Abstract
In this paper we abandon the intuition that the minimum length/latency schedule of aggregation convergecast has to be constructed in an order mimicking the flow of data, i.e., from leaves to the sink. Instead we adopt a top-down construction process (from sink to the leaves), and thus forego the need to construct an aggregation tree a-priori of the (collision free) schedule construction. The resulting scheme, called MINT, is a family of heuristics whose performance is controlled by a candidate link selection strategy. We report on the best found combination of MINT and candidate link selection strategy which specifically attempts to minimize the "cost" incurred by a scheduled transmission where the cost is the number of blocked transmissions that could have occurred in the same timeslot. The results produced show a significant and systematic improvement over the state-of-the-art. We analyze the results to uncover that the improvement due to MINT hinges on the even spatial reuse of the wireless medium.
Item Type: | Journal article |
---|---|
Faculties: | Mathematics, Computer Science and Statistics > Computer Science |
Subjects: | 000 Computer science, information and general works > 004 Data processing computer science |
ISSN: | 2163-4033 |
Language: | English |
Item ID: | 47460 |
Date Deposited: | 27. Apr 2018 08:13 |
Last Modified: | 27. Apr 2018 08:13 |