1: \begin{abstract}

2:   This paper deals with the stabilization of linear systems with

3:   process noise under packet drops between the sensor and the

4:   controller. Our aim is to ensure exponential convergence of the

5:   second moment of the plant state to a given bound in finite

6:   time. Motivated by considerations about the efficient use of the

7:   available resources, we adopt an event-triggering approach to design

8:   the transmission policy. In our design, the sensor's decision to

9:   transmit or not the state to the controller is based on an online

10:   evaluation of the future satisfaction of the control objective. The

11:   resulting event-triggering policy is hence specifically tailored to

12:   the control objective. We formally establish that the proposed

13:   event-triggering policy meets the desired objective and quantify its

14:   efficiency by providing an upper bound on the fraction of expected

15:   number of transmissions in an infinite time interval.  Simulations

16:   for scalar and vector systems illustrate the results.

17: \end{abstract}

18: