You are here: Resources > FIDIS Deliverables > HighTechID > D3.3: Study on Mobile Identity Management > 

D3.3: Study on Mobile Identity Management

mCrowd  Study on Mobile Identity Management
 Comparison of Anonymous Communication Mechanisms for ad hoc Networks


for anonymising WAP surfing

mCrowds (Andersson et al. 2003; Andersson et al. 2004) is a low-latency anonymity technology developed at Karlstad University. The purpose of mCrowds is to minimise the dissemination of personal information on the mobile Internet. It does so by enabling anonymous Wireless Application Protocol (WAP) browsing and by minimising the disclosure of personal information when using location-based services (LBS). In cases where the location is measured by the mobile device itself, location based services can be used anonymously if this mode is supported by the LBS provider.

Introduction to Architecture

mCrowds is based on Crowds (Reiter and Rubin, 1997), a system for anonymous web browsing on the traditional Internet. Crowds works by grouping users into a large anonymity set, a so called crowd, which issues requests to web servers on behalf of its members. Crowds is a peer-to-peer technology where each user runs his / her “Mix” in the network (called a jondo) and the communication is routed along virtual paths consisting of many such jondos. A dedicated node called the blender is taking care of membership management. In mCrowds the concept of a traditional Crowds system applied in a mobile Internet setting is combined with a personal privacy proxy that acts as a filter tailored to anonymise mobile requests. The figure below illustrates the mCrowds system. Note that the crowd itself resides on the wired Internet domain.

Figure 5-3: mCrowds overview


Performance Issues

Performance was one of the primary design goals in the development of mCrowds. The traditional Crowds system was chosen to provide a base for mCrowds, since Crowds as a base is supposed to offer better performance properties than the more common anonymity technologies based on Mix-nets (Chaum, 1981), such as JAP Web Mixes (JAP, 2003) or Onion Routing (Andersson, 1996). This is because Crowds as a base provides better scalability properties and further the use of public-key cryptography is minimised. Further performance enhancements have been implemented in the communication protocol of mCrowds.

The performance of mCrowds has been measured in a performance evaluation that measured the performance overhead introduced by mCrowds when browsing anonymously on the mobile Internet (Andersson et al., 2004). To make the conditions realistic, an experimental crowd was simulated where the nodes in the crowds were separated by a relatively large geographical distance. The results of the performance evaluation were encouraging, as the performance overhead was relatively small compared to the total latency. In figure 5-4 below, the total response latency while fetching data from a WAP server is measured. The results are plotted for firstly the case where mCrowds is enabled and the path length is four and secondly for the case where mCrowds is disabled.

Figure 5-4: Performance evaluation 


Mobile Internet introduces new privacy risks and privacy legislation alone is not sufficient to secure informational privacy for users. Thus there is a need to develop privacy-enhancing technologies in addition to privacy legislation. One contribution is mCrowds, which is a privacy-enhancing technology that enables anonymous WAP browsing on the mobile Internet.

A number of experiments have been made to evaluate the performance of mCrowds in practice, in which the performance overhead generated by mCrowds was measured. The subsequent results of this performance evaluation were encouraging as the overhead in performance introduced by mCrowds was relatively small compared to the total response latency when fetching WAP pages via the mobile Internet. The results of this performance analysis can serve as a comparison to other approaches for anonymity on the mobile Internet. The area of anonymity and identity management on the mobile Internet is growing fast and such technologies will become more common in the coming years. The contribution in the form of mCrowds can be seen as one of the initial steps.


mCrowd  fidis-wp3-del3.3.study_on_mobile_identity_management.final_04.sxw  Comparison of Anonymous Communication Mechanisms for ad hoc Networks
24 / 36