LEM: a Cooperative and Self-Organizing Flow Relaying Middleware for Multimedia Continuity in Dense Hybrid Wireless Networks

The recent advances in wireless client devices, that host heterogeneous wireless technologies spanning from IEEE 802.11 (WiFi) and Blue-tooth (BT) to cellular 3G and beyond, and the crucial role of multimedia communications in our society are motivating new multimedia pro-visioning scenarios. In particular, industrial in-terest has recently started to concentrate on the opportunities of tailoring multimedia content delivery depending on client location, with dif-ferent goals, from personalized location-dependent advertising/marketing to enhanced forms of touristic assistance. Those services show their main benefits when applied to highly populated regions where specific contents should be distributed to geographically bounded Areas of Interest (AOIs), e.g., personalized and differ-ent promotional videos toward the areas used by different exhibitors in an exhibition pavilion. Despite their recognized potential benefits, the development and deployment of those new classes of AOI-targeted multimedia services are still challenging tasks, also due to their quality requirements, e.g., data arrival time, jitter, and packet losses. In addition, densely populated spots introduce new challenges: i) they are typi-cally short-termed and that makes it very costly to plan and deploy in advance needed net-work/provisioning infrastructures; ii) fixed wire-less infrastructure requires installation of power lines and wired connections to the Internet, which is unfeasible in several cases; iii) node mobility further exacerbates the above issues, especially the seamless provisioning of service flows with quality and continuity requirements. This letter overviews the research work we are recently doing for an original self-organizing middleware for seamless multimedia delivery in dense wireless regions with AOIs (dense AOIs). We claim that the proposed solution is relevant for three primary technical contributions. First, it automatically and continuously re-organizes dense AOI network topology using a lightweight localization technique (based on received signal strength indication - RSSI, infrastructure-less, and anchor-free). Second, it takes advantage of dense network population and wireless technol-ogy diversity (rather than considering it a techni-cal issue) to achieve the needed positioning ac-curacy and multimedia distribution scalability. Third, it shows how it is possible to effectively exploit mobile relays, automatically re-elected within the community of dense AOI nodes, as both markers for AOI identification and com-munication gateways for other nodes in the same neighborhood. Here, we rapidly describe the open-source implementation of our middleware, called Localized rElay-based mobile Multimedia (LEM), and a selection of relevant experimental results about node mobility monitoring, dynamic relay determination, and client-side multimedia buffering. Additional information, further ex-perimental results, and the LEM prototype code are available at the project website [13A].