This paper is based on the results of the ‘GNSS Support Task’ study for the European Commission, DGXIII. It summarises the results of the cost benefit analysis in terms of coverage, accuracy and safety requirements for different types of user and describes the most cost-effective GNSS 2 architecture. These analyses also assume that the overall system is layered into wide area, regional and local systems. The future planning of satellite navigation is essentially driven by the wide area requirements and ensuring that these are global and seamless. There is some flexibility in coverage and accuracy for a wide area system in that it can be augmented regionally or locally if required, but it must provide the highest level of safety required. The paper approaches the architecture for a future navigation system from this safety aspect. An analysis of the chosen architecture shows that the required safety performance can be met. An implementation plan is described which allows a gradual evolution from the first system to be realized for safety critical operations to a fully civilian owned and operated system.