Build a transitor. Big, big, big ask up front, but it seems like it'd make all other mars missions trivial by comparison. Start with a lunar transitor to facilitate the lunar base missions.
I'm assuming that's what OP means. An Aldrin cycler is a cool idea and I'd love to see one developed some day, but I don't see how it helps unmanned missions significantly. You still need to independently launch and accelerate whatever you want to send to the same orbit for the rendezvous.
Makes sense for creature comforts (like life support and radiation shielding) for the long cruise but why bother for a robotic mission?
Don't bring them back maybe, invest in developing ways to get more complex testing equipment to mars? Obviously there are tests where this is impractical, if not impossible, but I always loved the idea of being able to land like a robotic controlled science center there.
Break the problem into parts. Design a system to get sample canisters into martian orbit. Design a Martian orbital system to catch the canisters and send them towards earth. Design an earth orbital system to intercept the canisters. Design a system to get the canisters from Earth orbit to the surface.
Each of these systems will be separate structures with different goals. Each one will be technically difficult and require new methods but each smaller piece will be easier to solve than the problem as a whole.