Making the smart grid a reality will entail communication capabilities that do not exist in current power grids and their legacy control systems. New grid elements such as distributed energy generation and PHEVs will necessitate changes in these existing communication networks. These changes will be dictated by the need for new control operations that will allow efficient management of these new elements and that cannot be supported by the existing communication infra structure. Moreover, the smart grid will let households man age energy usage more efficiently through AMI and HEMSs. The realization of these systems will require pushing communication networks far out from the network core and deep into homes and buildings. We have presented the communication requirements of the smart grid, and we have identified the main technical challenges that need to be tackled. The only viable way to fulfill these requirements is to design a new communication architecture that can support smart grid services and control operations. This future communication architecture will need to take advantage of the recent progress made in communication technologies and protocols. It will require the introduction of communications technologies that were not considered in the past. In addition, this communication architecture will need to be reliable, scalable, and extendable to future smart grid services and applications.