10-1 (Invited)
Future of Mobile Devices - Energy Efficient Sensing, Computing, and Communication (Invited Paper) Tapani Ryh¨anen Nokia Research Center, Eurolabs, and the University of Cambridge, Nanoscience Centre 11, JJ Thomson Avenue, Madingley Road Cambridge, CB3 0FF, United Kingdom Email:
[email protected] Abstract—The mobile phone is becoming a trusted personal device with fundamental new capabilities. New form factors of mobile device and their user interfaces require new concepts for transformable mechanics. Integration of electronics and user interface functions into structural components will be necessary. Modular architecture will enable use of optimal technology for any particular functionality and optimization of power consumption. Nanomaterials, new manufacturing solutions and energy sources together with increased memory and computing capacity will enhance the capabilities of mobile devices. Nanotechnologies will also enable embedding of intelligence into human everyday environments and body area networks. We have presented a concept device called the Morph that illustrates use and benefits of nanotechnologies in real life applications. Index Terms—Mobile devices, Sensors, Cognitive radio, Nanotechnologies, Distributed computing, Ambient intelligence.
Fig. 1. Sensing, computing and communication platform based on mobile devices as gateways between local and global information [1], [3].
I. I NTRODUCTION
Context awareness, including location, is the fundamental underlying capability of the future mobile devices. These context sensitive devices will open wide range of solutions
for Internet services and mobile communication. Sensors, positioning and powerful signal processing embedded in mobile devices make it possible to detect, observe and follow different events and patterns in user’s behavior and surrounding environments with precise location. Mobile device becomes a cognitive user interface that is continuously connected to the local environment and to the Internet services. Context awareness has also profound influence on the development of future communication and computing solutions by enabling intelligent allocation and sharing of resources. Nanoscience means capabilities to image, measure and manipulate physical and chemical processes at molecular level. These capabilities convert into nanotechnologies that are based on physical and chemical phenomena that emerge at nanoscale. Thus nanotechnologies are not just a continuation of the miniaturization roadmap but offer new capabilities to create solutions for health care, information technologies, materials and manufacturing. These pervasive capabilities will affect mobile communication [2]. Nanotechnologies for sensing, computing, radios, displays, structural and surface materials will enable creative design of future mobile devices and services. In this article, we will discuss the major foreseen changes in mobile device technologies, focussing on sensing, cognitive radio and transformable devices based on nanotechnologies.
98
2009 Symposium on VLSI Circuits Digest of Technical Papers
During the following ten years mobile communication and the Internet will converge into a global information platform. Mobile phones have already become an enabling platform for digital services and applications. Mobile phones are powerful multimedia computers with wide range of functionality, e.g., imaging, navigation, music, content management, browsing, email, and time management. Increasingly they will have advanced multi-access communication, information processing, multimedia, mass storage and multimodal user interface capabilities. In the continuation these trusted personal devices will also have new capabilities, illustrated in Fig. 1: • • • •
•
Interacting with local environment via embedded short range radios, sensors, cameras, and audio functionality. Functioning both as servers for global and local internet services and as clients for global internet services. Serving as gateways that connect local information and global internet based services. Carrying the digital identity of the user and enabling easyto-use secure communication and controlled privacy in future smart spaces. Sensing local context and the behaviour of its user.
978-4-86348-010-0
II. ROADMAP OF MOBILE DEVICES Mobile communication and the Internet are converging: wireless communication will find optimal solutions based on both regulated mobile communication (3GPP track) and unregulated local access (IEEE track) solutions. Flexible and efficient local access will support sensing, computing and actuation in mobile devices that are continuously connected to the Internet services. Implementation of sensors and multimodal user interface features together with energy efficient local connectivity will enable new mobile services and new paradigms of communication, e.g., ad hoc social networking. Context awareness and machine learning will create the user experience seamless connectivity and information access but require powerful embedded computing solutions. Form factors and user interface concepts of mobile multimedia computers will vary according to the usage scenario. The tendency towards smaller and thinner structures as well as towards reliable transformable mechanics will continue. The desire to have curved, flexible, compliant, stretchable structures and more freedom for industrial design sets demanding requirements for displays, keyboard, antennas, batteries, electromagnetic shielding and electronics integration technologies. A possibility to integrate electronics and user interface functions into structural components, such as covers, will be necessary. Modular device architecture of mobile multimedia computers will consist of several functional subsystems that are connected together via very high speed asynchronous serial interfaces [5], [6]. The modular approach enables the use of optimal technology for any particular functionality, optimization of power consumption, and the modular development of device technologies and software. The same modular architecture can be extended from one device to a distributed system of devices that shares the same key content, e.g., a remote mass storage, display or a printer. III. S ENSING AND SIGNAL PROCESSING A. Sensing paradigm Sensors can already be found as key features of various battery powered, hand-held devices. Especially, location, motion and gesture recognition are new pervasive elements of applications, user interfaces and services. One of the enablers of this rapid development has been microelectromechanical systems (MEMS) based on micromachining of silicon (see a review in [4]). The need for low cost, reliable sensors for automotive applications initiated the mass manufacture of silicon MEMS sensors. The requirements of consumer electronics, especially of sport gadgets, mobile phones and game controllers, have driven further the miniaturization of MEMS devices. Today MEMS and CMOS technologies provide a solid basis for large scale deployment of sensor applications. The opportunity to connect locally measured information to Internet services and to incorporate this local information into structured global information might be even more significant. Example of benefits include real time tracking of the spread
Fig. 2. a) Sensor subsystem in mobile device. b) Wireless sensor based on similar architecture [4].
