Huawei Technologies ERC Position Statement: Towards a Future Internet Public Private Partnership Kostas Pentikousis, Mirko Schramm, and Cornel Pampu Huawei Technologies European Research Centre Carnotstrasse 4, D-10587 Berlin, Germany Email:{k.pentikousis, mirko.schramm, cornel.pampu}@huawei.com (1) What use case and scenario in your area would you consider the most appropriate and representative one for large-scale experimentation with the Future Internet platform to be built starting from 2013? The use cases we are primarily interested in revolve around the theme of smart cities. In particular, we would like to work together on a common platform that can provide next generation content delivery services, and support ehealth, egoverment, and eeducation applications. These applications and services have many common characteristics, including a heavy media-centric orientation. In our view, the infrastructure that is already in place and will be rolled out by operators in the coming years can be capitalized upon to make our cities smarter, safer, more energy-efficient, and more open to citizens and visitors alike. Although over the last decade there have been a large number of EU projects on ehealth, egovernment, and eeducation, the results of this line of work have been both fragmented and limited in scope, did not receive wide acceptance, and failed to create the enablers that will form a unique, open platform for global deployment. Operators have invested extensively in infrastructure which is still to be fully capitalized upon in a unified manner. Instead, it has been mainly used to deliver flat-rate mobile broadband services, which are forming a business cycle of accelerating diminishing returns. This is unsustainable in the long-term and threatens to choke future innovation and infrastructure deployment. We believe that Europe can lead the way in developing an open network architecture of services where infrastructure providers will be vital in developing smart cities and will be compensated according to their contribution to the citizens well-being. We propose to go beyond point solutions in these domains and develop a common platform that will be suitable not only for creating the future European smart cities but smart cities around the world, in a fashion similar to the way previous European network architectures, such as GSM/UMTS, managed to dominate mobile communications in the past. 1
(2) What innovative Internet functionality and technologies would you consider important for your suggested use case and scenario (e.g. context awareness, sensor networks, advanced real time processing capabilities handling huge volume of data, ad hoc service composition and mash-up, managed broadband connectivity and services, embedded media support for interfaces easing the interpretation of processed contextual data, etc.)? In recent years, a common success factor in the Internet has emerged. It is the development of architectures for applications and services that although at their very core remain proprietary and close-guarded, they provide hooks for third parties to develop further services and applications in a rather liberal fashion. This is based on APIs, SDKs and protocols/interfaces that expose part of the full functionality and increase dramatically the value of the overall proposition. Large distributed computing platform operators, ranging from Google to Amazon to Facebook have successfully used this model to greatly enhance their offering. In contrast, telecommunication operators and in particular mobile network operators have deployed infrastructure based on open standards but without any hooks for further application development. Location services easily come to mind, for example. Although a mobile operator has much finer granularity information at its disposal, applications cannot use it. Context awareness, for example, is nearly non-existent, primarily application based and mainly relying on presets and static preference lists. We argue that future smart cities applications and services will benefit immensely from an interface to the network infrastructure which enables unobtrusive context-awareness, fine-tuned for individual users, not aggregate user groups. This level of granularity as well as privacy cannot be based solely on distributed computing/cloud infrastructures located far away delivering services over bare bones mobile broadband connectivity as is done today. Instead, what is needed is a combination of local, network, and cloud resources in order to capitalize on the all-ip infrastructure to deliver context-aware services in a variety of domains. Machine-to-machine (M2M) applications and services based on wide deployment of sensors and actuators, for instance, are expected to storm our homes and cities in the coming years. Yet, without ways to, one the one hand, orchestrate operator and usercontributed resources and, on the other, create mash-ups of applications and services the huge amount of additional information cannot be exploited to its full potential. What is more, if operator resources are not optimally utilized we will very soon hit a brick wall with respect to our ability to continuously deploy exponentially increasing capacity. Infrastructure cannot continue to be treated as a passive cut-through were traffic volumes explode. As we revisit the communication/computation/storage tradeoffs and consider energy efficiency as a KPI, we need to be able to handle the expected data growth in a short time span using local and remote resources, using a common network to service interface. Finally, we need to highlight the potential that such a standard but to-bedeveloped interface will have in the midterm. In today s Internet-based services, widgets, based on standard technologies and open APIs to distributed computing resources, have become a staple, used in a variety of modern web, mobile and desktop applications extending the reach of cloud resources. Future widgets created for smart city applications, content distribution/online storage/cloud-based services will be dramatically more sophisticated if they can access information available in the mobile operator network. On one hand, this information can be used to better provide managed broadband connectivity. 2
