Privacy in location-based services is often viewed as reaching a desirable trade-off between performance and a user's privacy; the more privacy provided, the less likely it is that the service can function as it would under a no-privacy scheme, or without suffering alterations in their architecture or application layer. As LBS offers a great variety of ever-changing features that keep up with users' needs while making use of the latest available technologies and adjusting to social behavior, they provide a similar scenario to LPPMs that aims to cover these services.
In the case of proactive location-based services (PLBS), where users are constantly reporting their location [4], the purpose of LPPMs is to alter the route as much as possible, while still providing a minimum level of accuracy that will allow the LBS to provide relevant information. This can be challenging because many PLBS, like traffic guidance apps, require the exact location of the user. So, unless the original data can be recovered or used in the altered format, it would be very complicated for these applications to implement an LPPM. Other services, like geomarketing or FriendFinder, may tolerate a larger alteration of the data, even if the change cannot be undone.
On the other hand, mechanisms intended for reactive location-based services (RLBS) often do not require critical accuracy, and therefore it is tolerable to alter the subject's position in order to provide location privacy.
Some LPPMs require special features alongside the usual client-server architecture, such as special database structures, extra data processing layers, third-party services, proxies, special electronics, a peer-to-peer approach between the LBS users' community, and so on.
Based on this, a proposed way to classify LPPMs is based on the application to PLBS and RLBS. Some of the techniques are general enough that they can be used in both worlds, but each has different implications:
In this chapter, two examples of LPPM implementations will be shown: noise-based location obfuscation, and private-information retrieval. Each of these imply changes to the design of the LBIS and the geographical database.