Lund University breaks the world record for transistor performance
A team from Lund University has developed transistors with a new world record of performance. The improved performance can be translated to a reduction of power consumption in computer and mobile devices, allowing for new features, more memory and longer battery life time.
The transistor is the smallest constituent of computer processors (CPUs) and memories (RAMs), and the continued development of it, since its invention in 1959, has ushered in the digital era.
“The primary transistor performance metric is the amount of current that is able to pass through the transistor under a certain voltage. This, in turn, depends on many factors, such as material properties and device design”, says Cezar Zota, who presented the work together with Erik Lind, at the 2016 Symposia on VLSI Technology and Circuits in Honolulu, Hawaii.
With improved transistor currents, the voltage powering them can be lowered, so that the current falls down to the minimally required level. The net gain is a reduction of the power consumption of the circuit which implements the transistors.
“In our research, we use so-called III-V compound semiconductors, rather than the silicon-based alloys used in industry. III-V materials have the potential of improving transistor performance, but also present new challenges to be solved”, says Cezar Zota.
With these results, Lund University now holds the world record in transistor performance, beating stiff competition from both academy and industry.
Text and Picture: Erik Lind
A new take on asleep and awake
-Congrats Nafiseh! You are a doctor! Tell us, did you have a nice party after the duties during the big day? -We had a dinner but no big party. Maybe later... Now it's relax and vacation time. It's been a tough spring.
Preparing for thesis presentation
Assistant Professor Sofie Pollin, KU Leuven, Belgien opened the event and the well attended E:1407 learned that modern wireless sensor devices are often constrained by batteries that determines the size of the device. Even though the nodes in a networked sensing system are mostly asleep and doing nothing, they have to wake up now and then. Waking up and checking activity consumes power which also goes for the processing and sending of data. Finding clever approaches on scalable, powertight wake-up schemes for wireless sensing systems is what our latest doctor Nafiseh Seyed Mazloum has been occupied with in her PhD-research.
Opponent Sofie Pollin (left) and Nafiseh Seyed Mazloum (right).
Nafiseh presented her thesis "Duty-cycled Wake-up Schemes for Ultra-low Power Wireless Communications" and talked about network topology, data traffic, cross-layer communication protocol solutions and more, all targetting extremely powertight wireless sensing networks.
Nafisehs approach to describe her research in a popular way is the following:
-Whenever you have a very small resource of anything, you want to use it efficiently. -My research addresses very small devices like pacemakers, hearing aids etc. Things that need to communicate without having a lot of battery. That's what I work on. This way of putting it usually works at a dinner conversation. -It's a bit like telling someone that you should switch off your mobile phone screen to make the batteries last longer. If the listener wants to know a little bit more, I continue with "standby-mode" and explain that entering this is done autonomously and not by a press of a button.
Nafiseh and LTH Dean Viktor Öwall
-DCW, or spelled out as "Duty-Cycled-Wakeup" means that you send your receiver to standby-mode in duty-cycles where it's on for a period and off for another, says Nafiseh. -Waking it up needs to be done autonomously since noone human is there to do so. Energy consumption also relate to communication delay so when sleep-time is long, the communication is delayed. -And duty-cycle settings depend on application and use cases, she tells us.
On the opponents question why a 65 nm process is used to verify the functionality of the circuits and not something more state-of-the art, Nafisehs opinion was clear.
-Well, whether to use older or newer technologies can always be discussed but I used the technology that was easily available at the time of testing. The focus of my particular research was not on state-of-the-art CMOS-technology, rater on proof-of-concept of the wake-up functionality.
From engineering in Teheran to research in Lund
Nafiseh worked as a hardware engineer in Teheran and then moved to Sweden 11 years ago where she did her master studies and PhD. The master thesis however, was performed at Philips research in the Netherlands. It all ended up with a PhD-position at LTH. The background in low-power electronics was a good match for this and Ove Edfors has been her supervisor (a very good one) ever since.
Supervisor Ove Edfors and Nafiseh Seyed Mazloum
Swedish integration going on
Nafiseh is now looking for new opportunities, probably in a research institute or similar. Engineering in industry is over and done with for the time being. -Sweden is nice, I'd like to stay here, she says. -And living in Lund represents a cosy lifstyle. -My swedish will improve now when I've more time, she promises.
