agents have been used in the invention viz. Search-for-Resource (SƒR) agent and Provide-a-Service (PaS) agent. The former is capable of searching and finding a resource or a free node or device capable of executing a required service. A mobile agent that carries the source code or a program or a set of rules or an application that can effect the desired service, as its payload, constitutes the PaS type. The invention considers a device to be in any of the following three states: (a) Free: The device in this state can engage itself in performing the services it is capable of, (b) Allocated: In this state the device has been allocated a service which it is capable of performing and (c) Execute: In this state the device executes the service related program or code or rules and completes the service request. After execution of the same it returns to the Free State. The invention also mentions how existing and new devices could host the MAF.