Topics

• WP 2.2.C.2 Model Verification Protocol has started
• WP 2.2.C.3 Model and Simulator Interconnection has started
• WP 2.3.2 WiSeLPAs has started


• Jobs on www.target-net.org

Events

• TARGET Tutorials
• Review of the 2006 IEE-Workshop on "High Efficiency Power Amplifier Design for Next Generation Wireless Applications"
• Review of the TARGET 3rd Summer School on "Linear Amplifier Design & Wireless Systems"

TOPICS

The intention of this WP is the development of procedures for verifying models for transistors as well as amplifiers. By help of these results, customers have a chance for the first time to compare different nonlinear models resp. different data sets for the same model. Thus, information on best available models for designing reliable circuits can be generated.

At the beginning of a new design, in many cases either no nonlinear model or more than one model/parameter sets are available. Therefore, the decision as to which model is the most accurate one is a very critical one that has to be taken carefully. Only if a proper selection is given a good prediction of the expected performance of e.g. a power amplifier can be done. Hence, the right selection will give good yield, whereas a wrong decision will enforce time-consuming redesigns or lower yield.

Within WP 2.1.2.2 (Quickshot – Model verification), a first attempt was done in order to compare different nonlinear transistor models. Several test benches have been created, which allow an easy comparison between measured verification data and simulation. In the new WP 2.2.C.2 not only single transistors but also complete amplifiers will be taken into account. Of course, basic verifications for transistors and for amplifiers are the same, but these parameters are not the key performance parameters for today’s amplifiers for communication systems! At transistor level we compare small signal S-parameters, compression, harmonics, intermodulation and DC parameters. These are typical parameters of CW operation while communication systems require information about linearity and adjacent channel power ratio with respect to the modulation scheme. Unfortunately, it is not so easy, perhaps impossible, to convert the much simpler results e.g. for intermodulation into ACPR behaviour! Therefore, it is necessary to define a set of most important communication systems, which will be considered for the development of new verification test benches. For this purpose, a schematic of an amplifier including a nonlinear transistor model will be used in order to generate data for the nonlinear amplifier (behavioural) model and also the verification “measurements”.

Already in the first part of model verification it could be seen that the selection/definition of verification ranges is very important. If e.g. the considered frequency range is quite small, a specific model may fit very well while it would almost fail for a wider frequency range. The same is true for bias conditions, temperature, RF power, impedances etc. Therefore, if a model verification protocol shall be established it must be clear how to select all these ranges. Of course, there are two problems: First, this selection will depend on the application. If another application has to be considered a new verification is necessary. Second, sufficient measurement data have to be generated. That sounds easy but it isn’t! Typically, only a subset of the required verification measurements have been performed due to limited time and money.

In addition to the verification of nonlinear models, this procedure can also be used for the determination of homogeneity of production. Due to fabrication tolerances, transistors as well as complete amplifiers differ. Comparing two sets of measurements (for different devices) will give an indication about deviations.

Dietmar Köther

back to top

The availability of efficient CAD tools is a main issue in the design of non linear microwave circuits and, in particular, power amplifiers. In fact, for efficient PA design purposes it is not sufficient to have an accurate nonlinear model for the active device, but a reliable interface with circuit simulation tools is also needed.

This WP focuses on the determination of critical issues regarding model-simulator interfacing for a wide variety of models to be implemented in the most widely used CAD tools. The main objectives of the WP are:

• Investigating the problems arising in the interconnection between LS device models and Simulation environments
• Providing evidence and documentation for the efficient and reliable use of advanced CAD tools for PA design
• Stimulating research cooperation within the Network, aimed at improving the reliability and accuracy of existing CAD tools
• Stimulating, as a collective TARGET user, SW houses to overcome drawbacks/limitations in commercially available CAD tools

The activity started officially in July 2006 and will extend for 18 months. A first WP meeting was held last September in Manchester during the European Microwave Week.

There are twelve partners from eight different European countries participating in this WP. The work has been divided into four major Sub Tasks, namely:

Task 1:Large-signal model selection.

Task 2:Evaluation of numerical simulation tools.

Task 3: Efficiency evaluation of LS model implementation in simulation tools.

