AbstractIn this paper, we address the recurrent problem of node redundancy in heterogeneous Wireless Sensor Networks (WSNs). In addition to presenting a new analytical approach for estimating redundancy and unlike existing works, we propose a new solution where sensors are competing for the identification and relocation of redundant sensors. In our solution, which is mainly based on a comprehensive formulation of redundancy as a dynamic matter, sensors with low redundancy weights force neighboring peers with high redundancy weights to relocate or go to sleep according to a bully strategy. First simulations with the ns3 tool are showing promising results.