Mixed-signal simulation on show at DATE

IMEC has developed a design methodology for efficient mixed-signal exploration, based on the coupling of two C++-based in-house tools: Fast and Ocapi.

The design of high-performance low-cost receivers for advanced wireless LANs lies at the most challenging edge of technology and design capability.

Digital compensation techniques can either relax critical analogue design specifications or increase receiver performance.

IMEC is focusing on three important receiver problems: automatic gain control and DC offset compensation, I/Q mismatch reduction, and common phase noise reduction.

Mixed-signal exploration and simulation methodologies offer new insights and potential timesaving for the system designer.

IMEC has developed a design methodology for efficient mixed-signal exploration, based on the coupling of two C++-based in-house tools: Fast and Ocapi.

At this year's DATE Conference, IMEC will simulate a complete OFDM WLAN transceiver using its in-house tools featuring an integrated and easy-to-use Matlab/Java user interface.

The demonstration couples two IMEC-developed simulators: Fast, a high-level, mult-rate, multicarrier dataflow simulator for analogue systems, and Ocapi, the Fast counterpart in the digital domain.

The analogue blocks of the transceiver are represented with high-level models that include all major nonideal effects, such as phase noise, I/Q imbalance or nonlinearities.

This coupling, based on compact models, allows for efficient bit-error-rate simulations of the whole system.

The user interface provides all the tools required by an RF designer for architecture exploration, specifications determination or designed systems simulation.

While the actual simulation is performed using Fast/Ocapi C++ code, Java is used to describe and manipulate the data structures, and Matlab is used as a top-level layer providing a complete set of post-processing and plotting capabilities.

The Java interface of Matlab provides a seaming less integration of the Java code into the Matlab environment.

This modular framework provides for the integration of other components such as system-level and circuit-level compact modelling tools.

Its main features include: C++ code generation from a high-level description; on/off switching of different non-ideal effects using a model levelling scheme; easy storage and comparison of both the simulation results and the system under analysis; scripting of the simulation process; and easy integration of user-provided Matlab models.

Access to IMEC's Ocapi and Fast tools is now widely available, both to industrial partners involved in IMEC's wireless and reconfigurable systems industrial affiliation programs and to systems and EDA houses through technology transfer routes or commercial agreements.

Licences for the use of the extensive Ocapi and Fast patent portfolios may also be obtained.

Back to top