Wireless Power Transfer (WPT) services require a communication network to ensure that the power transfer from grid to vehicle is carried out in a precisely controlled manner such that it is fast, safe and most efficient. A comprehensive scalable WPT ecosystem will comprise of not only vehicles and charging controllers but will also include multiple value chain providers such as charge station operators, power generation and grid management companies and vehicle manufacturers etc. So a communication system should not only address the immediate first-hop communication between the vehicle and charging controller but also provide a robust end-to-end communication solution that addresses scalability, real-time response characteristics, in a cyber-secure manner. This paper summarizes the combined efforts of Clemson University and Cisco Systems related to the development of a communication system for stationary, quasi-dynamic and dynamic WPT under DOE FOA 0000667 that addresses these aspects. The implementation used Dedicated Short Range Communication (DSRC) technology as the wireless technology and applied the Fog computing framework as the base for the overall architecture. This approach provided 1) ultralow latency communication between the vehicle and the infrastructure required by WPT real-time control loop, and 2) edge analytics for vehicles and grid components and interaction with WPT cloud. The design of the communication system includes: 1) Identification of data characteristics for the charging session, 2) Implementation of wireless link based on DSRC, 3) Design of network/transfer protocol over WSMP used by DSRC radios and 4) Design of infrastructure and back end architecture based on Fog computing. In addition, this paper presents a discussion of the characteristics of this network architecture and the lessons learned during this project.