Task 4: Documentation and dissemination of results.

So far, the modelling experience/background of each partner has been preliminarily surveyed so as to have an assorted number of model-simulator interconnections examined that covers the most important device technologies and modelling strategies, together with the most popular CAD simulators. In particular, efficiency and limitations in the interconnection of different types of nonlinear device models, ranging from conventional equivalent circuits to the black-box look-up-table models and the more complex non-quasi-static modelling approaches, with different simulation algorithms (e.g. Harmonic-Balance-based frequency domain, Envelope-based or conventional time domain) will be evaluated in the framework of different CAD environments (ADS, Microwave Office, CADENCE, etc..).

The first results of this activity have already stimulated cooperation between the partners, with the aim of providing information and promoting joint research activities aimed at improving the state of the art in CAD tools for microwave PA design.

Fabio Filicori, Ilan Melczarsky

back to top

1. Introduction and Objectives:

Although TARGET has been especially beneficial in connecting researchers from various microwave technologies, it has been felt through different work-packages that, due to its matrix structure, more could be done to take full profit of that potential. Indeed, it was recognized by several of its partners that the TARGET FUTURE programme was the appropriate frame to launch a transversal work in which a group of these people could be placed together to study, in an integrated way, the whole path of microwave electronics that starts at the device foundry plant and ends up in the performance of the wireless system. So, a group of 17 partners, mainly from academia, plus 2 associate partners from industry, got together to set the WiSeLPAS project.

Conceived to fulfil the underlying ideas of the EC Network of Excellence concept, WiSeLPAS aims at promoting transversal cooperative research from the physics device level, device modelling and circuit design up to the system level, profiting from the connections between research groups already built under TARGET. Creating a back and forth flow of data from the device foundry to the wireless service application, it will enable the extraction of other benefits from currently funded projects of its participant institutions and will promote competitiveness of European research groups in comparison to their north-American and Asian counterparts.

From a technical perspective, the objectives of WiSeLPAS can be summarized as follows. First, the link between the device, circuit and system design concepts should be demonstrated through the implementation and test of a linearised power amplifier presenting a state-of-the-art compromise between linearity and power added efficiency (PAE) for a selected scenario representative of the emerging wireless technology applications. For that, a dedicated test platform will be developed to demonstrate the system-level linearisation achievements.

Second, WiSeLPAS will also investigate the feasibility of future wireless transmitter architectures in which the RF signal is no longer created by a Cartesian I/Q modulator, but by a Polar modulator, and the RF PA is not the conventional linear RF transducer amplifying an AM and PM modulated RF carrier, but a highly efficient RF switch handling a constant envelope PM RF carrier dependent on a AM modulated power supply.

2. Technical Description:

Future wireless systems are expected to provide increased information capacities, lower handset costs and better integration of voice, video and data information. Each of these issues poses new and more stringent challenges on the systems’ and circuits’ design, namely, the spectral occupation and power consumption efficiency, and the coexistence of different service providers. Hence, the near future wireless services will lead to tighter compromises between linearity and PAE, which will demand for better PA linearisation schemes or even more ingenious transceiver architectures. The four research tasks of WiSeLPAS are exactly conceived to improve the design of wireless transmitters based on the present techniques of RF PA linearisation, but also to address completely different visions of wireless transmitter and PA design.

Therefore, WiSeLPAS starts with Task A - Device Processing and Characterisation, by evaluating the present possibilities of improving the compromise between transmitter efficiency and linearity, identifying to what extent it is now possible to change the device processes and what impact this has on the transistor’s behavioural characteristics.

Then, in Task B - Circuit Design and Modelling the workpackage concentrates on the PA circuit design to generate guidelines for the desirable device process adjustment and PA circuit optimisation.

Finally, in Task C - System Level Modelling and Linearisation, it will address the system-level aspects of transmitter and PA linearisation architectures. As an experimental validation step, a hardware test platform shall be designed to develop and implement efficient linearisation algorithms and to derive power amplifier specifications for optimal linearisation. With these specifications, power amplifiers shall be developed using optimised power devices (for linearity and efficiency) and advanced characterisation techniques.

Figure 1 illustrates the workflow of these three Tasks and maps it into the TARGET strands.

