The Smart Grid Infrastructure (SGI) has emerged as a necessary and critical platform for provisioning intelligent and accurate services to consumers of the electric grid, in recent times. With the emergence of this infrastructure and accompanying technologies, the need for securing the same from malicious attempts by the adversary class to disrupt routine operations, cannot be understated. A standard SGI may consist of disparate and heterogeneous devices, cooperating and exchanging customer-specific data (readings), obtained from smart meters. Some of the devices connected to the SGI, such as sensors and actuators, are resource-constrained in nature. In addition, an omnipresent threat to the SGI, from the adversary class in cyberspace, does indeed exist. In this paper, a brief description of various types of malicious attacks against SGI operations, is presented. In addition, a light-weighted pattern matching technique for detecting such attacks, is discussed. The proposed scheme is capable of detecting anomalous device behavior at various levels of the SGI hierarchy, at the same time imposing minimal overhead in terms of communication and storage needed.