Emanuele Montanari

System-Oriented Harmonic-Balance Algorithms for Circuit-Level Simulation

This paper discusses a self-consistent set of modern computational concepts providing an effective approach to the circuit-level harmonic-balance (HB) simulation of nonlinear microwave systems of complex topology. The system is automatically split into the interconnection of a near-optimal number of nonlinear blocks at run time. The resulting structure is then exploited by the domain-partitioning concept. A block-wise constant spectrum is used rather than a common spectrum by considering for each block only the set of spectral lines that are relevant to its electrical function, which leads to a very significant reduction in the number of problem unknowns. System simulation under digitally modulated RF drive is reduced to a sequence of modified multitone HB analyses that are backward coupled through the envelope dynamics. Besides providing high numerical efficiency, this set of techniques opens the way to an effective co-simulation of RF and baseband transceiver sections.

A New Wireless Displacement Sensor Based on Reverse Design of Microwave and Millimeter-Wave Antenna Array

This paper introduces for the first time a new low-cost, nondestructive sensor to remotely monitor structural displacements (patent pending). The sensor consists of a two-element antenna array, operating at microwave or millimeter wave frequencies, remotely interrogated by a reader. The elements are stuck on the two sides of a crack to be monitored, no matter how large the crack is. The operating principle is based on the dependence of the array radiation characteristics on the element distance normalized to the wavelength. A combination of electromagnetic (EM) interferometry and Radio Frequency IDentification (RFID) technology allows the displacement of the antennas with respect to a reference configuration to be remotely determined. It is demonstrated that null measurements of the radiation pattern at microwave frequencies allow structural displacements to be predicted with uncertainties of the order of millimeters. The sensor accuracy is directly related to the wavelength of the radiated EM field.

A fully automatic domain partitioning technique for the efficient circuit-level simulation of large nonlinear microwave subsystems

The domain partitioning concept is applied to the circuit-level simulation of complex nonlinear microwave subsystems of arbitrary topology, such as entire front-ends. The subsystem is described as a whole at the data entry level, but is automatically split into the interconnection of a near-optimal number of nonlinear blocks at run time. Nonlinear analysis is based on a scattering-matrix version of Krylov-subspace harmonic balance (HB) in order to avoid Y-matrix singularities in the decomposition process. This leads to high numerical efficiency and allows an excellent tradeoff between the piecewise and nodal HB approaches to be established.

Domain-Decomposition Harmonic Balance with Block-Wise Constant Spectrum

Domain-decomposition techniques provide a powerful approach to the circuit-level simulation of complex RF/microwave transceivers by envelope-oriented harmonic balance. Once the circuit has been suitably subdivided into blocks, a further significant enhancement in analysis performance may be obtained by limiting each blocks spectrum to the set of lines that are relevant to the block electrical function. Besides providing high numerical efficiency, this technique opens the way to an effective co-simulation of the RF and baseband transceiver sections

Highly Efficient Envelope-Oriented Analysis of Large Autonomous RF/Microwave Systems by a Trust-Region Algorithm Coupled with Krylov-Subspace Harmonic-Balance

The paper demonstrates the use of trust-region algorithms coupled with Krylov-subspace harmonic balance to enhance the robustness and the computational efficiency of large non-linear RF/microwave circuit analysis. The capabilities of the algorithm are demonstrated by the turnon transient analysis of an entire microwave front-end.

CAD Procedures for the Nonlinear/Electromagnetic co-Design of Integrated Microwave Transmitters

The paper introduces a new approach to the nonlinear/electromagnetic co-design of microwave transmitters including integrated antennas. Nonlinear analysis techniques and EM simulation are coupled to produce an accurate description of the transmitter performance. The antenna is not treated as a separate component but as an integral part of the linear subnetwork in order to fully account for any coupling effects that the circuit layout may introduce. For optimization purposes the passive circuit layout is described by neural models derived from rigorous EM simulation. A separate artificial neural network (ANN) is used in the neighbourhood of each harmonic of interest in order to overcome the bandwidth problem. Special ANNs are used to model the radiated field, allowing the design goals to be formulated in terms of far-field specifications, and suitable figures of merit encompassing radiation and nonlinear performance are introduced.

Coupled electromagnetic/nonlinear optimization of self-oscillating microstrip antennas with far-field performance specifications

The broadband design of self-oscillating microstrip antennas by direct numerical optimization based on electromagnetic (EM) simulation coupled with harmonic-balance (HB) analysis is demonstrated for the first time. The design goals are formulated in terms of far-field performance such as radiation intensity and cross-polarization suppression. The optimization problem is transformed into a system-solving problem, and the solution is found by a Newton iteration globalized by a trust-region method. This results in an order-of-magnitude reduction In the number of expensive EM analyses that are required to achieve convergence.