tag:blogger.com,1999:blog-5688599.post4352185955519488075..comments2023-10-12T04:54:05.108-07:00Comments on Deep Freeze 9: RF Mirror Worlds: Supercomputing meets propagation models, 3D terrain data and ubiquitous sensorsJP (Pierre) de Vrieshttp://www.blogger.com/profile/02311009024575927588noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-5688599.post-67709418098364355762014-03-04T10:58:20.032-08:002014-03-04T10:58:20.032-08:00Very interesting. You've laid out a vision of ...Very interesting. You've laid out a vision of making RF environment models better through increased computation, better propagation/terrain models, and more sensor data, then running the models continuously as a "mirror" of the real world. But you don't say precisely in what ways the models should be improved and why improvements of those types would matter. I suggest there may be 3 axes of RF model quality to consider.<br /><br />Accuracy – how big are the error bars in the RF model<br />Granularity – what is the smallest spatial and temporal unit in the model<br />Timeliness – is the model predictive, real-time (how much lag?), or post-facto (how much delay?)<br /><br />To justify and guide work on RF mirror worlds, we need to think about:<br /><br />1. What improvements in the model outputs would offer high value? For example, the granularity of the model directly controls the size of spectrum holes that could be exploited in a dynamic spectrum access scheme. Are there enough bands and users where dynamic spectrum access requires city-block granularity that development of a city-block granularity Mirror World is a priority?<br /><br />2. In what ways does a continuously running RF environment model (a "mirror world") offer higher value than one which is invoked on request?<br /><br />If you want to push forward in this area, I recommend you build a list of different applications for the RF mirror worlds – what could they do or be used for to provide benefits in the operation or development of RF systems – and categorize each one by its requirements for Accuracy, Granularity, Timeliness, and Continuity.<br /><br />John<br />John Chapin, DARPA RadioMap Program Managernoreply@blogger.com