In this research project, we are building a scalable, robust and time-energy aware network architecture for sensor networks. We address scalability, robustness, real-time and energy conservation issues in different layer of the proposed architecture. From bottom up, the proposed network architecture has following essential layers: (Figure 1)
Figure 1 Overarching Plan for the PH.D. Research
Power and Coverage Management Layer: this energy-aware layer is responsible for node duty-cycle scheduling for power conservation and sensing coverage management, which provide full sensing coverage to a geographic area while at the same time minimize energy consumption and extend system lifetime by leveraging the redundant deployment of sensor nodes. [Publication]
Robust and Reliable MAC Layer: This layer solves the issues related to robustness and self-stabilization. It can provide per-hop reliability if required by higher layer. Moreover, this layer deals with the node failure and mobility issues internally as much as possible before signal the network layer for further assistances. [Publication]
Application Independent Data Aggregation Layer: This time-energy aware layer provides data aggregation in order to reduce control overhead in MAC layer and energy consumption in radio communication. Adaptive control is provided in order to achieve aggregation without jeopardizing the end-to-end delay. [Publication]
Differentiated Packet Scheduling Layer: This QoS-time-aware layer support differentiated forwarding service among the packets with different priorities. The criterion for the differentiation includes not only the time constraints but also the spatial constraints. [Publication]
Location-Address Soft Real-Time Routing Layer: This time-aware scalable routing layer provides soft real-time communication for end-to-end packet delivery. Network congestion control is achieved by localized rerouting and traffic policing to the lower layer. Constant delivery speed is maintained through a combination of non-deterministic forwarding and neighborhood-based feedback control. The scalability issue is solved by the location-based routing and localized control scheme. [Publication]
Entity-Aware Transport layer: this layer abstracts communication endpoints into entities. This layer maintained robust connections between entities while both ends are mobile. Entity abstraction allows aggregation among a group of nodes in order to reduce the communication overhead and energy consumption. [Publication]
Localization Service: Localization is a cross layer service for this network architecture as a whole. It provides location information for 1) the sensing coverage management, 2) the velocity calculation for the differentiated packet scheduling, 3) the location-address soft real-time routing, 4) the entity formation and 5) the location service for sensor network applications such as the enemy tracking and temperature mapping. [Publication]
Besides new research issues to be addressed inside individual layers, the major challenge is to build an integrated architecture, which takes care of scalability, robustness, real-time and energy conservation issues simultaneously. Since the point solutions we proposed for individual layers are conceived with this integration in mind, we expect this major challenge will be solved in this thesis research.
Hermes: Hermes' main role was as a messenger. As the special servant and courier of Zeus, Hermes had winged sandals and a winged hat and bore a golden caduceus, or magic wand, entwined with snakes and surmounted by wings. He is also a Greek god of commerce, eloquence, invention and travel.