Technique for communication with mobile data processing devices by way of mobile software agents

1. Field of the Invention

The present invention describes a technique for communication with mobile data processing devices. More particularly, the present invention describes a technique for communicating with mobile devices, such as chipcards or mobile phones, via mobile software agents.

2. Description of the Related Art

Software agents, or also so-called intelligent agents, are becoming more and more important as the networking of systems becomes more widespread. Software agents are software modules located in a complex environment. The software agent has a sensor input, is able to change its environment, and has a set of goals which it wishes to attain by means of its actions. At present there are a large number of software agents with a wide variety of functions, such as interface agents, information agents, delinerative agents, reactive agents and mobile agents.

Of particular importance for the present invention is the mobile agent. Mobile agents are capable of roving around WANs (for example the World Wide Web). Usually the property of mobility is immediately associated with the term “agent” anyway.

This is not the case, however, because they interact with third-party systems (hosts), collect information and return the information to the user.

In addition to its autonomy, the mobile agent also has the functional property of cooperation capability. The advantage of the mobile software agent is that, by suitable programming, communication costs and times can be saved.

Whereas in the normal client/server sector all communication between the participants is over the network, when mobile software agents are used an agent can be sent over the network to handle the entire communication process locally, which means the computer's user does not have to be permanently online for the entire runtime.

Since mobile software agents rove between networks, software developers are primarily concerned with the problem of security. The issues of authentication (mutual knowledge of identity), authorization (right to use the desired function) and all problems connected with electronic funds transfer (ability/willingness of the agent to make payments; limitation of liability by users; performance guarantees) are key elements.

In the field of mobile software agents there is a state of the art which is concerned with migration, splitting and merging of agents. However, it is usually assumed that the mobile agents begin a task when they arrive at a system and return a result when they have completed that task.

No state of the art is known which describes a mechanism which in a particular way enables communication between the mobile agent and mobile data processing devices.

The approaches applied to date specify software running on terminal equipment which processes data received from backend systems and which can be updated by means of code update mechanisms. But it is not possible for code to migrate to terminals autonomously for specific tasks and interact at the destination with mobile devices.

It is therefore the object of the present invention to deliver a system and method which permits communication with mobile data processing devices by way of mobile software agents in a simple manner.

By additional implementation of a runtime environment (interface component) for distributed mobile software agents for administration of mobile data processing devices, e.g. chipcards, on every node of the network (client), administration of the mobile data processing devices in the network can be controlled from the backend system (server). The method adapts automatically to the network structure. Mobile software agents introduced into the system by way of the backend system automatically spread through the part of the system on which the runtime environment (“mobile software agent” interface component) is installed.

One example of application of the concept is a method for administration of chipcards in a complex networked system with the aid of event-controlled mobile software agents operating on a suitable infrastructure. Agents sent out into the network from backend systems spread according to predefined strategies across the network and respond to chipcards which are inserted into devices in which they have located themselves. On certain chipcards the local mobile software agent performs actions such as downloading a new application, removing an application or invalidating the entire card, where appropriate in cooperation with other agents in the network or with backend systems.

A further example of application of the concept is management of Pervasive Computing devices. Agents sent out into the network from backend systems spread according to predefined strategies across the network and respond to devices which enter the action range of the nodes on which agents are located and begin a communication. On certain PVC devices the local agent performs actions, where appropriate in cooperation with other agents in the network or with backend systems.

On the nodes of the network agent runtime environments (“mobile software agent” interface components) are preinstalled which enable the migration of administration agents onto the said nodes. The runtime environments provide the agents with the key basic functions, such as activation when mobile software agents appear, communication with cooperating software agents, communication with mobile software agents and communication with backend systems. The special feature of the runtime environment is that it includes a mechanism permitting it to inform the local agents of specific events, such as insertion or removal of a chipcard or the introduction or withdrawal of a mobile phone into or out of the action range. An agent can spread itself across several nodes by means of several migration and division processes, and then becomes inactive. Only when a specific event occurs, such as insertion or removal of a card, does the agent become active and begin its task.

The present invention will now be described with reference to the following drawings, wherein like reference numbers denote the same element throughout.

FIG. 1 shows the network architecture underlying the invention, comprising several backend systems, connecting nodes and terminal devices with chipcard readers;

FIG. 2 shows the processing of a mobile software agent in ccordance with the invention in an architecture in accordance with FIG. 1 ;

FIG. 3 shows the functional components of the mobile software agent in accordance with the invention;

FIG. 4 shows the interface component for communication between the mobile software agent and the mobile data processing device;

FIG. 5 shows an authentication architecture for mobile software agents; and

FIG. 6 shows a further authentication architecture for mobile software agents.

