Dynamic Buffer Sizing for Out-of-Order Event Compensation for Time-Sensitive Applications
This addresses data reliability for time-sensitive applications like sensor networks, but it is incremental as it builds on existing buffering methods.
The paper tackled the problem of out-of-order data in time-sensitive sensor networks by developing dynamic buffer sizing algorithms, showing that dynamic time-out buffering outperforms static buffers, especially under high network variation.
Today's sensor network implementations often comprise various types of nodes connected with different types of networks. These and various other aspects influence the delay of transmitting data and therefore of out-of-order data occurrences. This turns into a crucial problem in time-sensitive applications where data must be processed promptly and decisions must be reliable. In this paper, we were researching dynamic buffer sizing algorithms for multiple, distributed and independent sources, which reorder event streams, thus enabling subsequent time-sensitive applications to work correctly. To be able to evaluate such algorithms, we had to record datasets first. Five novel dynamic buffer sizing algorithms were implemented and compared to state-of-the-art approaches in this domain. The evaluation has shown that the use of a dynamic time-out buffering method is preferable over a static buffer. The higher the variation of the network or other influences in the environment, the more necessary it becomes to use an algorithm which dynamically adapts its buffer size. These algorithms are universally applicable, easy to integrate in existing architectures, and particularly interesting for time-sensitive applications. Dynamic time-out buffering is still a trade-off between reaction time and out-of-order event compensation.