of a disease or epidemic or interpretation of changes in traffic patterns on roads through a combination of local sensors and the Internet. The Internet is becoming a massive store of heterogeneous data and linked information. Extremely efficient search and data mining technologies are creating a dynamic and real time map of the physical world with its various economical and social networks. B. Smart sensors A smart sensor system consists of transducers, signal conditioning electronics, analog-digital conversion, digital signal processing, memory and digital communication capabilities, as shown in Fig. 2. Typically, sensors and their primary signal conditioning can be integrated into the same module. An integrated low power digital signal processor core and specic hardware accelerators can be used for processing sensor signals. Feedback to the actuators controlled by the sensor signals can be computed locally to minimise control loop delays. Efficient power management is needed to control various sleep and active state of the device. Voltage regulation and system clock can be integrated into the system. General purpose signal processor can be added for more complex computational tasks. The driver for such a high level of integration is autonomous, energy efficient sensing and signal processing. In many cases, continuous measurements are needed. Modular implementation and careful design of the internal power management of the sensor module are essential: only the necessary functions are powered. Various wake-up mechanisms, based on a dened signal threshold or an interrupt from the host processor, can be used. C. Impact of nanotechnologies Nanotechnologies may not revolutionize sensor technologies and applications. Existing sensor technologies based on MEMS and CMOS platforms have not yet fully met their potential to provide sensor applications and networks that improve the human everyday environment. However, nanotechnologies, i.e., different nanoscale building blocks and fabrication processes, will affect the development of sensors, their signal processing and actuators. Nanotechnologies will
2009 Symposium on VLSI Circuits Digest of Technical Papers
99
extend the applications of sensors to new potential fields, such as smart spaces, body area networks, remote health care, and pervasive environmental monitoring (see a review in [7]). Many nanoscale sensors are related to chemical and biochemical sensing where nanoscale transducers create a possibility to derive more detailed information on observed phenomena. Nanotechnologies offer a new possibility to create nanoscale transducers, memory and computing elements and to merge these elements together to form an intelligent sensor system. The same technology, e.g., silicon or ZnO nanowires or carbon nanotubes, can be used to create various functional elements for these systems. Several possible architectures, e.g., coupled resonator arrays, nanowire crossbars, plasmonics, and spiking neuron networks can be used for both sensing and signal processing. IV. T OWARDS COGNITIVE RADIO A. Intelligent radio Cognitive radio [8]–[10] can be defined as a radio system that has capabilities to obtain knowledge of radio operational environment, monitor usage patterns and user needs, dynamically adjust its operational parameters and protocols for more efficient performance and spectrum utilization, and to learn from results of its action to improve the performance. Development happens in radio, transport and application layers creating agile connectivity to other network nodes, transport independency in terms of optimization of most appropriate transport resources and context aware applications and services. Next generation local area access solutions will converge into intelligent and flexible solutions based on cognitive radio. Another important development is the integration of wireless sensors and tags into various objects of our everyday devices, the Internet of Things [11]. Manufacturing and logistics of goods are the natural starting point of distributing digital identity and communication capability to physical objects. However, this profound capability links the artifacts to the Internet via various mobile and fixed gateways (see Fig. 1). Consequently, consumers will benefit of various possible digital services that are linked to their everyday physical objects. Cognitive radio and the Internet of Things set new requirements for future devices and their radio solutions. First of all, future devices need to be able to measure the surrounding radio spectrum. Secondly, they need to dynamically adapt to the changes in the radio environment. Implementation requires power efficient spectrum analysis together with machine learning algorithms. Furthermore, the radio front-end requires efficient tunable components.
