Communication system has a key position in smart grid (SG). Smooth operation among different functional blocks of smart grid, i.e. generation, transmission, distribution, utility, and consumers require reliable and efficient exchange of information. Hybrid architecture in smart grid, in the form of Advanced Metering Infrastructure (AMI) consists of gateways/access points and smart meters. AMI network facilitates consumers to take benefits of real-time pricing, demand side management, load scheduling and upgrading the software updates through access points to multiple smart meters without visiting every place. Information sharing can take place in the shape of readings from meters to AMI, from meters to utility and from AMI to utility. This information exchange is either scheduled, on-demand or in the form of mass transfer. Information exchange in the form of time of use (TOU) and real-time pricing also occurs from utility to meters. Broadcasting is one possible solution in these scenarios for information exchange. Data broadcasting in AMI networks can be achieved using many wired and wireless communication technologies including Zigbee, Wireless LAN, Powerline Communication (PLC), WiMAX 802.16, Cellular, and Digital Subscriber Lines (DSL). But these technologies either have limited range or can be used for licensed users only. For these reasons, Cognitive Radio Networks (CRNs) are one possible solution that provides large data transmission utilizing available spectrum resources from licensed and unlicensed bands. Broadcasting is challenging in CRNs due to Licensed/Primary User (PU) activity, availability of multiple channels, and channel diversity. Therefore, PU protection must be provided at all cost and data transmission should be quick and reliable for a smooth operation. In this thesis, two novel reliable schemes are proposed, i.e., PRobability-based Broadcasting scheme for CR-based Smart Grid (PRB-CRSG) and Area-based Broadcasting scheme for CR-based Smart Grid (ARB-CRSG) specifically designed for AMI networks. These schemes provide reliable data to smart meters with sufficient PU protection. Both these schemes have been compared with two flooding techniques, i.e., "Key Policy Attribute Based Encryption" (KP-ABE) and "Key Management Scheme" (KMS). Simulation in NS-2 validates these schemes as a viable solution in cognitive radio-based smart grid systems.