Network Working Group H. Tschofenig
Internet-Draft Nokia Siemens Networks
Intended status: Informational J. Arkko
Expires: March 03, 2012 Ericsson
August 31, 2011

Report from the 'Interconnecting Smart Objects with the Internet' Workshop, 25th March 2011, Prague
draft-iab-smart-object-workshop-02.txt

Abstract

This document provides an overview of a workshop held by the Internet Architecture Board (IAB) on 'Interconnecting Smart Objects with the Internet'. The workshop took place in Prague on March, 25th. The main goal of the workshop was to solicit feedback from the wider community on their experience with deploying IETF protocols in constrained environments. This report summarizes the discussions and lists the conclusions and recommendations to the Internet Engineering Task Force (IETF) community.

Status of this Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet- Drafts is at http://datatracker.ietf.org/drafts/current/.

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."

This Internet-Draft will expire on March 03, 2012.

Copyright Notice

Copyright (c) 2011 IETF Trust and the persons identified as the document authors. All rights reserved.

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (http://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Simplified BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Simplified BSD License.


Table of Contents

1. Introduction

The Internet Architecture Board (IAB) holds occasional workshops designed to consider long-term issues and strategies for the Internet, and to suggest future directions for the Internet architecture. This long-term planning function of the IAB is complementary to the ongoing engineering efforts performed by working groups of the Internet Engineering Task Force (IETF), under the leadership of the Internet Engineering Steering Group (IESG) and area directorates.

Today's Internet is experienced by users as a set of applications, such as email, instant messaging, and services on the Web. While these applications do not require users to be present at the time of service execution in many cases they are. There are also substantial differences in performance between the various end devices, but in general end devices participating in the Internet are considered to have high performance.

There are, however, a large number of deployed embedded devices and there is substantial value in interconnecting them with the Internet. The term "Internet of Things" denotes a trend where a large number of devices employ communication services offered by the Internet Protocols. Many of these devices are not directly operated by humans, but exist as components in buildings, vehicles, and the environment. There is a large variation in the computing power, available memory, (electrical) power, and communications bandwidth between different types of devices.

Many of these devices offer a range of new possibilities or provide additional value for previously unconnected devices. Some devices have been connected using proprietary communication networks in the past but are now migrating to the use of the Internet Protocol suite in order to share the same communication network between all applications and enabling rich communications services.

Much of this development can simply run on existing Internet protocols. For instance, home entertainment and monitoring systems often offer a web interface to the end user. In many cases the new, constrained environments can benefit from additional protocols and protocol extensions that help optimize the communications and lower the computational requirements. Examples of currently ongoing standardization efforts targeted for these environments include the "Constrained RESTful Environments (CoRE)", "IPv6 over Low power WPAN (6LoWPAN)", "Routing Over Low power and Lossy networks (ROLL)", and the "Light-Weight Implementation Guidance (LWIG)" working groups at the IETF.

This workshop explored the experiences of researchers and developers, when considering the characteristics of constrained devices. Engineers know that many design considerations need to be taken into account when developing protocols and architecture. Balancing between the conflicting goals of code size, economical incentives, power consumption, usability and security is often difficult, as illustrated by Clark, et al. in "Tussle in Cyberspace: Defining Tomorrow's Internet".

Participants at the workshop discussed the experience and approaches taken when designing protocols and architectures for interconnecting smart objects to the Internet. The scope of the investigations included constrained nodes as well as constrained networks.

The call for position paper suggested investigating the area of integration with the Internet in the following categories:

The goals of the workshop can be summarized as follows:

Note that this document is a report on the proceedings of the workshop. The views and positions documented in this report are those of the workshop participants and do not necessarily reflect IAB views and positions.

2. Constrained Nodes and Networks

An observation that lead to the scheduling of the workshop was the presence of constrained devices that are more and more interconnected to the network. So, it is quite natural to ask how these limitations impact the design of the affected nodes. Note that not all nodes suffer from the same set of limitations.