There are some fundamental dimensional constraints that have made the integration of RF MEMS components to mobile devices challenging. The conductivity of even highly doped bulk silicon or polysilicon is not sufficient for high enough quality factor of capacitors, inductors or switches in the frequency range of 1-5 GHz. Thus the use of metal film based micromechanical structures is required, leading to several other challenges. The temperature dependence in devices that consist of multiple materials is very difficult to control and reduce. Even practical solutions have been found [13], [14]. Low cost system-in-package integration of RF ICs, passive components and MEMS devices is challenging. Voltage levels required to actuate RF MEMS devices are typical much higher than the supply voltage levels of modern ICs. Oscillators based on MEMS resonators at 10 MHz frequency range with low losses √ (Q ∼ 200000) and very good phase noise (−155 dBm/ Hz) have been developed [15]. However, the thermal stability of the devices is still a challenge. MEMS resonators are limited to roughly 10 MHz frequency. At higher frequencies, e.g., in 1-5 GHz range, the fabrication tolerance of narrow electrode gaps, very high control voltages and the lower Q values of resonators limit the development of practical devices. Recently, the development of piezoelectric actuators for MEMS resonators has improved the efficiency of electrical coupling [16]. Also the complexity of the value chain has so far lead to slower development than expected: the MEMS devices, RF ICs and RF modules have been manufactured in many cases by different manufacturers, according to the specications of the system integrators. Both technical and commercial challenges in creating integrated solutions have existed. Recent development is clearly bringing the technologies together. V. M ORPH - NANOTECHNOLOGIES IN FUTURE MOBILE DEVICES
A. Transformable device Transformation of the device can essentially happen in many levels: transformation of graphical user interface, mechanical configuration, available applications and services. The Morph device [1] is transformable in many different ways. The user interface of the device can adapt to the context of the user in terms of functionality but also its appearance. Transformability can be used to enable the ease of use of the device, applications and services. The Morph device is transformable in its form and conformation. The Morph is a cognitive user interface, capable of sensing both the user and the environment, making decisions based on this information, adapting to the context and give feedback to the user. The Morph learns about its user and becomes a trusted personal companion.
B. MEMS enabled radio front-end
B. Bifurcation of technology development
Tunable MEMS RF components can reduce necessary discrete elements in the radio front end. MEMS RF switches, tunable MEMS capacitors and MEMS resonator based devices, such as, delay lines and filters, have been the focus of extensive development in both university and industry laboratories [12].
The last forty years of development in electronics have targeted to ever increasing integration of functionality, i.e., very large scale integration. There is no doubt that this development will continue to build even more efficient solutions for sensing, computing and communication. However, interfaces
100
2009 Symposium on VLSI Circuits Digest of Technical Papers
978-4-86348-010-0
VI. C ONCLUSION We have discussed some major technological challenges related to mobile communication. The convergence of mobile communication and the Internet will bring digital services even nearer to human everyday life and the physical world. Major technology disruptions may be related to pervasive sensing, cognitive radio, distributed computing and more flexible and efficient integration of electronic functionality based on nanotechnologies. ACKNOWLEDGMENT
Fig. 3. Nokia Morph, tranformable, transparent, flexible and partly stretchable mobile device [1], [2].
of future devices with the physical world and their users require new type of intelligent and energy efficient sensors and actuators that can benefit of development of low cost electronics manufacturing and functional materials. Printed electronics creates capabilities to integrate functionality on low cost large area substrates, enabling new user interfaces, sensors and RFID tags. Functional materials research enables intelligent and responsive structural and surface materials. C. Technology requirements The Morph has some new capabilities that are not possible with the existing technologies: a flexible and stretchable device made of transparent materials with embedded optical and electronic functions. We can list some of the technology requirements: • •
• • • •
•
Transparent device with display capability Flexible and partly stretchable mechanics with non-linear spacial and directional control of elasticity embedded into the materials themselves, with even rigid-on-demand actuators Distributed sensors and signal processing in the transparent structures, e.g., pressure and touch sensor arrays Transparent and flexible antenna, electronics and energy storage Externally controllable and dynamic surface topography and roughness Multifunctional, robust surface coatings providing protection of device functionality, dirt repellence, antireflection, etc. Transformability and conformability with intelligence that can extract conformation and context and adjust the functionality accordingly
The transformable compliant mechanisms need to be built deep into the material solutions of the device [2], [17]. Complex mechanical and electromagnetic metamaterials and artificial nanoscale material structures enable controllable flexure and stretch in the macroscopic mechanisms creating the desired functions.
