Network boundary recognition via relative azimuth angle

To recognize the boundary of wireless networks to meet some specific needs, relative azimuth angles of nodes distributed in the 1-hop neighborhood of each node in one network were used to recognize the boundary of the network. A novel recognition algorithm based on local generalized convex hull was proposed for one node to decide if it self is a boundary node. The flexibility, efficiency and consistency of the algorithm are shown in its performance by extensive simulations.