The premise for grid computing is resource sharing with the intensive use of Internet protocols. The architecture for grid computing has leveraged the benefits of various evolutionary architectures, including client-service and distributed architectures. The basic purpose of this architecture is to use the inherent strengths of networks, the Internet, databases, and storage, and to provide better services. The benefit of grid architecture is the extension of services with optimized cost and without compromising security and quality of service (QoS).

This paper discusses network service descriptions and discovery for high-performance grids across heterogeneous networks. It is an interesting paper that presents a model to describe service capabilities and service discovery technology for selecting network and resource allocation schemes. This model, therefore, is contemporary in terms of offering approaches and solutions for providing enhanced QoS with flexibility. The author supports the developed model with algorithms, and examines two cases to validate the results obtained through the application of this model.

There is scope for improvement, however, in establishing certain claims. Pervasiveness and ubiquitous features of a networked environment are quite instrumental in enhancing QoS. In this model, the author does not comprehensively present these two features. First of all, the paper could have included two separate sections to discuss pervasive and ubiquitous features of networks with the support of literature and algorithms. Through the cases discussed, the author could have validated deliveries of the model involving these features. With regard to the cases, the author does not discuss whether the simulations were done in an intranet environment. A comparative assessment of the results obtained through the application of Internet bandwidth with multiple service providers and the intranet environment in a campus network could have provided better insights. Benchmarking the results could have supported the findings more succinctly.

Still, this paper is quite current and will benefit researchers who are interested in this evolving area.