The author would like to thank his current and past teams at Nokia Research Center, working in fascinating fields of MEMS, sensors, nanotechnologies and their applications. R EFERENCES [1] Nokia Morph is a mobile device concept created by Nokia Research Center and the University of Cambridge, Nanoscience Centre. Members of Nokia design team were Jarkko Saunam¨aki, Tapani Ryh¨anen, Asta K¨arkk¨ainen, Markku Rouvala, Tomi Lonka, Teemu Linnermo, and Alexandre Budde. Members of the University of Cambridge design team were Stephanie Lacour and Mark Welland. First published in P. Antonelli, ”Design and the Elastic Mind”, the Museum of Modern Art, New York, February 2008. Web page: http://www.nokia.com/A4852062 [2] O. Ikkala, A. K¨arkk¨ainen, T. Ryh¨anen, M. Uusitalo, and M. Welland, Eds., ”Nanotechnologies for Future Mobile Devices”, Cambridge University Press, 2009 (in press). [3] T. Ryh¨anen, M. Uusitalo, and A. K¨arkk¨ainen, ”When everything is connected”, in ”Nanotechnologies for Future Mobile Devices”, Eds. O. Ikkala, A. K¨arkk¨ainen, T. Ryh¨anen, M. Uusitalo, and M. Welland, Cambridge University Press, 2009 (in press). [4] T. Ryh¨anen, ”Impact of Silicon MEMS - 30 Years After”, in ”Handbook of Silicon Materials and Technologies”, Eds. V. Lindroos, M. Tilli, A. Lehto, and T. Motooka, Elsevier, 2009 (in press). [5] www.notaworld.org [6] www.mipi.org [7] P. Andrew, M. Bailey, T. Ryh¨anen, and D. Wei, ”Sensing, actuation and interaction”, in ”Nanotechnologies for Future Mobile Devices”, Eds. O. Ikkala, A. K¨arkk¨ainen, T. Ryh¨anen, M. Uusitalo, and M. Welland, Cambridge University Press, 2009 (in press). [8] S. Haykin, ”Cognitive radio: Brain-Empowered Wireless Communications”, IEEE on SAC, vol. 23, pp. 201-220, 2005. [9] Joseph Mitola III, ”Cognitive Radio Architecture - The Engineering Foundations of Radio XML”, Wiley, 2006. [10] ITU definition of cognitive radio. November 11, 2008. [11] ITU report, ”The Internet of Things”, the World Summit on the Information Society, Tunis, 16-18 November, 2005. [12] V. Ermolov, H. Nieminen, K. Nybergh, T. Ryh¨anen, and S. Silanto, ”MEMS for Mobile Communications”, Part I and Part II, Circuit Assembly, June and July 2002. [13] H. Nieminen, V. Ermolov, K. Nybergh, S. Silanto, and T. Ryh¨anen, ”Micro- electromechanical Capacitors for RF Applications”, Journal of Micromechanics and Microengineering, vol. 12, pp. 177-186, 2002. [14] H. Nieminen, V. Ermolov, S. Silanto, K. Nybergh, and T. Ryh¨anen, ”Design of a Temperature-Stable RF MEM Capacitor”, IEEE Journal of Microelectromechanical Systems, vol. 13, pp. 705-714, 2004. [15] T. Mattila, J. Kiiham¨aki, T. Lamminm¨aki, O. Jaakkola, P. Rantakari, A. Oja, H. Sepp¨a, H. Kattelus, and I. Tittonen, ”12 MHz micromechanical bulk acoustic mode oscillator”, Sensor and Actuators, vol. A101, pp. 1-9, 2002. [16] G. Piazza, R. Abdolvand, and F. Ayazi, ”Voltage-tunable piezoelectrically-transduced single-crystal silicon resonators on SOI substrate”, IEEE the Sixteenth Annual International Conference on Micro Electro Mechanical Systems 2003, MEMS-03 Kyoto, pp. 149-152, 2003. [17] S.P. Lacour, J. Jones, S. Wagner, T. Li, and Z. Suo, ”Stretchable Interconnects for Elastic Electronic Surfaces”, Proc. IEEE, vol. 93, pp. 1459-1467, 2005.
2009 Symposium on VLSI Circuits Digest of Technical Papers
101
