他大意是說,利用車子上的sensor感應heading(方向) and speed(速度)來找出能連接AP的時間最長
VEHICULAR NETWORK PATH SELECTION
In general, broken paths increase overhead and latency.
For this reason, selecting paths with the longest expected connection time may be a good idea.
We present three connection time estimate (CTE) metrics that use heading and speed hints to estimate whether a path in a vehicular network is likely to be long-lived.
Let the ordered sequence (u1,...,uj) denote a j − 1 hop path, dh(ui,ui+1) denote the difference in heading of nodes ui and ui+1 (from 0 to 180 degrees), and s(ui) denote the speed of ui (in m/s).
Our three CTE metrics, called cte1,cte2, and cte3, are deﬁned for a path R = (u1,...,uj)
The metrics use the assumption that a small difference in heading indicates nodes are moving in the same direction on the same road, and are therefore likely to stay connected longer.
The cte3 metric includes the additional assumption that speed is inversely correlated with connection duration.
Because cte1 multiplies the inverse of heading differences at each hop, it is biased toward single-hop paths.
The cte2 metric, by contrast, evaluates a path only by its worst hop, scoring multi-hop paths higher than cte1.
The cte3 metric multiples the inverse of the sum of node speeds with the cte1 value.
== 簽名檔 ==