Analysis of a crowd-computing search algorithm in mesh networks

This paper presents the principles and performance analysis of a general search algorithm to locate the best matched resources from the candidate nodes in a wireless mesh network. This algorithm uses a cross-layer network protocol and the concept of crowd-computing which lets all the wireless nodes in a mesh network to perform the searching and comparing. It shows much higher efficiency than the conventional methods. Our lab test has validated the algorithm in a small-scale network. Event-based simulation results are given to prove its efficiency and scalability in mesh networks with more than 1000 nodes.