Art and sport during leisure hours
Any life outside work? -Yes! Painting in watercolour and also consuming any type of art. Visiting Louisiana Museum of Modern Art in Denmark is a must at least once a year. And I do like exercising in any kind of sport even though on an amateur level. But trying to keep in shape is important for me.
Gender balance potential
An encouraging fact with your disputation is that we saw a female opponent, two women in the committee and last but not least you defending your thesis. Pretty good marks on gender balance in this case compared to most disputations in our area! Gender balance is addressed more and more in the society on all levels and we know that there is a lot of talk but not always much walk-the-talk. How would you say LTH/EIT perform? -I've enjoyed my time and colleagues at EIT a lot but we are not champions in this particular sport. Nafiseh ends the conversation by diplomatically mentioning that there is potential to build a good culture here. As long as there is will.
Congratulations from EIT and good luck in the future!
Text and Picture: Anders Borgström
2016-06-17 Master Thesis presentatio by Sebastian Andersson and William Tidelund
Date: June 17, 2016
Title: Adaptive Beamforming for Next Generation Cellular System
The work has been performed at Ericsson AB in Lund.
In this work a Matlab model of a simpliﬁed LTE system has been implemented. The PUSCH and PDSCH signal chains has been used for reception and transmission of data and DM-RS symbols are used as pilots. Moreover does the model support communication and interference between multiple users and base station antennas. The wireless channels are modeled as multi-path Rayleigh fading and are continuous in time such that multiple frames can be transmitted on a correlated channel. Three standardized multi-path delay proﬁles have been used for modeling users in pedestrian, vehicular and urban environments. Three beamforming algorithms have been implemented, maximum ratio transmission, zero-forcing and regularized zero forcing. This model is an extension to the current standard of LTE in the sense that the parameters of the model are scalable beyond what is currently in the standard. The different algorithms are compared in many different scenarios, including different modulation levels, delay proﬁles, number of user sharing the same resources, number of base station antennas, multi-layer transmissions and complexity. Maximum ratio transmission is shown to be computationally much faster , while the zero-forcing algorithms is better at removing inter-user interference, especially as the users sharing the same resources grows for a constant number of base station antennas. Regularized zero-forcing is shown to outperform the other algorithms when looking at the entire SNR range.
Anders Karlsson tar guld i institutionsgolfen
Anders Karlsson tar guld i institutionsgolfen
Årets institutionsmästerskap i golf avgjordes torsdagen den 9 juni i strålande sol på fantastiskt fina Erikstorpsbanan, Landskrona GK. Efter en hård batalj kunde fyra medaljörer koras efter förnämliga insatser. De fyra är:
Guld och institutionsmästare: Anders Karlsson
Stora silver: Ulf Körner
Lilla silver: Richard Lundin
Brons: Bengt Mandersson
Hål 3: Richard sätter en lång putt och får fyra poäng på hålet. Det är dagens högsta poäng. Grattis till denna prestation Richard!
Hål 10: Skyltarna för väganvisningar är inte de bästa och det gör att förvirring ibland uppstår i ledarbollen. I bollen efter ledarbollen går en grupp av fyra damer som från och med hål 10 ser som sin uppgift att hjälpa ledarkvartetten med orienteringen. Så fort tvekan uppstår kommer därigenom en dam resolut rusande över fairway och ger instruktioner. Mycket praktiskt.
Hål 13: Anders slår ut först och får till en lång men mycket slicad drive och hamnar i ett svårt läge, på tok för långt till höger om fairway. Ulf slår ut som andreman och även han slicar grovt och hamnar långt till höger om fairway. De båda herrarnas lägen förbättras dock väsentligt när Bengt förklarar att de slagit ut mot fel fairway. Det slutar med ett par för Anders och en bogey för Ulf.
Hål 14: Birdie från damtee av Ulf Körner! Ulf slår i vattnet på detta svåra par tre hål. Han droppar på damtee, med ett slags plikt, slår upp bollen på green och sätter en 10 metersputt.
Hål 15: Richard hålar ut efter 12 slag. Dagens högsta siffra. Heder åt Richard för att han minsann inte ger upp i första taget!