      +---------+------------------+--------------------+
      | Class   | Volatile Storage | Persistent Storage |
      +---------+------------------+--------------------+
      | Class 1 | ~ 10 KByte       | ~ 100 KByte        |
      |         |                  |                    |
      | Class 2 | ~ 50 KByte       | ~ 250 KByte        |
      +---------+------------------+--------------------+
          

Energy constraints:


Since wireless communication can be a large portion of the power-budget for wireless devices, reducing unnecessary communication can significantly increase the battery life of a low-end device. The choice of low-power radio being used also has a lot of impact on the overall energy consumption. Sleeping periodically, and often aggressively, when not in use. In some cases, these nodes will only wake periodically to handle needed communications. This constraint is quite in contrast to the “always on” paradigm found in regular Internet hosts. Even in case of non-battery operated devices power is a constraint with respect to energy efficiency goals.
Bandwidth constraints:


Various low power radio networks offer only ~100 KBit/s or even only a few KBits/s, and show high packet loss as well as high link quality variability. Nodes may be used in usually highly unstable radio environments. The physical layer packet size may be limited (~100 bytes).
Memory constraints:


Despite the overwhelming variety of Internet-connected devices that can be found in deployments it may be useful to consider a number of classes of constrained memory availability. For example the following classes could be used:

A system designer also needs to consider CPU constraints, which often relate to energy constraints: a processor with lower performance consumes less energy. As described later in this document the design of the mainboard may allow certain components to be put to sleep to further lower energy consumption. In general, embedded systems are often purpose built with only the hardware components needed for the given task while general purpose personal computers are less constrained with regard to their mainboard layout and typically offer a huge number of optional plug-in peripherals to be connected. A factor that also has to be taken into consideration is the intended usage environment. For example, a humidity sensor deployed outside a building may need to deal with temperatures from -50 C to +85 C even. There are often physical size limitations for smart objects. While traditional mainboards are rather large, such as the Advanced Technology eXtended (ATX) design with a board size of 305 × 244 mm available in many PCs or the mini-ITX design typically found in home theater PCs with 170 × 170 mm, mainboard layouts for embedded systems are typically much smaller, such as the CoreExpress layout with 58 × 65 mm, or even smaller. In addition to the plain mainboard additional sensors, peripherals, a power adapter/battery, and a case have to be taken into consideration. Finally, there are cost restrictions as well.

The situation becomes more challenging when not only the hosts are constrained but also the network nodes themselves.

While there are constantly improvements being made, Moore's law tends to be less effective in the embedded system space than in personal computing devices: Gains made available by increases in transistor count and density are more likely to be invested in reductions of cost and power requirements than into continual increases in computing power.

3. Workshop Structure

With the ongoing work on connecting smart objects to the Internet there are many challenges the workshop participants raised in more than 70 accepted position papers. With a single workshop day discussions had to be focused and priority was given to those topics that had been raised by many authors. A summary of the identified issues are captured in the subsections below.

3.1. Architecture

A number of architectural questions were brought up in the workshop. This is natural, as the architectural choices affect the required technical solutions and the need for standards. At this workshop questions regarding the separation of traffic, the need for profiling for application specific domains, the demand for data model specific standardization as well as the design choices of the layer at which functionality should be put were discussed and are briefly summarized below.

3.1.1. One Internet vs. Islands

Devices that used to be in proprietary or application-specific networks are today migrating to IP networks. There is, however, the question of whether these smart objects are now on the same IP network as any other application as well. Controlled applications, like the fountains in front of the Bellagio hotel in Las Vegas which are operated as a distributed control system [Dolin], probably are not exchanging their control messages over the same network that is also used by hotel guests for their Internet traffic. The same had been argued for the smart grid as well. The question that was raised during the workshop is therefore in what sense are we talking about one Internet or about using IP technology for a separate, walled garden network that is independent of the Internet?

Cullen Jennings compared the current state of smart object deployment with the evolution of voice-over-IP: "Initially, many vendors recommended to run VoIP over a separate VLAN or a separate infrastructure. Nobody could imagine how to make the type of real-time guarantees, how to debug it, and how to get it to work because the Internet is not ideally suited for making any types of guarantees for real-time systems. As time went on people got better at making voice work across that type of IP network. They suddenly noticed that having voice running on a separate virtual network than their other applications was a disaster. They couldn't decide if a PC was running a softphone and whether it went on a voice or a data network. At that point people realized that they needed a converged network and all moved to one. I wouldn't be surprised to see the same happens here. Initially, we will see very separated networks. Then, those will be running over the same hardware to take advantage of the cost benefits of not having to deploy multiple sets of wires around buildings. Over time there will be strong needs to directly communicate with each other. We need to be designing the system for the long run. Assuming everything will end up on the same network even if you initially plan to run it in separate networks."

