Connecting wireless sensor networks (WSNs) to the Internet is a very interesting topic. As useful as WSNs are, they used to be constrained in a separate compartment of the network. The collected data needs to be hauled in by a base station that creates a single point of failure. I for one had the unpleasant experience of resurrecting one to regain access to the sensor network and its data. Address allocation is certainly a big stride toward true integration. Internet protocol version 6 (IPv6) accommodates more devices for its massive address space compared to IPv4.

The current state of this research topic is not entirely clear through the literature survey in this paper. Several conventional address configuration/allocation algorithms are surveyed and compared with their assumptions and potential issues. However, as stated, this prior literature did not take advantage of the location information that has proved to be accessible for sensor networks. This paper is probably not the first to present such an idea. A survey of truly related materials would prove helpful. Such materials conveniently show up later in the simulation results when the authors attempt to demonstrate the effectiveness of the proposed solution.

The main solution is described in detail. Built on top of the sensor geo information already obtained, the WSN area can be divided up into zones. Each zone acts independently, thus avoiding the typical issues introduced by flooding the entire network. Based on a certain schedule/priority, the nodes within the same zone are assigned an address. When the address space is used up, address reclamation takes place.

For the performance measurement, it would be great to see work done in a realistic test bed. Unfortunately, we have to rely on an NS2 simulation and theoretical analysis. The results are straightforward and not surprising. One would certainly be interested in more in-depth exploration of this topic with a realistic scenario that would bring much more complexity into the picture.