FIG. 1 shows a communication architecture underlying the present invention. The communication architecture comprises several backend systems 100 and several connecting nodes 102 by way of which communication with respective terminal devices 104 is routed. In the preferred embodiment of the invention, the terminal device 104 is a data processing device to which a chipcard reader is connected.

Mobile software agents are introduced into the chipcard system from the backend systems 100 over the network. They spread automatically from the backend systems 100 via the connecting nodes 102 to the terminal devices 104 on which administrative tasks, such as downloading new applications onto cards, changing expiry dates or invalidating cards, can actually be carried out. By assigning a life span to the agents, it can be ensured that a generation of agents is automatically deleted after a certain period of time. Preferably, the generation of agents deletes itself.

FIG. 2 shows the spread of mobile software agents in the communication architecture of FIG. 1 .

The basic mode of functioning of a mobile software agent is as follows. Originating from one of the backend systems 100 , a mobile software agent A roves through the network and in the process is able to duplicate parts of itself and leave the parts behind on suitable nodes. Depending on the type of node on which they are located, agents can behave differently. An agent on a backend system 100 , for example, can wait for messages from agents in the network and answer queries contained in the incoming messages by accessing a database. Agents on the connecting nodes 102 could serve as caches or passthroughs. Agents on terminal devices 104 with chipcard readers could wait for insertion of a card 106 , read information from the card and send it to the backend system 100 from which they originate, receive and carry out instructions from the backend system 100 for administration of the card, and finally inform the backend system 100 of the result.

FIG. 3 shows the functional components of a mobile software agent 110 in accordance with the invention. The mobile software agent 110 includes an event-handling component 112 . Using the event-handling component 112 , the mobile software agent responds to certain events of which it is informed by the runtime environment, such as the introduction of a mobile device into the action range or the insertion or removal of a chipcard into or out of a chipcard reader. The agent 110 also includes a node recognition component 114 . The component 114 classifies network nodes. When the agent 110 migrates to a node, it determines whether the node is of a type on which it is capable of existing and carrying out its tasks. When the agent 110 has classified a node as suitable for it and has been accepted by the node, an installation/termination component 116 is utilized to install the agent 110 on the node. Further, when the agent 110 migrates to a node which already contains a predecessor agent now rendered obsolete by the migrating agent 110 , the component 116 terminates the predecessor.

A device classification component 118 is also part of the agent 110 . When a mobile device, such as a chipcard, appears in the action range of the agent 110 , the agent 110 classifies the device. Device action components 120 include logic to be carried out relative to the devices. Depending on the result of the classification, the agent 110 performs actions such as invalidating a card or installing an additional application on the chipcard. After a successful or failed action, a checkback component 122 may be employed to checkback to other agents or to a backend system.

A migration strategy component 124 contains information relating to the destination for the agent 110 and/or where copies of the agent 110 should be replicated. In order to reach their destination, such as a single node or an area of a network, agents 110 migrate from node to node and/or replicate themselves on one or more nodes, creating a copy of the agent 110 on such nodes.

A merger component 126 helps to avoid wasting resources. Agents with common or similar tasks can be merged into one agent. Agents may have a specific life span at the end of which they must terminate themselves in order to make the resources taken up by them available to new agents. A self-termination component 128 handles this function.

FIG. 4 shows an interface component (runtime environment) 138 for communication between mobile software agents 110 and a mobile data processing device (not shown).

To enable the mobile software agents to execute the functions mentioned in connection with FIG. 3 , an infrastructure which provides the agents with the necessary support is required. This infrastructure may be a network on the nodes of which an interface component is installed. A node must offer the certain interface components, whereby running of the components is dependent on the rights of the creator of the agent on the respective node. A support function component 140 handles a number of functions. These include reception of agents from the network, e.g., via a specific TCP/IP port, validation of the agents, e.g., by verification of a digital signature which guarantees the integrity of the agent and by byte code verification, running of agents on Java platforms, e.g., by executing the agent's byte code, remote monitoring of the agents running on the node, etc. (Java is a trademark of Sun Microsystems, Inc.)

A node capabilities component 142 provides agents with the information they need to make migration or installation decisions. Via a notification component 144 , agents can have themselves notified of certain events, such as when chipcards are inserted into a chipcard reader. A mobile device access component 146 permits communication with mobile devices, such as chipcards. For chipcards, this may be carried out by means of the OpenCard Framework.

A key management component 148 permits management of keys in the system. It enables agents to import encrypted keys they carry with them safely into local security modules which may be installed in terminal devices.