It is clearly possible to let sensors in a building communicate through the wireless access points and through the same infrastructure used for Internet access, if you want to. Those who want separation at the physical layer can do so as well. What is, however, important is to make sure that these different deployment philosophies do not force loss of interoperability.

The level of interoperability that IP accomplished fostered innovation at the application layer. Ralph Droms reinforced this message by saying: "Bright people will take a phone, build an application and connect it, with the appropriate security controls in place, to the things in my house in ways we have never thought about before. Otherwise we are just building another telephone network."

3.1.2. Domain Specific Stacks and Profiles

Imagine a home network scenario where a new light bulb is installed that should, out of the box without further configuration, interoperate with the already present light switch from a different vendor in the room. For many this is the desired level of interoperability in the area of smart object design. To accomplish this level of interoperability it is not sufficient to provide interoperability only at the network layer. Even running the same transport protocol (e.g., TCP) and application layer protocol (e.g., HTTP) is insufficient since both devices need to understand the semantics of the payloads for "Turn the light on" as well.

Standardizing the entire protocol stack for this specific "light switch/light bulb" scenario is possible. A possible stack would, for example, use IPv6 with a specific address configuration mechanism (such as stateless address autoconfiguration), a network access authentication security mechanism such as PANA, a service discovery mechanism (multicast DNS with DNS-SD), an application layer protocol, for example, Constrained Application Protocol (CoAP) (which uses UDP), and the syntax and semantic for the light on/off functionality.

As this list shows there is already some amount of protocol functionality that has to be agreed on by various stakeholders to make this scenario work seamlessly. As we approach more complex protocol interactions the functionality quickly becomes more complex: IPv4 and IPv6 on the network layer, various options at the transport layer (such as UDP, TCP, SCTP, DCCP), and there are plenty of choices at the application layer with respect to communication protocols, data formats and data models. Different requirements have lead to the development of a variety of communication protocols: client-server protocols in the style of the original HTTP, publish-subscribe protocols (like SIP or XMPP), store-and-forward messaging (borrowed from the delay tolerant networking community). Along with the different application layer communication protocols come various identity and security mechanisms.

With the smart object constraints it feels natural to develop these stacks since each application domain (e.g., health-care, smart grids, home networking) will have their unique requirements and their own community involved in the design process. How likely are these profiles going to be the right match for the future, specifically for the new innovations that will come? How many of these stacks are we going to have? Will the differences in the profiles purely be the result of different requirements coming from the individual application domains or will these mismatches reflect the spirit, understanding and preferences of the community designing them? How many stacks will multi-purpose devices have to implement?

Standardizing profiles independently for each application is not the only option. Another option is to let many different applications utilize a common foundation, i.e., a protocol stack that is implemented and utilized by every device. This, however, requires various application domains to be analyzed for their common characteristics and to identify requirements that are common across all of them. The level of difficulty for finding an agreement of how such a foundation stack should look like depends on how many layers it covers and how lightweight it has to be.

From the decisions at the workshop it was clear that the available options are not ideal and further discussions are needed.

3.1.3. Which Layer?

The end-to-end principle states that functionality should be put into the end points instead of into the networks. An additional recommendation, which is equally important, is to put functionality higher up in the protocol stack. While it is useful to make common functionality available as building blocks to higher layers the wide range of requirements by different applications lead to a model where lower layers provide only very basic functionality and more sophisticated features were made available by various applications. Still, there has been the desire to put application layer functionality into the lower layers of the networking stack. A common belief is that performance benefits can be gained if functionality is placed at the lower layers of the protocol stack. This new functionality may be offered in the form of a gateway, which bridges different communication technologies, acts on behalf of other nodes, and offers more generic functionality (such as name-based routing and caching).