Efter avslutad tävling avnjöt de fyra pristagarna en fantastisk lunch på Erikstorps Kungsgård med Kokt rimmad kalvbringa med kräm på örter, vårprimörer i dillsmör och potatispuré. Erikstorps Kungsgård ligger i direkt anslutning till golfbanan. Även icke-golfare kan rekommenderas denna pärla.
EIT projektleder open-sourcebaserad innovationsupphandling
EIT har tillsammans med LU Open, LU Byggnad, Inköp- Upphandling, MAPCI och Datordriftgruppen deltagit i en för LU unik Vinnovastöd innovationsupphandling. På sikt syftar upphandlingen till att åstadkomma ett open-sourcebaserat system med stor potential att samordna, förändra , förenkla och förbilliga administration och användning av olika organisationers lås- och passagesystem. Två slutkandidater för leverans av skarp prototyp har valts ut, Sensative AB från Lund (Årets skånska startup) och Osser Brosoft AB från Stockholm. Teknisk projektledare från EIT är Anders Borgström.
Mohammed fights low-voltage linearity issues
Mohammed Abdulaziz ready for his PhD-thesis presentation
What is it all about?
Mohammed Abdulaziz recently presented his PhD-thesis in an almost full E:1406. The title of his thesis is "Linearity Enhancements of Receiver Front-End Circuits for Wireless Communication" and for those who want to dive deeper into the knowledge he gained during his research-time, a quick glance at the abstract give a hint of what's to be found in the full thesis. The thesis is based on 5 papers covering topics like positive feedback of operational amplifiers, LTE-channel select filters, feed-forward linearization, low noise amplifiers, spectrum sensing and other appetizing stuff.
OK, and what does that mean?
For us "normal" people, not prepared to try to follow all details in the thesis we ask Mohammed about a high-level and more popular description of his area of research and we get the following story:
Demands on high-end wireless handsets (e.g. smart phones) are ever increasing. Fast data transfer, power efficiency and "always-connected" are properties that are regular subjects of improvement and also not to be traded away.
Well attended in E1406
Carrier aggregation is hard
Band fragmentation, where cellular frequencies are divided between different operators, requires sometimes special measures where several narrow channels are combined, called carrier aggregation. This is a complex thing to handle by the hardware.
Opponent checking paper four
To squeeze in more transistors on a given chip-area will mean smaller transistors, lower power consumption, improved speed but also reduced supply voltage. The latter has a negative impact on signal quality because of worse linearity of the transistor in this regions. Higher distortion is inevitable.
Stronger signals taking over
Stronger signals can also saturate the receiver preventing reception of the wanted signal. This problem is similar to trying to listen to one talking while a foghorn is sounding at the same time.
Committee at work
Verify with measurements on chips
Mohammed has addressed the problem of receiving weak signals in the presence of strong interference and three channel select filters are designed and implemented with improved linearity. A full receiver front-end for carrier aggregation with spectrum sensing is also implemented. These solutions are analyzed and simulated, and chips are fabricated and measured to verify the performance and compare to the analysis.
Last minute check
Briefcase guided me to the PhD-track
Since I was a child I have always wanted to finish my postgraduate education, I think my mother was the one who motivated me the most. She promised to give me the leather briefcase when I asked her for it, but “after you finish you PhD” she added.
From Yemen to Lund via Erasmus program
I come originally from Yemen, did my bachelor degree in Jordan. After finishing my bachelor degree I worked for the university in my home town (Dhamar) as a teacher assistant. The Erasmus mundus program started and I was interested in doing my PhD. I saw that Lund University was a partner. I was very interested in LTH because I knew at that time that Bluetooth technology was born in the same department. I applied for my first and second wishes to be LTH and the third one for French university. I think if there was a third option in LTH I would have applied for it as well.
Future is like good unwritten book.
Currently I am engaged in a project in the department, you will notice I am still around in the lunch room and the research lab. After that future is not so clear but promising.
EIT says congratulations and good luck in the future!
Text and Photo: Anders Borgström
Färs och Frosta’s research prize to Xiang!
Färs och Frosta’s research prize to Xiang!
At the new engineering graduates ceremony last week, Sparbanksstiftelsen Färs och Frosta’s research prize went to EITs newly PhD-graduate Xiang Gao. Read more about Xiangs PhD-graduation here and about the prize ceremony here. Congrats to Xiang!
