Collaboration Between Wireless Sensor Networks using Information Exchange

My PhD proposal.

Wireless sensor nodes are usually composed by a micro-controller and powered by batteries. Additionally, each node has at least one of four possible interfaces to interact with the external world, which can be buttons, USB ports, radio transmitters and sensors. There is a large variety of sensor types such as those that monitor temperature, relative humidity, water levels, light, carbon monoxide levels, or other environmental parameters.

Nodes equipped with radio transmitters may make use of network protocols and be organized into a Wireless Sensor Network (WSN). There, they are able to collect and transmit data from remote places to a central station. Large networks may be composed by heterogeneous devices with different hardware from distinct manufacturers. In order to make this integration feasible and achieve scalability in the WSNs, the IEEE standard 802.15.4, which specifies the physical layer and media access control, has been developed.

Wireless sensor nodes have different constraints other than those that traditional computers have–such as limited size and cost–which bring restrictions to their computational power and energy sources. Therefore, researchers from this field have been focusing mainly on new algorithms that explore their characteristics. The progresses may be done by improving their components, developing new MAC and transport protocols, or using data aggregation, data compression, collaboration and specific applications.


Some WSN applications enable nodes to take decisions based only on local data. For example, nodes that detect fire can match information about the CO level, temperature, ionization and photoelectric sensors in order to infer whether a fire is present or not. Moreover, nodes may share their knowledge or their resources with other nodes of the same network. This is called as collaboration and can be classified into different categories:

  • Cooperation – when nodes work cooperatively according to their own contribution degrees to the main goal of that network. For example, cooperatively sensing the environment.
  • Resource sharing – when resources must be shared between different nodes. Examples of resource sharing involve algorithms for processing data in a node equipped with a more powerful micro-controller, or transmitting data through the shortest path.
  • Self-organization – when the data sensed by different nodes is joined and used to react to the environment. For example, a set of nodes may reach a consensus state and detect holes in their coverage. Hence, they are able to update their behavior and cover the area of interest.

A point that has not being studied so far is how to build the collaboration between the WSNs. Nowadays, a high number of sensor networks have been deployed close to each other, either monitoring cities, forests, buildings, machines or military areas. Therefore, they measure semantically linked data, for example, temperature and humidity, but these relations have not been exploited to the maximum.

The Goal

The main goal of this work is to maximize the lifetime of the WSNs and the the main requirement is not compromise the quality of measurements. The way to do that is using data from external WSNs in a collaborative infrastructure, taking humans out of the loop. As a part of its responsibilities, the proposed system will compute–at runtime–sets of nodes that must be activated and their workload according to the changes in the environment. In other words, it will be capable of self-configuring and self-managing its WSNs according to the information sent by internal or external WSNs.



Leave a Reply

Your email address will not be published. Required fields are marked *