Smart meter collects the electrical energy consumption in short time intervals and report this information to the utility. Such high resolution data can be used by the utility in the purpose of monitoring, billing and efficient scheduling. However, this information can be used to make inference about what each single consumer does in any specific time interval, and thereby it is natural to ask about the privacy, specially of the residential consumers.
There is a tradeoff between the accuracy of the transmitted data and the user privacy. We look at this problem in an abstract way, and use the rate-distortion theory to precisely quantify the tradeoff between the accuracy of reported data (mean square distortion) and privacy (information leakage) for our proposed model. We show that the privacy-utility tradeoffs on the total load are achievable using an interference-aware reverse water-filling solution, which intuitively translates to suppressing low energy components.