Ocelot is a Lucent tool for improving the overall performance of cellular networks. It is the result of a collaborative research project between the Computing Sciences Research Center, the Mathematical Sciences Research Center, the Wireless Research Laboratory, and Lucent business units.
Ocelot uses a general nonlinear optimization procedure, called with AMPL, to adjust certain parameters of cellular networks and increase a particular performance measure. Ocelot adjusts antenna tilts, azimuths, pilot fractions, power levels, and (for GSM) frequency plans. The performance measure combines coverage, the percentage of the served area from which a call can be made, and capacity, the total amount of traffic that can be carried. When Ocelot optimizes a cellular network, the user sees a Tradeoff Curve window, with points representing difference combinations of coverage and capacity; clicking a point invokes a detailed analysis and visualization of the proposed design, in a graphical display of the market area.
Ocelot models and optimizes CDMA, UMTS, 3G1X, and GSM cellular systems, including 1xEV-DO.
Ocelot uses many sources of information: the elevation and clutter of the terrain; data about the antennas; data about the locations of likely users, gleaned from road maps, service measurements, census tracts, and other sources; information about pathlosses obtained from drivetests; and many others.
Ocelot runs under Windows, and was written in C++. It is currently about 125,000 lines of code.
Ocelot was designed and developed by Computing Sciences researchers Chandra Chekuri, Ken Clarkson, John Hobby, Howard Trickey, and Lisa Zhang, in collaboration with: Georg Hampel, Larry Drabeck, John Graybeal, and other members of Wireless Research, and with Sem Borst, Phil Whiting, and other members of Mathematical Sciences Research. Paul Polakos, of Wireless Research, began the project with Ken Clarkson and John Hobby.
Other contributions, from former center members: