0% away from on−grid 1e−04% away from on−grid 1e−03% away from on−grid 1e−02% away from on−grid 1e−01% away from on−grid 5e−01% away from on−grid

This chapter introduces the Random Modulation Pre-Integrator (RMPI) and details the status of an integrated circuit implementing the architecture. To our knowledge, this is the first full-speed hardware RMPI in existence. The goal of this chapter is to describe some key parameters of the RMPI that are essential to its performance. This is not a hardware chapter and does not go into detail of the designs; the reader does not need any knowledge of Op-amps, though standard mathematical and signal processing background is helpful. The idea behind the parameter selection is that of trade offs. We presuppose that a competent engineer is able to make several designs, and the task is to decide which design is better. This introduction describes the setting of the RMPI and its basic principles, and tries to make clear the intuition behind the receiver, as well as outline when the RMPI is an appropriate architecture and when it is not. Further sections of the chapter describe the system in more detail, and report a some of the many results from 3 years of study and simulation. The hardware implementation of the RMPI has been returned from the manufacturer, and is in the process of being calibrated, which has required significant work (to be discussed in §2.3.4), but results are encouraging. An earlier chip, which we refer to as “version 1,” was manufactured and some reconstructed pulses obtained, but otherwise all plots in this chapter are from simulated data. The simulations took place at various levels of realism, which will are described in §2.3. Most computational work was performed in Matlab and/or Simulink. The RMPI is due to the work of a large number of people, but we especially want to mention the efforts of Juhwan Yoo. He has done most of the hardware design, collaborated on the system design, and contributed figures and occasional text to this chapter. Funding for work was graciously provided by DARPA grant FA8650-08-C-7853.