A method includes determining a number of drops of a plurality of messages sent to a first node of a plurality of nodes within a mesh network. Based at least in part on the number of drops of the plurality of messages exceeding a threshold number of drops for a time period, decrementing a first rating assigned to the first node to a second rating assigned to the first node. Based at least in part on the second rating being below a rating threshold, determining that the first node is a potentially malicious node. Based at least in part on a first distance to the first node being larger than a distance threshold, identifying that the first node is a malicious node. The method may further include ending communications with the first node.