FIG. 5 shows an authentication architecture for mobile software agents 110 . The nodes are configured such that only secure mobile software agents 110 , i.e., agents digitally validated by a backend system 100 , are executed. Nodes and backend systems only accept messages from other nodes or backend systems which have a digital signature. This means it is not possible to communicate with the backend system 100 from a node outside the system in order to utilize the keys located there to create a signature which would enable actions to be carried out against a mobile device.

The software agent 110 can be authenticated to a mobile device 200 as follows. The software agent 110 receives from the runtime environment 138 a notification that the mobile device 200 has been activated. The software agent 110 requests a random number from the mobile device 200 , which may be a chipcard. The software agent 110 sends the random number to a backend system 100 from which it itself originates. The backend system 100 signs the random number with a key 204 and sends the signed random number by way of the software agent 110 and the runtime environment 138 back to the mobile device 200 . The mobile device 200 verifies the signature and only accepts the actions of the software agent 110 if it is correct.

This embodiment of the authentication architecture is based on the fact that only the mobile device 200 and the backend system 100 have a key.

Another embodiment of the authentication architecture may involve a key 204 being stored only on the mobile device 200 and on the respective node. This variant is represented in FIG. 6 . The authentication can then occur as follows. The mobile software agent 198 receives from the runtime environment 138 a notification that a mobile device 200 has been activated. The software agent 110 requests a random number from the mobile device 200 . The software agent 110 has the runtime environment 138 sign the random number with a key 204 and sends the signed random number to the mobile device 200 . The mobile device 200 verifies the signature and only accepts the actions of the software agent 110 if it is correct.

In the following, two preferred examples of the application of mobile software agents in accordance with the present invention are presented.

Invalidation of Cards

Cards can be invalidated in the following manner. The backend system which wants to invalidate certain cards creates its own agent and sends it out into the network. The agent spreads by migration and replication according to a strategy defined by the backend system over the areas of the network in which the card being invalidated may appear. On nodes already containing other, compatible agents likewise tasked with invalidating cards, the agent merges with them, whereby the resultant agent is tasked with invalidating the combined set of chipcards, including those to be invalidated by the original agents. One of the cards to be invalidated is inserted into a terminal on which one of the created agents is located. The agent detects the card and invalidates it. The agent notifies the backend system from which it originates as well as its associated agents. All notified agents remove the now invalidated card from their lists of cards awaiting invalidation. At the end of its life or on completion of its tasks, the agent terminates itself.

Installation of New Applications

New applications can be installed on chipcards by agents in the following manner. The backend system which wants to install an application creates its own agent and sends it out into the network. The agent spreads by migration and replication according to a strategy defined by the backend system over the areas of the network in which the card on which the application is to be installed may appear. On nodes already containing other, compatible agents likewise tasked with installing the same application, the agent merges with them, whereby the resultant agent is tasked with installing the application on the combined set of chipcards, including those on which the original agents were tasked to install it.

One of the cards is inserted into a terminal on which one of the created agents is located. The agent detects the card and installs the application. It may need to communicate with the backend system 100 from which it originates to do so, for example, if the necessary card commands can only be assigned the correct signatures or Message Authentication Codes (MACs) there. The agent notifies the backend system from which it originates as well as its associated agents. All notified agents remove the card from their lists of cards awaiting installation of the application concerned. At the end of its life or on completion of its tasks, the agent terminates itself.

When the runtime environment for distributed agents to administer chipcards has been installed on the nodes of the network, the administration of chipcards in the network can be controlled from the backend system. The method adapts automatically to the network structure. Mobile software agents introduced into the system automatically spread through the part of the system on which the runtime environment is installed.

The invention also permits the deployment of agents in environments with limited resources. Since key, frequently needed functions are provided to the agents from an Agent Framework on the nodes, the agents themselves can be very small. In addition, the merging of agents allows the number of agents in the system to be limited and thus conserves system resources.

The method can also be used to distribute off-card applications in the network. A banking application permitting cash withdrawals from ATMs could be represented as an agent which migrates only to terminals with a cash dispensing function, and based on the insertion of EC cards/credit cards dispenses cash instead of performing an administrative function.

The method adapts automatically to the network structure. Mobile software agents introduced into the system by way of the backend system automatically spread through the part of the system on which the mobile software agent interface component is installed.

While the present invention has been described relative to a specific preferred embodiment thereof, various modifications and changes may be obvious to one skilled in the art, and it is intended that the present invention encompass such changes and modifications as fall within the scope of the appended claims.