Two examples of functionality offered at the network layer discussed during the workshops were location, and name-based routing:

Location:


The notion of location gives each network node the understanding of proximity (e.g., 'I am a light bulb and in the same room as the light switch.'). Today, a router may provide information as to whether other nodes belong to the same subnet or they may provide location information (for example, in the form of GPS based coordinates). However, providing a sense of the specific environment (e.g., a room, building, campus, etc.) is not possible without manual configuration. While it has been a desirable feature for many ubiquitous computing applications to know this type of information and to use it for richer application layer interactions, widespread deployment has not happened yet.

Name-based Routing:


With the work on recent clean slate architecture proposals, such as the named data networking, flexible naming concepts are being researched to allow application developer to express their application layer concepts in a way that they can be used natively by the underlying networking stack without translation. For example, Jeff Burke provided the example of his work in a theater with a distributed control system of technical equipment (such as projectors, dimmers, and light systems). Application developers name their equipment with human readable identifiers, which may change after the end of a rehearsal, and address them using these names. These variable length based naming concepts raise questions regarding scalability.

The workshop participants were not able to come to an agreement about what functionality should be moved from the application layer to the network layer.

3.2. Sleep Modes

One limitation of smart objects is the available energy. To extend battery life, for example of a watch battery or single AAA battery for months, these small, low power devices have to sleep from 99% to 99.5% of their time. For example, a light sensor may wake up to check whether it is night-time to turn on light bulbs. Most parts of the system are off-line most of the time and particularly communication components are put into a sleeping state (e.g., WLAN radio interface) and only very few components of an embedded system board, such as sensors, are triggered periodically. When interesting events happen then these components wake-up other parts of the system, for example a radio interface to connect to the Internet. Every bit is precious, so is every round trip, and every millisecond of radio activity.

Many IETF protocols implicitly assume that nodes in a network are always-on and respond to messages, i.e., to maintain a persistent presence on the network in order to respond to periodic messages that are required in order to maintain persistent sessions, connections, security associations, or state. These protocols work well on networks with sufficient network bandwidth, where there is a low cost to receiving/sending messages, and nodes are persistently available on the network.

In the early days a machine had gotten a specific IP address allocated and it could use it when it wanted to send an IP packet. You might need to execute an ARP exchange first before sending the packet but you could keep the mapping in the cache for 15 minutes.

Nowadays we want to make sure that we are on the right network before we send an IP packet, we run neighbor discovery, we cannot keep neighbor discovery for 15 minutes and so when a node wakes up again it essentially has to re-do it to refresh the cache, we want to run Detecting Network Attachment (DNA) procedures to check that hosts are on the same network either by re-getting an address using the Dynamic Host Configuration Protocol (DHCP) or by noticing that the node is using the same default gateway because of a received Router Advertisement (RA). Essentially, a number of steps have to be taken before sending a packet.

However, these protocols do not work well, if at all, when the cost of sending/receiving those messages is high (in terms of bandwidth or battery life) or in cases where nodes sleep periodically and are not persistently available to receive those messages. A number of issue arise from these almost-always-off nodes.

Also a lot of our protocols are getting more chatty. Keeping the receiver up for an additional roundtrip costs extra energy. Protocol messages can also be lengthy, e.g., many protocols carry XML-based payloads.

There are a couple of ways to think about how to make the situation less worse:

  1. The Always-On Assumption

    When designing a protocol that assumes that the nodes will always be there at least consider an alternative paradigm. For example, with duplicate address detection an alternative would be not to use it. There might be also the option to consider an architecture with a proxy node that these nodes can poll when they boot up to find out whether anyone tried to contact them or whether anything they care about has happened, or the proxy answers on behalf of another node.

  2. High Cost to Rejoin the Network

    The goal of these things is to determine whether a wireless node is not moved to another network while it was asleep and that might be a viable thing to do. Expecting a wirelss node to go through all these steps every time it wakes up before it is allowed to send an Ip packet could be considered rather excessive.

    Can we find ways to reduce the number of protocol interactions that have to be executed for network attachment, including determining whether a node has moved or the environment has changed otherwise?

  3. Verbose Protocols

    Some protocols involve multiple roundtrips and/or lengthy messages. As a result, low-powered nodes have to use more power in sending messages and have to stay powered on for a longer period of time as they wait for the protocol exchanges to complete. The best way to address these problems is to ensure that the simplest possible protocol exchanges are used when the protocols in question are designed. However, in some cases alternative encoding formats and compression may also help.

