Targeted data breaches and cybersecurity attacks involving IoT devices are becoming ever more concerning. To combat these threats and risks, the IETF standardized Manufacturer Usage Description (MUD), which allows IoT device vendors to specify the intended communication patterns (MUD profile) of an IoT device. MUD profile enables validation of the actual communication pattern of an IoT device with the intended behavior at run-time. However, the MUD specification was primarily intended for enforcement at the Local Area Network (LAN) of the IoT device, thus fragmenting the solution across multiple heterogeneous networks. MUD enforcement at higher levels in the network hierarchy (e.g., private edge for enterprise networks) eases security policy management and reduces processing overheads on the existing security infrastructure. To realize MUD enforcement at the edge, there are mainly two challenges: (1) How to identify an IoT device at the edge so that enforcing device-specific MUD profile on the IoT traffic is possible. (2) How to scale MUD enforcement to a large network of IoT devices. In this paper, we present our approach to address these challenges and validate IoT device communication at the edge. In order to scale MUD enforcement to a large IoT network, we leverage multi-stage pipeline architecture and stateful ALUs of P4 programmable switch and process IoT traffic in the dataplane. © 2022 ACM.