Casimiro, AntónioMarques, Luís2012-09-172014-11-142012-09-172014-11-142012-09-174th International Workshop on Dependable Network Computing and Mobile Systems (DNCMS 2011), in Proceedings of the 30th IEEE International Symposium on Reliable Distributed Systems Workshops, pages 26-35, Madrid, Spain, October 2011.http://hdl.handle.net/10451/14145http://repositorio.ul.pt/handle/10455/6887Achieving dependable and real-time operation in Wireless Sensor Networks (WSNs) is a hard and open problem. This can be an obstacle for many applications, namely in the automotive and medical domains, particularly if safety-critical control is envisaged. To overcome the communication uncertainties that are intrinsic to wireless and dynamic environments, a generic approach is to constantly adapt to environment conditions. This requires appropriate solutions to characterize such conditions. This paper contributes with a lightweight solution for a dependable characterization of network QoS metrics, which is appropriate to support dependable adaptation in WSNs. The proposed solution offers probabilistic guarantees, building on non-parametric stochastic analysis to achieve fast and effective results. The paper also provides an evaluation of the solution.engadaptationnon-parametricdependabilitylightweightQoSWireless Sensor NetworksLightweight Dependable Adaptation for Wireless Sensor Networksreport