DOT's SPACE

The NORS open source contribution (described in an earlier blog) is now also available on SourceForge!

Please download, try and provide feedback…

Lately, we see more and more of these great services for mobile devices: messengers (e.g., YahooGo!), uploading photos via FlickR, LifeBlog, permanent recording of events, wireless sensing, AI on our phones and what not. And not to mention here presence, VOIP, surfing etc. as rather basic IP services.

But I keep wondering, in particular after I’ve gotten my N80 lately. Does nobody actually consider energy management as crucial here? I mean, we’re thinking of placing all kind of stuff on the phone, assuming the ever increasing speed and memory sizes for our small little mobile gadgets. We also seem to assume the ability to run all this in a mobile (i.e., power-disconnected) way.

But it is here that we’re seemingly wrong. If we look at the development of smartphones in the last three to four years, we can indeed observe an increase in speed and memory (the new Nokia E series phones come with around 70MB of memory – my first PC had 32MB!). But the battery capacity per size has barely increased. Today, a phone like the N80 is sold with an appalling capacity of around 800 mAh. In normal usage (to me), i.e., without fancy WiFi surfing, IM and VOIP, I still end up charging the phone once a day. I do not leave anymore without spare battery in my bag in the fear to run out of power because I might use some fancy features on my great new phone. This observation is seemingly supported by forecasts around battery capacity. I just remember an article I read a few weeks ago that forecasted a mere increase of the factor of two for the next ten years in this space! So what functionality increase do we expect to place on the mobile devices with that outlook?

What I miss is a orchestrated effort that tackles this problem. And this is truly multi-disciplinary. It starts on the hardware platform, encompasses software issues but also includes interaction (e.g., developing interaction modes that simply do not require to ‘fire-up’ your display as a major power source). I have seen little in this respect and wonder why? For instance: the new smartphones from Nokia come with a small little blinking LED. It took about four years of smartphone generations to develop a mechanism that actually switches off the display as a main power source rather than displaying a useless “screen saver” (LCD screen don’t have burn-in effects in any case). Unfortunately, the LED though signals only that the device is still running but nothing else (at least in the current N80). There is hope though since the new E50 does use the LED also for signaling new SMS, MMS and emails so that you do not need to wake up the display all the time in order to check in a meeting if anything happened.

These are small steps but I see more is necessary. The outcomes of such efforts could be design guidelines for hardware and software, the development of tools that automatically create the mobile device software with energy-optimized runtimes. Interaction design guidelines as metrics for future “good software development” could be other outcomes.

I’m truly concerned that this will hit us very soon if we don’t address this problem sufficiently. Power does not exist in abundance, generally (looking at sustainability of our planet) and specifically not for mobile scenarios!

After a long uphill battle with Nokia’s legal and IPR departments, the NORS (Nokia Remote Sensing) platform is finally going public (see here)!

NORS provides a platform solution for mobile-centric wireless sensing, implemented as a centralized server with a collection of mobile devices acting as gateways towards local WSNs. This solution has been developed based on the original N-RSA (Nokia Remote Sensing Architecture) specifications (also available on the website above), work that started back in 2004 with Dana Pavel from Nokia Research and myself as project members in an internal remote sensing project. The mobile device solution is implemented in Java MIDP2.0 and provides therefore the basis for a large scale user base (several 100 million enabled phones)!

NORS is intended as a basis for research in this space and is, apart from being an OS contribution as such, seen as a contribution to the scientific community within the idea of SensorPlanet. NORS allows for easy extension of the actual platform (with LGPL enforcing to keep the platform open) while stimulating innovation on the application level. Research areas envisioned to stimulate include, for instance, research in adaptive communication (from centralized over P2P to ad-hoc), swarming, dynamic aggregation, integration of a variety of local WSNs and of course the research into particular vertical applications. For instance, NORS is used within a project at Cambridge University on pollution monitoring. In this project, bicylists and pedestrians are collecting environmental data from moving and stationary sensors, transfering the data in real-time towards a back end server, which in turn provides alerts to the mobile community. Also MIT MediaLab and other US universities have expressed interest in downloading and using the package within their efforts. Architecturally, the ideas of N-RSA and its possible extensions towards more flexible communication paradigms are currently integrated into the NSF GENI (Global Environment for Network Innovations) system specifications in a consortium of international researchers in this space.

