1: \begin{abstract}

2: A distributed consensus algorithm for estimating the maximum value of the initial measurements in a sensor network with communication noise is proposed. In the absence of communication noise, max estimation can be done by updating the state value with the largest received measurements in every iteration at each sensor. In the presence of communication noise, however, the maximum estimate will incorrectly drift and the estimate at each sensor will diverge. As a result, a soft-max approximation together with a non-linear consensus algorithm is introduced herein. 

3: %A design parameter controls the trade-off between the soft-max error and the dynamic range of the state value. The error is small when the dynamic range is large, making linear approaches such as amplify and forward infeasible. Therefore, a non-linear consensus algorithm is used to control the transmit power.

4: A design parameter controls the trade-off between the soft-max error and convergence speed. An analysis of this trade-off gives a guideline towards how to choose the design parameter for the max estimate.

5: We also show that if some prior knowledge of the initial measurements is available, the consensus process can converge faster by using an optimal step size in the iterative algorithm. 

6: %A trade-off between the convergence speed and the estimation error is described.

7: A shifted non-linear bounded transmit function is also introduced for faster convergence when sensor nodes have some prior knowledge of the initial measurements.

8: %Finally, a consensus algorithm with different edge step sizes is proposed to improve the convergence speed when no prior knowledge of the initial measurements is given.

9: Simulation results corroborating the theory are also provided.

10: %It is shown that for some distributions of the initial measurements, a modified soft-min consensus can also be used to calculate the max. %

11: \end{abstract}

12: