The development of multi-constellation, multi-frequency GNSS is ongoing, with the aim to enable a robust and reliable navigation and approach service to airspace users. While this will greatly reduce vulnerability to space weather, unintentional interference and constellation weakness, some residual vulnerabilities will remain. In the current, predominantly GPS L1 GNSS environment, aviation has accepted that alternate positioning, navigation and timing capabilities based on terrestrial systems remain necessary. These reversionary area navigation capabilities are based primarily on DME/DME, while still providing some residual VOR/DME services. However, this reversionary capability has not been demonstrated to support the stringent RNP requirements that GNSS can support. Also, DME is criticized as being spectrum inefficient, and aviation-internal and aviation-external pressures to share the DME band with other services are increasing significantly. A key question for the future evolution of Communication, Navigation and Surveillance systems is what type of a reversionary capability will be needed in the future (terrestrial or space based), and what performance levels it needs to provide. To answer this question, supported by specific technology options, a project under the SESAR Horizon 2020 Framework (PJ14-03-04) is working on this topic under the title “Alternative Positioning, Navigation and Timing, A-PNT”. A-PNT is a complex, multi-disciplinary topic, with technical and operational aspects going across the CNS domains, and spectrum concerns being an underlying driver. The research activities in PJ14-03-04 are covering a selected set of potential technical solutions: take full advantage of the actual DME performance, DME enhancements (ensuring compatibility with legacy systems), LDACS NAV function and eLORAN. This paper will focus the discussion on the performance levels achievable by each of these technologies and their major advantages and drawbacks. The concept of a modular approach will be introduced as well (which allows the computation of a position solution with integrity based on inputs from various types of sensors).