On the other hand, new services can be rolled out much faster and mobile network operators can increase revenues by monetizing information they manage in their network (e.g. user identifier, security and privacy; user location; statistical information) to 3rd parties. Previous FP research projects have touched upon this issue, but never really worked out how to implement a common network to service interface. Creative mash-ups based on combinations of local, network and cloud resources will be the ultimate enabler for future smart cities, where innovation can be based on both technology push and demand pull. (3) Which of the identified functionalities would you expect the Future Internet core technology platform to deliver to support your and other usage area scenarios? First and foremost, the Future Internet core technology platform will comprise all the functions which are required to run a telecommunication network and especially, a mobile network. Of course, we do expect enhancements compared to e.g. existing 3GPP Release 9 standards to address new aspects like the order of magnitude increase of end devices (M2M), dynamic network composition (mobile sub-networks, local access network operators) and a more efficient handling of the massive increase in data traffic (new services and applications, local and user provided content). In addition, the Future Internet core technology platform should offer generic enablers for managing the treatment of end devices in the access network in a service and application specific manner. This could entail the controlled use of all available access technologies, flow routing especially with regard to the selection of the inter-operator peering points (also known as local breakout), Quality of Service guarantees, and end device configuration. Data management and rights management play an important role here as the network infrastructure becomes far more empowered than it is today and have to be considered as part of the generic building blocks when designing the common platform. In addition, information about the user and the access network (e.g. user location, network load, end device capabilities) could be offered to the services and applications to broaden their usage area. Another aspect that smart cities need to incorporate is making sure that communications are secure and privacy is respected throughout. In future smart cities, end user will interact with a host of services and objects. Current technologies, often left unused, are well-tested and proven robust in a variety of scenarios, but infrastructure providers will have to explore many more opportunities. SIM cards along with existing mechanisms are a valuable asset which will have to be enhanced in order to provide optimized solutions for the new opportunities arising. In particular, identity and access management as well as accounting, rating and billing in the common platform will require the definition of generic functional blocks which along with the rest of the enablers will facilitate the fast roll-out of new applications and services for the smart cities of the future. Note that this refers to dynamically creating mash-ups similar to the way widgets work today based on the common platform de-facto defined by web technology standards and open APIs from distributed computing providers. To sum up, we expect a variety of applications/services that can be built based on an open platform where the aforementioned generic enablers can be mixed and matched to create new offerings with a common network to service interface. The new interface 3
would allow services to interact with the network operator infrastructure in a dynamic way and thus will cater different needs, user groups and application domains (ehealth, egovernement, eeducation). Furthermore, the Future Internet core technology platform ought to enable making the most out of available information, which currently remains hidden and underutilized and provide the means to expose them in a controlled manner so that infrastructure can be optimally used. (4) What kind of experimentation environment would you consider necessary for broad large scale testing of the platform to be developed in your use area? What would be needed to experiment new services and applications cutting across use areas (services and application mash-up) and building a new services and application ecosystem around the prototype implementations of the platform? Huawei Technologies can provide a distributed test-bed environment spread across several European cities which is ready for testing a range of new services which can form an application ecosystem. In particular, the prototype implementation of a Future Internet core technology platform is a key priority in the context of this joint effort. Telecommunication infrastructure components are available as well and can be used and enhanced when necessary to form the foundation for the platform. With regard to our preferred use case scenario ( Smart City ), we envisage a prototype implementation which realizes optimum access to content through local and remote content distribution mechanisms. The location of the end device in the (mobile) operator access network will play a role and steer the selection of the local content store as well as the selection of the inter-operator peering points. The format of the content could be chosen according to the end-device capability. The characteristics of the content and the associated QoS requirements could influence the use of the optimum access technology in the foreseen common use case where multiple network accesses and paths are available. Specific settings for QoS and charging can be triggered depending of the nature of the content as well. Different schemes for content retrieval, compensation, and of course authentication and access management could be realized to adhere to a variety of business models for operators and content providers based on the generic enablers of the common platform. In short, content distribution seems to be an excellent use case for evaluating a flexible, dynamic and context aware interaction between the service layer and the network operator infrastructure. In effect, by testing this use case a range of ehealth, egoverment, and eeducation applications will become easily deployable using mash-ups of the generic platform enablers. 4
(5) How do you see the potential role of your organisation in the FI-PPP, in the context of Usage areas taking a prominent role in the Initiative, to ensure an appropriate application driven approach? The Huawei Technologies European Research Centre (ERC) is a rising first-class contributor in the European R&D scene. Due to our experience in the telecommunication infrastructure market, Huawei Technologies, through ERC, can contribute to the Initiative in conceptual design, functional description and protocol implementation of the Future Internet core technology platform. Standardization and interoperability of smart city systems will be key for the success of the developed functionality and solutions in the real world, well beyond what previous academic-mainly efforts have achieved. We care about creating a common platform with a real value proposition to citizens and business stakeholders alike. To this extent, standardization is vital. Huawei Technologies actively participates in 91 international standardization organizations including ITU, 3GPP, 3GPP2, ETSI, IETF, OMA and IEEE and will play a very active role in pushing forward standardization and harmonization of the envisioned smart city solutions. As we strive to develop a truly universal platform for world-wide deployment, Huawei Technologies ERC can play an instrumental role in design, implementation, experimentation and standardization bringing into the consortium a unique point of view for a truly scalable word-wide accepted architectural solution. 5