One can argue that certain features are not useful in an environment where most nodes are sleeping. The main focus of past investigations has been on IPv6 and ND but other protocols do also deserve a deeper investigation, such as DNS, and DHCP.

During the protocol design phase certain protocols were assumed to be used in a human-to-device context and therefore it was argued that the verbose encoding is helpful. Examples are the Hypertext Transfer Protocol (HTTP), the Session Initiation Protocol (SIP), and Extensible Messaging and Presence Protocol (XMPP). Nowadays these protocols are also being considered and used in device-to-device communication and the verbose nature is not helpful.

While the principles seem to be most useful for low-power, battery powered devices they would also be useful for other devices as well. Energy efficiency is useful for normal devices as well, such as laptops and smart phones.

For example, consider energy consumption in a home environment. The question is whether it will save more energy than it uses and therefore one has to consider the overall energy consumption of the entire solution. This is not always an easy question to answer. IEEE 802.11 nodes, for example, use a lot of power if they cannot be made to sleep most of the time. A light bulb may use less power but there is also the device that controls the bulb that may consume a lot of energy all the time. In total, more energy may be consumed when considering these two devices together.

3.3. Security

In the development of a smart object applications, as with any other protocol application solution, security has to be considered early in the design process. As such, the recommendations currently provided to IETF protocol architects, such as RFC 3552 [RFC3552], and RFC 4101 [RFC4101], apply also to the smart object space.

While there are additional constraints, as described in Section 2, security has to be a mandatory part of the solution. The hope is that this will lead to implementations that provide security features, deployments that utilize these, and finally that this leads to use of better security mechanisms. It is important to point out that the lack of direct user interaction will place hard requirements on deployment models, configuration mechanisms, and software upgrade/crypto agility mechanisms.

Since many of the security mechanisms allow for customization, particularly with regard to the cryptographic primitives utilized, many believe that IETF security solutions are usable without modifications in a large part of the smart object domain. Others call for new work on cryptographic primitives that make use of a single primitive (such as the Advanced Encryption Standard (AES)) as a building block for all cryptographic functions with the benefit of a smaller footprint of the overall solution. Specifically the different hardware limitations (e.g., the hardware architecture of certain embedded devices prevents pipelining to be utilized). In the excitement for new work on optimizations of cryptograhpic primitives other factors have to be taken into consideration that influence successful deployment, such as widespread support in libraries, as well as intellectual property rights (IPR). As an example of the latter aspect the struggle of Elliptic Curve Cryptography (ECC)-based cryptographic algorithms to find deployment can partially be attributed to the IPR situation. The reuse of libraries providing cryptographic functions is clearly an important way to use available memory resources in a more efficient way. To deal with the performance and footprint concerns investigations into offloading certain resource-hungry functions to parties that possess more cryptographic power have been considered. For example, the ability to delegate certificate validation to servers has been standardized in the IETF before (see Online Certificate Status Protocol (OCSP) in the Internet Key Exchange protocol version 2 (IKEv2) and in Transport Layer Security (TLS)).

Focusing only on the cryptographic primitives would be shortsighted; many would argue that this is the easy part of a smart object security solution. Key management and credential enrollment, however, are considered a big challenge by many particularly when usability requirements have to be taken into account. Another group of challenges is seen in the privacy area where the ongoing work on smart grids could be mentioned where concerns regarding the ability of others to keep track of the user's energy usage consumption (and the associated conclusions) even in an aggregated form have been voiced. As another example, it is easy to see how a scale that is connected to the Internet for uploading weight information to a social network could lead to privacy concerns. While security mechanisms used to offer protection of the communication between different parties also provide a certain degree of privacy protection they are clearly not enough to address all concerns. Even with the best communication security and access control mechanisms in place one still needs additional safeguards against the concerns mentioned in the examples.

While a lot can be said about how desirable it would be to deploy more security protocols on the entire Internet, practical considerations regarding usability and the incentives of the stakeholders involved have often lead to slower adaption.

