Friday, April 9, 2010

Cut Throat World of Electromagnetics

The government's newly established health care reform policy has made many people worry that the government is becoming too big and controlling. There's many reasons for this. Personally I do not yet know how I feel about it, mainly because I haven't taken the time to fully understand what changes are going to happen and how they are going to be implemented. Right now I'd say I am both pleased and upset with what I do know. I am pleased that insurance companies will no longer be able to reject children with pre-existing conditions. Although I am displeased that apparently one of the revenue raising tactics is going to be an increased tax on investment income. Obviously forcing insurance companies to take on pre-existing conditions will raise prices. But raising taxes on investment incomes doesn't seem right. I'm not a politician but I say let's tax the stuffing (how's that for family friendly wording) out of alcohol and cigarettes. Let our main source of revenue for paying for health care come from the purchases of unhealthy products. That's health care reform. In fact positively this is already happening to some extent by the 10% increase placed on tanning bed services. And I realize that would mean a higher tax on my pop and junk food but I'd be happily willing to pay that tax before one on my investments. Especially with all this talk on needing to improve the economy. Putting higher taxes on investors doesn't seem like a good way to stimulate and encourage economic growth.

While I am still on the fence about health care reform and whether or not the government is overstepping it's bounds, there are also other areas that the government is trying to exert some control. And one of those hits very close to home.

About four years ago the government decided to create an inhouse program called CREATE. The purpose of CREATE is the development of a new electromagnetic solver. Electromagnetic solvers have a variety of purposes. Say for example you're a cell phone company seeking to bring coverage to a new city. Some EM solvers will let you model that city in 3D, and place towers at different locations. Then the solver will graphically show you the coverage region and strengths. The cell phone company can then use a bit of trial and error to decide where the best place to put a tower is to ensure the greatest amount of coverage. There are a ton of EM solvers out there, from those written by grad students to those that have been consistently developed for 20 years or more by some of the major US corporations like Boeing. Basically they all revolve around the same math, and the answers they give should be about the same. Differences come from what approximations are made. A full-wave solver for example will solve for Maxwell's Equations without any approximations. There are 5 Maxwell's Equations and they describe the behavior of electromagnetics. For example, cell phone signals obey Maxwell's Equations. A full-wave solver will produce results that are essentially exactly what the real-world scenario will exhibit. The only limitation is the accuracy of the models being simulated. For example, that town with the cell phone tower may be modeled by a few cubes representing buildings, or it could be represented with full architectural precision where even the material properties like glass, concrete and the leaves and trunks of trees are accounted for. But these full-wave solvers take forever to perform simulations. Typically only small canonical problems can be solved because the processing power and memory required to solve many problems of interest simply does not exist right now. So approximations are made to solve Maxwell's Equations. With these approximations, some problems that might take years to solve using a full-wave solver may be reduced to only minutes. And the answers will still usually look reasonably close to what we would observe in the real world.

You may be curious to know that these solvers can cost anywhere from nothing at all because they are freely distributed, to over one-hundred thousand dollars, to not even readily available because the government has deemed the need for them to be classified.

Since forever, the government has been using these programs developed by the major corporations to solve its electromagnetic problems of interest. But with CREATE the government is going to mandate that it is used by all corporations instead. Now in reality this is really just a dream, because CREATE is a full-wave solver and the technology to reasonably finish the government's problems of interest simply does not exist. Although in the next 8 years, the time period left for CREATE to be developed, maybe sufficient advances will be made. At any rate, one of the hats I wear is to help one of those EM solvers that the government is seeking to replace with CREATE. So in 8 years I may very well find myself with a cold head.

0 comments: