Location privacy is a concern in mobile networks monitoring systems. We propose
a model using homomorphic encryption, where researchers can compute
aggregations, learning nothing more about users' location than the result of
the aggregation. We build a location privacy layer on top of Adkintun Mobile
using a Partial Homomorphic Encryption Scheme (Paillier) and a Fully
Homomorphic Encryption scheme (HElib) and compared the performance. The result
shows that both systems are practical for off-line statistical computation.