An effort based social diffusion approach for mobile encounter networks

Mobile information systems use publish-subscribe mechanisms to distribute content between users. Unfortunately, content that is close to the user is not necessarily accessible unless users have specified so earlier. Mobile encounter networks (MENs) represent an alternative, which spread content in a spatial diffusion process. A drawback of MENs is their high resource usage. Some researchers have proposed diffusion methods that utilize knowledge about the social network behind the user community. Such approaches work well in the wired Internet since robust bandwidths exist. This, however, cannot be assumed for MENs where the bandwidth is determined by the encounter duration. We use an effort based knowledge diffusion approach to solve this problem. Effort based diffusion mechanisms differentiate between simple and complex content types and adapt content exchanges to the capabilities of involved peers. We emulate content diffusion between humans with an urban simulator. During the simulation we obtain data about content exchanges and cached contents. Results show that the least effort theory outperforms traditional approaches. Being aware of content transfer efforts, such routing approaches increase the overall transparency of large scale urban communities and to the same time enable richer content exchanges within friend networks.