This document provides a 3-step process for protecting a network peering edge: 1. Tag prefixes received from peers and customers uniquely in BGP and FIB tables. 2. Tag external packets at peering locations based on the longest prefix match to identify packets from peers and customers. 3. Classify packets via MQC based on the tags and apply policies such as dropping packets tagged as coming from peers to protect the network. Hardware support is needed for classification.