Probabilistic Congestion of Wireless Sensor Networks: a Coloured Petri Net based Approach

Abstract

Analysing probability properties on Coloured Petri Nets (CPNs) model is one of a favorite topic on system verification recently. This paper focuses on verifying congestion probability on Wireless Sensor Networks (WSNs) which is modelled by CPN. Actually, WSNs are the collection of sensors. A WSN topology is formed by the interaction among sensors via Wi-Fi connections. However, sensors can be consider as unsteady devices when working in the harsh environment due to limited processing capacity, non-replacement battery, etc. Hence, each sensor needs to attach a reliable probability so that users can know the probability of reaching the sink of data. Such probabilities are added into the transitions in our CPN probability model before checking congestion. Whole verifying process introduces also in order to emphasize the purpose of this paper via a straight example.

A new generalized stochastic Petri net modeling for energy‐harvesting‐wireless sensor network assessment

“…To find such a compromise, these systems need to be evaluated before their actual implementation. Two approaches can be used to analyze a WSN behavior 8 . The first one uses simulation software, whereas the second resorts to formal modeling.…”

“…For example, when a researcher introduces a new WSN protocol of communication, he can conduct his assessment according to both approaches. However, by opting for the first method, the researcher must encode the new algorithms in the language of the simulation framework 8,9 . Therefore, coding difficulties and long simulation time issues must be addressed.…”

“…In addition to that, the researcher must be aware that the simulation environment can significantly affect the obtained results. However, the simulation provides platform‐dependent results 8 …”

“…Two approaches can be used to analyze a WSN behavior. 8 The first one uses simulation software, whereas the second resorts to formal modeling. Despite being related, these approaches have quite different requirements.…”

“…However, by opting for the first method, the researcher must encode the new algorithms in the language of the simulation framework. 8,9 Therefore, coding difficulties and long simulation time issues must be addressed. In addition to that, the researcher must be aware that the simulation environment can significantly affect the obtained results.…”