Text: Anders Borgström. Photo: Gunnar Menander
Exjobbare från EIT vinner SER- Prize.
|Max Andersson||Gabriel Jönsson|
Sveriges Elektro- och Dataingenjörers Riksförening, SER.se, har i år tilldelat sitt SER junior Prize, för årets bästa exjobb, till Gabriel Jönsson och Max Andersson för sitt examensarbete runt en FPGA baserad PCI Express controller med motiveringen: Att kunna styra snabba processer i realtid kommer alltid att behövas inom processindustrin. Handledare var Fredrik Kristensen
Grattis från EIT!
EIT records in spectrum efficiency.
The Lund-Bristol team (left-to right): Benny Chitambira (Bristol), Paul Harris (Bristol), Wael Boukley Hasan (Bristol), Liang Liu (Lund), Joao Vieira (Lund), Steffen Malkowsky (Lund), Siming Zhang (Bristol), Mark Beach (Bristol)
Lund University (EIT) and Bristol University together break new grounds in the field of spectrum efficiency with a 128-antenna array in a prototyping platform from National Industries.
Compared to existing 4G-platforms, the team shows around 22 times better capacity using 256-QAM modulation, a shared 20 MHz radio channel at 3,51 GHz and a 128-antenna M-MIMO-array. Eight researchers and postdocs from Bristol and Lund demonstrated this and from Lund Liang Liu, Steffen Malkowsky and Joao Vieira participated.
The team beats their own earlier record and another one from Facebook with 24 users and a 96-antenna array. A 1000 fold increase in netwowrk capacity compared to 4G is claimed needed to meet future 5G-needs. Massive MIMO, mmWave and beamforming are technologies that will contribute and our Lund researchers are key persons.
Jun Wu, a proud nanoelectronic researcher.
Jun Wu, preparing for his thesis presentation. Photo: Anders Borgström
MOSFET-evolution going nano
-I am quite proud of my research, says Jun Wu who is EITs most recent PhD-graduate. -And my research covered a quite broad area.
In his thesis, titled "Vertical III-V/High-k Nanowire MOS Capacitors and Transistors", you can read that the transistor gate geometry has evolved from planar technology through doublegate-, tri-gate up to what Jun is now studying, namely GAA-technology. GAA stands for Gate-All-Around and using that means a rethink around the most fundamental buildingblock of electronics, the MOSFET.
-In general, we would like to realize a new type of MOSFET using standing III-V compound semiconductor wires called ”nanowire” (below 50nm) and try to further develop circuit implementation using the new type MOSFETs, says doctor Wu. -To do so is beneficial since one can further increase chip packing density and performance, he continues.
Many fields addressed
His research has been involved in several fields as...
...the field of growing nanowire crystals where he has developed recipes to grow defect-free nanowire crystals.
...the field of optimizing the MOS-gate electrode where he realized the detailed characterization and improvement for MOS-gate electrode for the first time.
...the field of fabricating and improving the MOSFET where he cooperated with his colleague to develop new type of nanowire MOSFETs with improved performance
...the field of developing circuit implementation where he realized 3-D cricuit integration along nanowires.
Again, a quite broad area has been addressed.
Jun Wu and committee. Photo: Anders Borgström
Research fun but applications also important
-Doing research may be the most interesting job in the world, since you get a chance to work in the frontier of mankind’s knowledge, and try every now and then to make breakthrough out of it, Jun says.
The subject of nanoelectronics is attractive to Jun because it is a branch of the emerging promising nanotechnology topic, and also beause it's closely related to industry implementations rather than pure scientific research. -The practical motivation is a need for me, says Jun.
From China to US via Sweden
Mr Wu, who comes from Beijing in China is interested in reading and traveling and his reasons for choosing LTH is Lund Universitys excellent research reputation within the field of Nano science and technology. -A field I'm interested in, he underlines again.
The future plans includes visa-applications to U.S.A. in order to pursue a post-doc offer by UCSB (University of California, Santa Barbara). Further to that, no longer term plans are in sight.
EIT congratulates and wishes all success in the future.