On what other visions of wireless transmitter and PA design are concerned, WiSeLPAS will investigate the feasibility of non-conventional transmitter architectures that are based on envelope elimination and restoration, but are being integrated with the latest digital techniques.

The idea, depicted in Fig. 2, and addressed in Task D - Emerging Wireless Transmitter Architectures, uses a polar AM/PM modulator architecture, which has recently been receiving an increasing attention from both industry and academia. The conventional linear, but inefficient, class-A or class-AB RF PA is substituted by a highly efficient RF switch handling a constant envelope PM RF carrier. Then, the AM information is impressed onto the amplified PM modulated carrier via the voltage supply of the RF switch. For guaranteed overall supply power consumption efficiency, this AM modulation path also uses switching amplification followed by an appropriate low-pass reconstruction filter.

Fig. 2 - Emerging transceiver architectures to be investigated under WiSeLPAS.

Jose Carlos Pedro, Pedro Lavrador

back to top

A new challenge is facing the TARGET community: the investigation of enabling PA technologies for the 4G wireless systems. This challenge was accepted by the new settled task force operating in the frame of the work package MUTATIS (WP 2.3.3).

Wireless technology has been evolving across various vectors from cellular to wireless broadband and to wireless personal area networks. As we move toward the third decade of commercial wireless technology deployment, there is a growing trend for a multitude of these technologies emerging on various standard platforms. This scenario reflects in the request for transceiver architectures capable to implement these new applications in an ‘always and everywhere’ global connectivity.

Extending the scenario to already experienced 3G voice/data systems, users may be moving while simultaneously operating in a broadband data access or multimedia streaming session. The need to support low latency and low packet loss handovers of data streams as users transition from one access point to another is clearly a challenging task, which may require the concurrent use of more than one frequency band at the time. For full-mobile data services, no user interaction will be required to adapt their service expectations because of environmental limitations that are technically challenging but not directly relevant to the user (such as being stationary or moving). For these reasons, the air interface must be able to anticipate the user expectations and deliver accordingly.

This is nothing else as the new paradigm of the communication technology also known as: Cognitive Radio (CR).

CR is a form of wireless communication in which a transceiver can intelligently detect which communication channels are in use and which are not, and instantly move into vacant channels while avoiding occupied ones. This optimizes the use of available radio-frequency (RF) spectrum while minimizing interference to other users. In spite of large efforts there is no consensus yet regarding the optimum multi-band radio topology. Anyway, basics level system solutions exist that provide technology concept descriptions. Generally, they refer to software-defined radio (SDR) as a radio communication system, which uses software for reconfiguration of the digital as well as analogue part for the modulation and demodulation of radio signals. A SDR performs significant amounts of signal processing, allowing the receiving and transmitting of new form of radio protocol just by running new software.

The success in harnessing the promised flexibility and incredible processing power of the SDR has led designers to consider implementing CR that adapt to their environment by analyzing the RF environment and adjusting the spectrum use appropriately. The key attributes for the successful implementation of cognitive radio are low latency and adaptability to the operating conditions. These are essential features that are needed for deployment of CRs in multi-mission scenarios such as 4G communications. CRs thus represent the chimera beyond the evolution of SDR.

MUTATIS addresses how new PA technologies may be used to implement functionality that would enable practical implementation of a basic cognitive radio using today’s SDR technology. Among the main topics focused in MUTATIS, a crucial and difficult one is the development of a linear and efficient flexible PA. The various transmitter architectures along with the linearisation scheme will also be investigated. The broadband direct quadrature conversion will represent the first choice although other emerging solutions appropriate for use in an SDR system will be analyzed, e.g. polar transmitters and LINC.

Chip designed in 0.25um BiCMOS technology to verify the main building bloks caracteristics for MMRR, implemented test structures:

Alessandro Cidronali

back to top

We would like to draw your attention to our homepage which is increasingly used as job exchange by our partner members. Currently, 6 vacancies are being offered so please check our website regularly for open positions.

Sue Ivan

back to top

EVENTS

Conventional Tutorials

Continuing the conventional tutorial series on 16 May 2006, the Tutorial on Device Modelling was held in conjunction to the MELECON conference in Málaga, Spain. The presentations focused on modelling techniques for wide bandgap devices, neural modelling techniques and model implementation issues. 28 participants attended this very interesting half day event.

