Memory Optimization Techniques For Computing Discrete Logarithms In Compressed Sike

The supersingular isogeny-based key encapsulation (SIKE) suite stands as an attractive post- quantum cryptosystem with its relatively small public keys. Public key sizes in SIKE can further be compressed by computing pairings and solving discrete logarithms in certain subgroups of finite fields. This comes at a cost of precomputing and storing large discrete logarithm tables. In this paper, we propose several techniques to optimize memory requirements in computing discrete logarithms in SIKE, and achieve to reduce table sizes by a factor of 4. We implement our techniques and verify our theoretical findings.