Text: Anders Borgström
Career talk with EITs new professor
Professor is a useful title
-And a career step which I’m glad to take even though my daily work will not change dramatically. Our new professor Buon Kiong Lau, or Vincent as most of us know him, says he is happy to be promoted as he grabs the coffee-mug and prepares for a chat on academic careers. The importance of reaching the professors level shall not be underestimated. -It is a good statement of qualifications and you don’t need to explain all your merits to everyone, he says.
Vincent realized that he was approaching the professors title around 2012 – 2013 -Enough qualifications and merits in place so I decided to go for it. Guidelines and checkpoints have progressively become stricter and it’s the Deans decision to allow the application by writing a letter of support. After improving a little on lower-level education 2013 the gap was filled.
The name Vincent does not appear in either his publications or his passport so why Vincent? -The reason I have this nickname is because my uncle suggested it being born around St Vincent’s day. As simple as that.
Research and education is of equal importance…
He emphasizes the importance of both research and education but admits that research is a little bit more fun. For Vincent, both sides have of course been mastered throughout the years as well as fund chasing, PhD-supervisions, IEEE-editorial tasks, directorship of postgraduate studies and other important duties being expected in higher level work.
…as well as science outreach (tredje uppgiften)
“Science outreach” i.e. how to “deliver” back to society is on the checklist for reaching professors level. Is that close to your heart? -I like to do research that is potentially fulfilling corporate and societal needs and challenges, says Vincent. But if I need to be go into more fundamental research in order to solve a problem, I’ll do it. Fundamental insights are necessary to be able to do more applied research.
Advice to a graduate
-A masters student that really enjoys the master thesis work will most likely be a suitable candidate for an academic career, Vincent let us know. You need passion for writing or at least consider it to be a natural part of the job if the academic world should be enjoyable. A potential PhD-students need to have a good attitude to research and if you can’t handle a 5 month project (Master Thesis) then you will not survive a 5 year research project.
All career steps count
In hindsight, which step in your academic career would you value the most? –Hard to say since every stage has its value, Vincent says. But my new role will hopefully mean more scientific leadership. My current research field is neglected by funding agencies.
Learn from Spain and Italy
What about gender balance among electrical engineering? –Well, I do my best in encouraging female students, says Vincent. -My first PhD-student was female and so many of the visiting students I have supervised. So I think my track record is pretty ok in that respect. –But even though equal opportunities, it takes time to change cultures, traditions and attitudes. Ending up with 80% male student buddies by selecting an engineering track may prevent many from choosing it. And role models matter. If mums, aunties and friends are not engineers, perhaps a young woman hesitate. –We could learn from Spain and Italy, Vincent suggests. They have a lot of female electrical engineers for some reasons.
Performance and design must walk hand-in-hand.
Also consumers needs to be enlightened in Vincents opinion. -We choose phones based on look and feel rather than on performance, he says. 2 Mb/s instead of 20 Mb/s seems to be ok for many as long as the phone is shiny with cool apps. As a professor I’d like to engage more in this and promote in a more balanced way. Design yes! But also perfomance, yes!
A bad antenna destroys a system
Antenna systems including interaction between antennas and the channel and focus on system performance has been Vincents silver thread in his research. His background in signal processing enables another important building block in achieving better data rates and throughput. Failing terminal antennas may destroy a 5G-performance. –We are benchmarking very well, claims Vincent. Close collaboration with propagation channel research projects and good experimental facilities makes EIT an excellent research environment within the field.
From Malaysia to Lund via Australia and Blekinge
Born in Malaysia and studying in Australia for 10 years took Vincent to Blekinge and then eventually as a postdoc. -Sweden is a good place to have a family and good career opportunities, says Vincent. The sometimes bad weather is nothing fixable so we just have to accept it.
Vincents spare time is mostly spent on his family with 9, 6 and 3 year old kids. Boy, girl, boy. Will some of them become engineers? –My oldest son is interested in technical things. There’s hope…
How to save the world with 10 dB
Asking Vincent finally about long term visions for the world we get an engineers answer: 10 dB improvement can be easier to achieve with an excellent antenna design than 1 dB with signal processing. Wise words from our new antenna professor.
Congratulations from EIT!
Text and photo: Anders Borgström
LuMaMi on SVT
Fredrik Tufvesson interviewed by SVT (Swedish Television) regarding the next generation of (wireless) Internet.
See the interview here (in Swedish):
Text: Johan Cedervall