Chetty and Buyya compare the emerging computational power grid and the developed electrical power grid and consider some of the issues involved in making the former as widespread as the latter. For the most part the comparison effectively highlights the similarities and differences in the questions of resources, network, system load, transmission capacity, and access interface, to name a few.
The authors rightfully conclude that their analogy enables them to identify a number of deficiencies, including lack of an operational model, stability issues, and ease of use, that must be overcome in order to implement the computational grid on the scale of the electrical grid.
Having said that, Chetty and Buyya do fail to carry to their analogy to its logical limits in one important aspect. The propagation of the electrical grid was not just a feat of technical engineering that automatically procreated itself like runaway ivy but a social phenomenon.
Other than the questions of regulation by a governing body and a standardization body, however, the authors ignored much of the human side of how actualize the grid on a grand scale. In particular, they ignore the question of accessibility, a glaring omission if one is comparing computational and electrical grids. By this I am not referring to the ability of, for example, a physically disabled individual (such as a blind persons) to interface his PC with the grid, although this is an important issue as well that the authors neglected. Rather, I mean the availing of the technology to human communities, particularly those that face obstacles to joining a grid.
Recall the history of electrification in the United States. Initially, electricity was only available in cities because of the expense of transmission lines and the seasonal use by farming customers. As it was not cost-effective to bring the countryside into the grid, power companies charged higher construction costs and higher usage rates, which rendered electrification largely unaffordable. By the 1930s, although the vast majority of urban dwellers were on the electrical grid, the vast majority of rural dwellers were not. To speed the entry of American farms online required federal intervention - the creation of the Rural Electrification
Returning to the computational gird, one may ask, will history repeat itself? Will it take an act of Congress to push the grid out of large population centers and into the less densely populated regions of America? Not necessarily, for technological advances may help reduce costs to make rural expansion of the computational grid profitable and practicable for private enterprises, as seventy years ago expansion of the electrical grid could not be. For example, wireless technology requires neither ditches dug nor telephone poles erected nor expensive wiring strung, so no public works entity needs to build the infrastructure.
On the other hand, drawbacks to integrating wireless devices on the grid abound - slower processors, smaller secondary storage, battery depletion, low-bandwidth connection, signal attenuation, and security concerns. Solutions are already under research, such as the proposal for a clustered proxy-base architecture wherein "minion" devices are connected to an "interlocutor" device that represents them on the grid.
However, these workarounds again do not address the needs of, for instance, those who live in the Appalachians where the geographic features would interfere with wireless communication. And expenditure of capital is needed to construct towers in the more remote regions to carry the signal. Would it be public or private monies?
It therefore remains to be seen if technological innovation alone will facilitate the entry of the more isolated have-nots into to computational grid. If the Roosevelt Administration had waited for technology alone (along with market forces) to solve the problem of rural electrification, the de facto policy of prior administrations, the modernization of the American farm and its improvements in the quality of life for rural communities would probably have proceeded much more slowly. Just as lack of electricity creates obvious social and economic barriers, so increasingly will information and communication poverty - Michael J.
Muller and others call "an inability of human beings to obtain, process, and communicate information critical to their life and livelihood" - constrain the abilities of individuals and communities to fully participate in society. Consideration of technological availability in the computational grid's infancy is therefore warranted to bring the benefits of this endeavor to more people in less time and make the grid a truly universal and beneficial accomplishment.
Perhaps one could argue that public policy was beyond the scope of their paper. The authors chose to compare the computational grid to the electrical grid, however, both exist in social and political contexts, and therefore a solution to the problem of computational grid technology must also take those contexts into account. While Chetty and Buyya make a case for the analogy between the electrical and the computational grid, it would have been more informed if they had considered the macro-issues of availability in human communities.
--- See, for instance, Ephraim P. Gilnert, "Ensuring Access For People with Disabilities to the National Information Infrastructure and Multimedia Computing," ACM SIGCAPH, September 1997, 59:10-16
Thomas P. Hughes, "Electricity for Rural America: The Fight for the REA." (rev.), Journal of Economic History 41 (June 81): 476.
Thomas Phan et al., "Challenge: Integrating Mobile Wireless Devices into the Computational Grid," Proceedings of the 8th Annual Conference on Mobile Computing and Networking International Conference on Mobile Computing and Networking, Atlanta, September 23 - 28, 2002.
G. Formand and J. Zahorjan, "The Challenges of Mobile Computing," IEEE Computer, April 1994, 27(4): 38-47.
Douglas R. Hurt, "REA: A New Deal for Farmers," Timeline, 1985-86 2(6): 32-47.
Michael J. Muller, et al., "Toward a New HCI Research Practice Agenda Based on Human Need and Responsibility," Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, Atlanta, 1997, 155-61.