|
Abstract: |
The next-generation networks are envisioned to be deployed as an infrastructure of sensor devices that are available anywhere and any time, autonomous, survivable against multiple faults and attacks, and highly secure for communication. Sensor networks hold the promise of revolutionizing sensing in a wide range of application domains because of their reliability, accuracy, flexibility, cost-effectiveness, and ease of deployment. One of the most fundamental purposes of sensing information is to immediately respond to any anomalies. In order to make an accurate and cognizant decision, there is a great need for a dependable wireless sensor network. Failure dependability relates to providing dependability against node and/or link failures.
In multi-layered wireless sensor networks with data fusion, we observe that as the data propagates from the sensor node to the sink, the data packet represents an increasing number of sensor nodes (sub-tree) in the network. Hence, it is critical that this data packet reaches the sink, since the loss of a single packet eliminates the information sensed by a whole (possibly large) sub-tree of sensor nodes. We evaluate two dual-homed routing techniques for providing fault tolerance, namely 1+1 dual-homing and 1:1 dual homing. Based on extensive simulation results, we observe that 1:1 dual-homed routing minimizes packet loss, increases network throughput, and increases network lifetime compared to the 1+1 dual-homed routing and the traditional single-homed routing techniques. |
