Can you show us what you have tried so far?
Both of my strategies I have come up with involve an initial epoxidation, whether that be diastereoselective (through coordination with the OBn) or asymmetric, like a shi-epoxidation, followed by some sort of ring-opening and alterations.
Pathway 1.
http://prntscr.com/rrxvfdIn pathway 1 I am counting on the relatively labile TMS group to fall off at some point before or during the tosylation, since a fluoride source would also deprotect, at least partially, the TBS group too. The TMS-CN reaction with the zinc iodide catalyst does have high regiochemical selectivity which is an upside to this reaction.
Pathway 2.
https://prnt.sc/rrxwlsMy main issue with pathway 2 is that I am worried cyanide wouldn't have that great of a diastereoselectivity here, and I am having difficulty finding an alternative. I am also worried about directing the epoxide with. In theory, the Ti(O-I-Pr)4 may coordinate to the OBn group and direct the peroxide (similar to a Sharpless epoxidation), but that peroxide will also want to attack from the other side due to the bulky Bn protecting group itself.