In parallel to the EUMW in Manchester another conventional tutorial was held on 14 September 2006. This Tutorial on Semiconductor Materials and Devices concentrated in its first part on the technological aspects and recent developments in the field of wide bandgap devices and circuits in GaN, InN and SiC technology. In a second part the tutorial presented thermal mapping methods for the always more important thermal characterisation of devices. The tutorial welcomed 31 attendees.

Online Tutorial

TARGET TUTORIALS work package had an excellent start of is first Online Tutorial on Device Modelling in April 2006. On 7 July 2006 this success was continued by the second Online Tutorial on PA Characterisation and Testing. 4 presenters reported on their work in the field of PA characterisation starting from general aspects on PA characterisation up to built-in tests in RF circuits. The high number of registrants (57) indicated strong interest in this topic but also in this form of tutorials. Early next year we will continue these online tutorials with different hot TARGET topics.

TARGET DAYS

Next week, from 16-18 October 2006, the next event organised by the TUTORIALS work package, the TARGET DAYS 2006, will be held in wonderful Villa Mondragone in Monte Porzio Catone, Italy. In the last three years a lot of scientific achievements were made in the framework of TARGET. While the previous TARGET tutorials/workshops were organised TARGET internally, we believe that it is now time to present the excellent work done by the partners also to the TARGET ‘outside’ world. The TARGET DAYS were created with the aim to initiate a scientific dialog between TARGET and the wireless industry to discuss the technical achievements of the TARGET network and to identify future requirements and trends in mobile communication systems. Wireless industry responded with a huge number of presentations and by sponsoring the event, thus showing large interest in microwave power amplifier research and in the work of the TARGET Network of Excellence. In total 33 oral presentations from academia and industry were selected for presentation, reflecting the wide range of TARGET research topics. With over 90 attendees from 15 countries the TARGET DAYS show promise to turn into a very successful, interesting and fruitful event for the microwave community.

Markus Mayer

back to top

On May 25, 2006, the workshop dedicated to the current development of GaN-based devices was hosted by MIKON 2006 Conference in Krakow, Poland. The decision to organise such a workshop was decided commonly by MIKON Conference representatives, the GAAS© Association and the European TARGET Network of Excellence.

The aim of the workshop was to give to our scientific community an overview of the research activity in Europe on GaN from material growth to devices. 11 presentations were given by speakers representing a large panel of European laboratories involved in GaN research. Additionally to classical electronic topics, papers on high temperature, photo-detection components and new plasmonic effects in electronic transport to hopefully reach THz operation were also presented. Arvydas Matulionis from the Semiconductor Physics Institute from Vilnius, Lithuania, also gave an additional vivid talk on the effect of alloy scattering on electron drift velocity in GaN HEMTs.

• Overview of wide bandgap semiconductor materials, A. Jelenski, ITME, Warsaw, Poland
• AlGaN/GaN epitaxial growth on SiC in a hot-wall MOCVD system, E. Janzen, A. Kakanakova, U. Forsberg and I. Ivanov, Linköping University, Sweden
• Material issues on Electronic Transport in HEMT Devices, M. Uren, T. Martin et al, QinetiQ, UK.
• Processing issues for GaN-based devices targeting high temperature application, E. Kaminska, A. Piotrowska, ITE, Warsaw, Poland
• Development of the GaN/AlGaN UV photodetectors, L. Dobrzanski, ITME, Warsaw, Poland
• AlGaN/GaN HEMT : Processing & Characterisation at TIGER laboratory, S.L. Delage, M-A. Poisson, E. Morvan, B. Grimbert, D. Ducatteau, C. Gaquière, Jean-C. De Jaeger, Alcatel-Thales III-V Lab & IEMN / TIGER, France
• Microwave Power Devices Based on AlGaN/GaN HEMTs, R. Lossy, R. Behtash, A. Liero, W. Heinrich, J. Würfl, G.Tränkle, FBH, Berlin, Germany
• AlGaN/GaN HEMT Microwave monolithic Integrated Circuit Technology: Selex-SI Roadmap and Status, M. Peroni, C. Lanzieri, M.Calori and A. Cetronio, Selex-Si, Roma, Italy
• InAlN/(In)GaN HEMTs for high power applications (UltraGan project), J. Kuzmík, J. -F. Carlin, T. Kostopoulos, G. Konstantinidis, S. Bychikhin A. Georgakilas, D. Pogany, TU Wien, Austria
• Plasmonic Electronic using AlGaN/GaN HEMT for THz emission, W. Knap, Montpellier University, France

