Autonomic Obstacle Detection and Avoidance in MANETs Driven by Cartography Enhanced OLSR

Algorithm 3

Concave hull construction step.

Require:

    : list of points retained after the filtering step.

    : the threshold based on which we decided to dig or not into the encountered concavities.

Ensure: ConcaveHullEdges: a List containing the edges forming the built concave hull of .

 ()    ConvexHullPoints ← BuildConvexHull()         ▹  build the points list forming the convex hull that

      encompasses scatterplot using the well known Jarvis March algorithm

 ()    ConvexHullEdges: list

 ()    for  (; ≤ ConvexHullPoints.size; )  do ▹  convert ConvexHullPoints to a list of edges

 ()        mod ConvexHullPoints.size

 ()       Point  ← env[]

 ()       Point  ← env[]

 ()       Edge ()

 ()       ConvexHullEdges.pushBack()

 ()    end for

 () ConcaveHullEdges ← ConvexHullEdges

 ()   − ConvexHullPoints

 () for each  edge   in  ConvexHullEdges  do

 ()     ← FindNearestInnerPointToEdge()

         ▹   is the nearest point of to such as neither of neighbour edges of is closer to than

 ()    if  ( exists)  then

 ()       ← length of the edge     ▹    =  euclidianDistance(.ext1, .ext2)

 ()      decisionDistance ← distanceToEdge(, )

 ()      if  ((/decisionDistance) > )  then

 ()          insert new edges (.ext1, ) and (.ext2, ) into the tail of ConcaveHullEdges

 ()         delete the edge from the ConcaveHullEdges

 ()                           ▹  delete from

 ()      end if

 ()    end if

 () end for

 () Return  ConcaveHullEdges