Overview

In the last few decades, the main advances in computer science and telecommunications have fostered the realization of new interesting scenarios in which mobile nodes are able to form impromptu mobile networks without the intervention of any fixed wireless infrastructure. Such a network, usually known as Mobile Ad-hoc NETwork (MANET), groups different mobile devices that address message routing in a completely decentralized fashion: each node can exchange messages only with the neighbors in its own transmission range, and can enact as router to enable multi-hop message routing.

Between different possible applications, MANETs are widely used in disaster area scenarios, i.e., in those areas that experiment sudden, unexpected and terrible disasters due to natural, e.g., earthquake or tornado, or human, e.g., terrorist attack, causes. In particular, since pre-existing traditional fixed wireless infrastructures (such as 3G networks) could be completely damaged, those scenarios use MANETs to address the communication task inside the disaster area. The communication between the different rescue teams is fundamental to coordinate the rescue operations, so to increase the possibility of saving human lives.

Context-awareness, namely the capability of providing the applications level with full awareness of current execution context (such as location, devices and people in the physical proximity, etc.) has a key role in above scenarios. Different and novel context-aware services can be deployed to assist the different rescue teams (e.g., to suggest the people that require immediate medical assistance) and to provide real-time access to physical surrounding status (e.g., to suggest the shortest/safest path to reach a particular building). Unfortunately, to enable the real world realization of these novel context-aware services, many different issues have to be addressed. First, these systems usually use bandwidth-constrained and unreliable wireless links, e.g., IEEE 802.11 (WiFi) and Bluetooth (BT). Second, these systems are based on devices with limited energy and processing capabilities. Third, these systems have to manage a huge amount of context data in a completely decentralized way.

To manage context data distribution, we need Context Data Distribution Infrastructures (CDDIs) able to transparently tackle above issues. Since context data distribution should be i) as least intrusive as possible to reduce management overhead, ii) as fast as possible to enable timely context access, and iii) as reliable as possible to guarantee reliable access to safe critical data, we consider context access a principal problem, and we focus on context management aspects in challenging environments.

RECOWER (Reliable and Efficient COntext-aware data dissemination middleWare for Emergency Response) is a joint project developed at the PARADISE Research Laboratory, University of Ottawa and at the Laboratory of Advanced Research on Computer Science, University of Bologna, that addresses the realization of context-aware services in disaster area scenarios by following innovative solutions:

• Heterogeneous wireless communication: RECOWER integrates with and uses both WiFi and BT, as the most widespread ad-hoc wireless technologies. In this way, it increases both the final available bandwidth (to improve scalability) and the communication opportunities between devices (to reduce the probability of MANET partitions).
• Distributed context data memorization: RECOWER realizes a distributed context data repository based on the tiny repositories offered by traditional mobile devices. Context data are replicated to increase data availability and to reduce the average path length followed to retrieve them. At the same time, data replication is carefully controlled to efficiently use the scarse available resources.
• Quality of Context (QoC)-based data distribution: RECOWER introduces Quality of Context (QoC) contracts to express quality requirements on exchanged context data and CDDI. These contracts, named Context Data Distribution Level Agreements (CDDLAs), detail the quality requirements RECOWER must grant: as long as those requirements are expected, RECOWER self-adapts to maximize context data distribution quality.
• Context-aware data distribution: RECOWER adapts context-awareness at the CDDI level. In particular, to prevent wireless channel congestion, RECOWER coordinates neighbor nodes and adapts data distribution to improve efficiency and reliability.
  
  

Involved people