The Jet Propulsion Laboratory has begun work on large arrays of small antennas for increasing the receiving capability of the NASA Deep Space Network (DSN). These receive-only arrays promise to be the lowest cost way to meet the downlink needs of future missions. The DSN also need new uplink capability: a) to provide higher uplink rates to new types of missions, b) to support more spacecraft simultaneously, c) to provide high effective isotropic radiated power (EIRP) for very distant missions and for spacecraft emergencies, and d) to back up and potentially replace the aging 70-m antennas. This paper discusses various approaches to providing new uplink capability, with the goal of minimizing the overall life cycle cost of the DSN. One proposed way to provide new uplink capability is by uplink arrays of small antennas. This approach has technical challenges, and it needs to be demonstrated that they can be overcome. The second major option is to use 34-m and/or larger antennas for the uplinks. In this option, 34-m antennas might be arrayed to obtain the same or greater EIRP as might be obtained on 70-m antennas. This work presents a scenario for increasing the overall DSN capacity in several steps. Building blocks are defined that can be used to implement these steps. The key building blocks are 34-m antennas with uplink and downlink, 34-m antennas with receive only, and various array configurations that are equivalent in performance to the 34-m antenna configurations. Making cost estimates for these building blocks facilitate estimating overall DSN costs for various approaches and levels of capability.