JADE Agents to Wireless Sensors: Easy Wireless Sensor Network Management

This particular paper describes our most recent work on our system called JAWS. This system is intended to control and monitor Wireless Sensor Network. In principle, JAWS is group of agents implemented on JADE platform. It consists of several agents that are able to communicate with Wireless Sensor Nodes and obtain values from particular sensor nodes called motes. We are also able to inject mobile code to each mote so we can change behaviour of that mote and in extension of the whole network. JAWS system uses services as a natural and most viable concept that helps us to control and monitor Wireless Sensor Network. We will briefly describe basic concepts of our system and describe services used for controlling the network. Concept of services is tightly bound to ability of the system to expose and use services. To expose and use services we created set of protocols. In this paper will be also described the most important protocols of our system. In the end of the paper we will describe some of the problems that we experienced using our system and we also describe some counter actions that we performed to troubleshoot some problems. Keywords-JAWS; JADE; Wireless Sensor Network; Service; Artificial Agent; Protocol. Wireless Sensor Networks (further WSN) are in the scope of researchers in the last years. It is because of their ability to measure various physical quantities in physical environment. The second reason is that their motes (nodes of network) are also able to communicate with each other via radio channel. By this two abilities, WSNs are mostly used for monitoring purposes in some environment. This is useful when we need to measure data with high density of measurements because WSNs are able to run with thousands of motes that are distributed within measured environment. For data gathering is used special node called Base Station (further BS) that can be connected to computer and that can deliver data sent by motes from network. WNSs are also likely used in environments that are hostile for human or where human presence is not suitable. One example for all is experiment on monitoring habitats on Great Duck Island[1]. In this experiment, the ducks were monitored on island and presence of humans could change animal behaviour and thus devalue whole experiment. There are various ways to control mote in WSN. Since our research is in area of artificial agents our systems are agent based. This approach has been proved viable in other agent based systems for controlling WSN (e.g. Agilla[2]). In our case we developed system called WSageNt[3] that is in principle similar to Agilla. It is agent based and it is a platform that runs on each mote. Agent is then understood as a small piece of code that is interpreted on platform. That is where the similarity ends. Agilla agent code is similar to assembler but our agents are written in Agent Low Level Language[4]. Since our ALLL agents can move among motes, ALLL language is designed to be minimalistic. It means that less bytes have to be sent via radio channel, which the is biggest consumer of the battery of a mote. As postulated in [5] sending 1 bit of information is equal interpreting about 1000 of instructions. We briefly describe WSageNt system in section II. We explained the system running on each mote. This system allows each mote in WSN to sense environment (via mote’s sensors) and it is able to talk to other motes in network. The only other thing we need now is BS that will collect data sent from motes. Motes can send data at any time so it seems useful to have a system that will gather this data from Base Station and that will process them in some, possibly intelligent, manner. Since “intelligent” agent runs on each mote, data can be processed directly on mote. For example sums of data or average values can be precomputed so less bytes will be transmitted over radio and some battery will be saved. Even though data can be preprocessed some of it arrives to the Base Station and in extension to some system that will process them further to obtain more complex information than data itself. And JAWS is exactly such a system. It can obtain data from WSN on the one side and it can send data or even ALLL agents to network on the other side. Since Journal of Network and Innovative Computing ISSN 2160-2174 Volume 1 (2013) pp. 260-269 © MIR Labs, www.mirlabs.net/jnic/index.html _______________________________________________________________________________________________ Dynamic Publishers, Inc., USA