Invariably all progress is made because we, as the saying goes, are able to see further standing on the shoulders of giants. Yet this is the first time I have seen a paper with such refreshing homage to that adage. The problem space that the authors attempt to solve is one of optimal intruder detection.

Given a rectangular surface with randomly distributed intruder sensors, the desire is to maximize the number of continuous detection belts that an intruder would have to cross in an attempt to cross from one side of the surface to the other. The authors propose to do this with the minimal number of sensors without compromising the probability that each detection belt will be able to detect the intruder. A secondary optimization is to use the discarded nodes as wireless forwarding nodes to minimize the hops between the sensor node and a given set of destination sinks for the intruder notifications.

By adding a dummy node on either side of the detection surface connected to all sensor nodes abutting that edge, the authors transform the problem into one of finding the maximum number of flows between these two dummy nodes--a problem already solved with the Edmonds–Karp max-flow algorithm [1]. To solve the secondary optimization, the authors connect the first dummy node to each sensor node and the second dummy node to each detection sink and optimize for the cheapest path from source to sink using the Orlin–Ahuja min-cost max-flow algorithm [2].

There are however two problems. First, by adding a discontinuity in the detection probability (equation 1), the authors cannot make the assertion in equation 2 and hence invalidate parts of their coverage algorithm. Second, despite having noticed a substantial decline in editorial quality over the last few years, this is the first time I have seen a paper not having passed through a spell-checker. Please notice “bettery,” “estimztion,” “coerage,” “min-csot,” “senosr,” “shoter” and “transmitting,” amongst others.