A common characteristic of the emerging generation of wireless networks is their heterogeneity: these networks consist of devices with very different capabilities and requirements sharing overlapping spectrum. Intelligent home networks consisting of HD streaming, gaming consoles, wireless routers, and energy monitoring devices is one example; cognitive networks (utilizing white spaces), and femtocell networks being two more. Somewhat surprisingly, nearly all engineering design, analysis, and knowledge is based entirely on the assumption of node homogeneity Our research involves obtaining fundamental new laws and limits for the connectivity and capacity of heterogeneous wireless networks. The nature of such limits is expected to provide a roadmap for better design principles.
The novelty of this research is based on a new application of marked point processes, where the points model the node locations and marks characterize essential traits of the node, like their bandwidth and power. To calculate the desired statistical properties of the network, we introduce and advance new mathematical tools such as Tauberian theory, series of random functions, Stein approximation theory, and sub-ergodic theory. In addition to the basic research aspect of this project, we are pursuing an energetic program of technology transfer with several leading companies developing heterogeneous networks and planning standards contributions and other intellectual property development. This is consistent with the PI's past track record.
", I'm pleased to report that we surpasssed even my most optimistic projections of what we would achieve during the lifetime of the project. The project resulted in seminal contributions on the modeling, analysis, and understanding of heterogeneous cellular networks, which are conventional cellular networks augmented with low power nodes such as picocells, femtocells, and distributed antenna systems. For the first time, we were able to derive a closed-form expression for the signal-to-interference-plus-noise ratio (SINR) in single-tier (i.e. homogeneous) cellular networks. This paper, published in Nov. 2011, is the most cited technical paper in the field of wireless communications in the last 4-5 years and has led to a very large volume of follow on work. It showed mathematically that cellular networks have a "scale-invariance" property, whereby making them denser and denser (by installing more base stations) has virtually no effect on the SINR. This paper won the Stephen O. Rice prize paper award in 2014 as the best paper published in the previous three years in the IEEE Transactions on Communications. Our own extension of this work to heterogeneous networks, i.e. having many different types of base stations, was published in Apr. 2012 in the IEEE Journal on Sel. Areas in Communications (JSAC). This paper showed, quite interestingly, that the above results and intuitions can be extended to a very complex network where the different types of base stations have different densities, transmit powers, and so on. A particular intuition is that even randomly adding small cells all over the network does not create "interference overload": rather the network still exhibits scale invariance as long as the mobile phone can connect to the strongest signal. Intuitively, the desired signal gets stronger at the same rate that the interference increases. This led to increased optimism about small cell deployments (operators were concerned that deploying nested small cells might bring down their carefully planned network due to interference). This paper too has received a very large volume of citations and was honored with the 2014 Leonard G. Abraham prize paper award as the best paper published in the previous three years in JSAC. In addition to these two seminal papers, this project supported a great deal of other highly-cited work which has had real-world impact. This includes our work on load balancing for cellular networks, as well as more theoretical work that develops new tools for stochastic geometric analysis of wireless networks. This work has already appeared in several textbooks, and has been integrated into the wireless communication curriculum at UT Austin.