The workshop took advantage of the excellent organisation and the support of the MIKON organizing committee and allowed to gather people from different parts of the European Union to meet, exchange ideas, show their expertise and the gave them the occasion to launch new collaborative actions between partners.

Sylvain Delage, Agnieszka Konczykowska
Alcatel Thales III-V Lab

Franco Giannini, as President of the GAAS Association and Gottfried Magerl, on behalf of TARGET, would like to extend their very sincere appreciation to Sylvain Delage, Agnieszka Konczykowska and Marco Peroni of SELEX S.I. for their excellent work in organizing this workshop.

back to top

During the last week of July, 2006, an intensive four-day Summer School focused on high-level design aspects of microwave linear power amplifiers (PA) was held in the Technical University of Catalonia (UPC), Castelldefels Technical School (EPSC), Spain. The contents of the Summer School ranged from the PA specification and evaluation at system level to other issues at circuit level, such as topics related with device linearity and power efficiency improvements.

Power amplifier linearisers were the key subject of the Summer School, complemented by tutorials on behavioural modelling, as well as guided practical lectures to provide experimental skills in PA modelling. The faculty was composed of 20 lecturers from 10 different institutions.

A set of university and industry students (composed of PhD students and practicing engineers) attended the course, sharing a common interest on microwave amplifiers design. Although the courses introduced some theoretical aspects, emphasis was also given on implementation issues, a topic that completely occupied the last day of the Summer School. The first day was devoted to high-level topics on wireless systems; TX & RX block diagrams, and PA design fundamentals. The morning of the second day was dedicated to circuit-level issues: linearisation at the circuit level, boosting the power efficiency and advanced design topics. The afternoon session was focused on behavioural modelling. It started with a plenary tutorial and was split later in sub-groups for laboratory sessions: modelling PA’s at system level (Matlab) and extracting circuit models (ADS). These practical sessions continued the next day in the afternoon, while the morning session was devoted to system level linearisers, emphasizing the digital predistorters and the feedforward structures.

Attendants were very interested in the lectures and they actively participated in them. There were 51 participants out of 53 registrants from 13 countries and 25 institutions. Some sessions exceeded the planned time in order to give participants the opportunity to ask and debate with the faculty in a nice scientific and intellectual atmosphere, a fact that positively evidences the vocation of the new generation of researches, who preferred to spend more time in lectures albeit the reduced free-time for visiting Barcelona and its surroundings. Another proof of this willingness to learn more was the significant number of attendants to the optional training course on circuit/system simulations held in continuation of the Summer School.

From the questionnaire filled by the participants, three aspects outstand:

• The usage of laptops (lent by UPC when necessary) and CDs containing the learning material (instead of the classic paper-handouts): excellent or good (94%)
• The interest of the practical sessions: excellent or good (94%)
• The organisation (schedule, conference room, lunches, social dinner...): excellent or good (100%)

Furthermore, the duration of 4 days was considered adequate for the majority of the attendants.

On the last day, a certificate of attendance was given to each participant, detailing the 25 academic hours corresponding to the Summer Schools’ full attendance. This figure corresponds to an ECTS credit, which the PhD student participants can present for consideration to their respective organisations in charge of their doctorate programmes.

Eduard Bertran

back to top

PLEASE NOTE:

To receive target news regularly, please subscribe on http://www.target-net.org/press/newsletter/subscription. You can also unsubscribe from this service at any time.

Imprint

published by: ftw. Forschungszentrum Telekommunikation Wien Betriebs-GmbH, Donau-City-Straße 1, 1220 Vienna, Austria, Phone +43/1/5052830-0

responsible for content: Prof. Dr. Gottfried Magerl, Mag. Sue Ivan