In addition to these external efforts, an internal demonstration is under preparation. In the recent weeks, Elena Balandina, Ville Nenonen and Marion Hermersdorf have been working towards an instant meeting demonstration where a collection of stationary M2M devices (called N12 – Nokia used to sell these) is used to determine the occupancy and environmental information within the meeting rooms here at Nokia Research in Helsinki. The demo is almost finished (we’re struggling with some nasty deployment issues) and will go live rather soon for the Nokia Research audience. With this demo, employees will be able to determine the availability of meeting rooms even if the rooms are indicated as reserved in our current tools. The near term plans aim at extending this demo with WLAN indoor positioning (integrating currently in NORS), based on Marion’s work on this throughout this year, giving us the possibility to determine the nearest room available.

I hope that NORS will see a wide-spread usage in the wireless sensing community, placing mobile devices into the landscape of this research area. Open sourcing our results bears the hope of accelerating the research in this space.

There will be a separate Wiki page at the Nokia Open Source website soon. A Sourceforge project is currently under preparation and will be used as the main dissemination tool. Please check out the websites frequently. Most importantly, use NORS and provide feedback, either in Wikis, blogs or directly to me!

The Academia category in this blog serves the purpose of presenting academia work and results. It is meant for discussion but also as a little advertisement for academia work that was and is conducted within the community.

As a first item, I’d like to present the Communications Futures Program (CFP) research consortium at MIT. I’ve been involved in CFP for quite some time now (more than two years). The consortium investigates different areas of interest related to the future of communication.

CFP is organized in several WGs around topics that are deemed important to the industry and the research community. These WGs are Value Chain Dynamics, Internet Architecture (with its sub-groups on QoS, Spectrum Policy, and Inter-connection), Privacy & Security, and Viral Communication. The Broadband WG was disbanded in June this year after running about two years (led by Sharon Eisner). The main PIs in CFP are Charlie Fine (Sloan), David Clark (CSAIL), David Reed (MediaLab), Andy Lippman (MediaLab) and Karen Sollins (CSAIL).

CFP was created in early 2004 with the Value Chain Dynamics WG being its first operational WG (then called Core-Edge Dynamics WG). I had the pleasure of co-chairing this WG together with Prof. Charlie Fine from MIT Sloan until recently. For about two years, until June 2006, the work was conducted in close collaboration with the sister-program called Communications Research Network (CRN) at University of Cambridge. This relation formally ended although some of the work, such as around Spectrum Policy, will continue to be conducted jointly.

Please note that many of the results of CFP are restricted to the members of the consortium. However, I intend to discuss some findings of CFP and its research in the future on this blog in order to make the work in this research consortium more public. For more detailed information however, I will need to refer you to the individuals or the public web sites.

So for now, please have a look at the consortium’s webpage.

There’s been a lot of hype around Web2.0 and the way novel applications in the Internet are created in an ever-increasing speed.

But why only Web2.0 (and Web1.0 as its predecessor for that matter)? Isn’t the Web built on top of major Internet protocols? Do we need similar developments for the basic networking functions to ensure an evolution of the Internet beyond Web2.0?

So I came up with some possible line of evolution. It starts (rightfully as I believe) with Net1.0, the Internet before the World Wide Web, evolves towards Web1.0 (see O’Reilly) and ends up at Web2.0 as we know of today. But what comes after it? This is where Net2.0 comes in, the application of Web2.0 principles of openness, community, adaptiveness and end-user creation on networking level. Fair enough, there won’t be as much hype around this as it is for Web2.0. But it argues for necessary developments on the networking level, along the lines of Web2.0, in order to evolve towards something that one could call the Next Generation Internet.

How far out is this? Well, some of the necessary work is already ongoing (even within Nokia) and some of it will definitely reach far out.

What does it mean for the industry landscape? If we look at the implications of Web2.0 regarding the role of large Internet service providers and the transition that companies are going through, we can expect a significant impact of Net2.0 on our relation to the operators as we know of today.

But have a look at it yourself: Here’s the presentation on Net2.0 that I gave at Cambridge University in December 2005.