3.4. Routing

A smart object network environment may also employ routers under similar constraints as the end devices. Currently two approaches to routing in these low power and lossy networks are under consideration, namely mesh-under and route-over. The so-called mesh-under approach places routing functions below at the link layer and consequently all devices appear as immediate neighbors at the network layer. With the route-over approach routing is done at the IP layer and none in the link layer. Each physical hop appears as a single IP hop (ignoring devices that just extend the physical range of signaling, such as repeaters). Routing in this context means running a routing protocol. IPv6 Routing Protocol for Low power and Lossy Networks (RPL) [I-D.ietf-roll-rpl], for example, belongs to the route-over category.

From an architectural point of view there are several questions that arise from where routing is provided, for example:

When multiple different link layer technologies are involved in a network design then routing at layer 3 has to be provided in any case. [I-D.routing-architecture-iot] talks about these tradeoffs between route-over and mesh-under in detail. Furthermore, those who decide about the deployment have to determine how to connect smart objects to the Internet infrastructure and a number of wired and wireless technologies may be suitable for a specific deployment. Depending on the chosen technologies the above-mentioned mesh-under vs. route-over approach will have to be decided and further decisions will have to be made about the choice of a specific routing protocol.

In 2008 the IETF formed the Routing Over Low power and Lossy networks (ROLL) working group to specify a routing solution for smart object environments. During its first year of existence, the working group studied routing requirements in details (see [RFC5867], [RFC5826], [RFC5673], [RFC5548]), worked on a protocol survey document comparing a number of existing routing protocols [I-D.ietf-roll-protocols-survey], including Ad hoc On-Demand Distance Vector (AODV)-style of protocols [RFC3561], against the identified requirements. Although the survey document was abandoned and not published an RFC it serves as a historical data point. The ROLL WG concluded that a new routing protocol satisfying the documented requirements has to be developed and the work on the RPL was started, as the IETF routing protocol for smart object networks. Nevertheless, controversial discussions at the workshop about which routing protocols is best in a given environment are still ongoing. Thomas Clausen, for example, argued for using an AODV-like routing protocol in [Clausen].

4. Conclusions and Next Steps

The workshop allowed the participants to get exposed to interesting applications and their requirements (buildings, fountains, theater, etc.), to have discussions about radically different architectures and their issues (e.g., information centric networking), to look at existing technology from a new angle (sleep nodes, energy consumption), to focus on some details of the protocol stack (neighbour discovery, security, routing) and to implementation experience.

One goal of the workshop was to identify areas that require further investigation. The list below reflects the thoughts of the workshop participants as expressed on the day of the workshop. Note that the suggested items concern potential work by the IETF and the IRTF and the order does not imply a particular preference.

Security:


A discussion of security is provided in Section 3.3. In general, security related protocol exchanges and the required amount of computational resource requirements can contribute significantly to the overall processing. Therefore, it remains a challenge how to accomplish performance improvements without sacrifying the overall security level taking the usability of the entire system into consideration.

Another challenge is how to balance the security and performance needs of smart objects with the interoperability requirements of hosts on the Internet since different suites of algorithms may tend to be chosen for these different environments. This involves trade-offs between performance on the smart objects versus end-to-end security. Solutions that mandate a "trusted" middlebox whose only role is to terminate security associations tend to be frowned upon from the security perspective, especially since end-users of challenged devices (where those exist) are unlikely to understand the security consequences of such middleboxes.

The discussion concluded with the following recommendations:

The difficulty and impact of choosing specialised algorithms for smart objects should not be underestimated. Issues that arise include the additional specification complexity (e.g., TLS already has 100's of ciphersuites defined, most of which are unused in practice), the long latency in terms of roll out (many hosts are still using deprecated algorithms 5-10 years after those algorithms were deprecated) and the barriers that IPR-encumbered schemes present to widespread deployment. While research on this topic within CFRG and the cryptographic research community is a very worthwhile goal, any such algorithms will likely have to offer very significant benefits before they will be broadly adopted. 20% less CPU is unlikely to be a winning argument no matter what an algorithm inventor believes.

Energy Design Considerations:


One part of the workshop was focused on the discussion of energy implications for IETF protocol design with proposals being made how to extend protocols to better support nodes that are mostly sleeping. Discussion are encouraged to take place at the RECIPE mailing list [RECIPE]. The workshop position paper [Wasserman] by Margaret Wasserman provides a good starting point for further investigations.
Information/Content Centric Networking:


Information/Content Centric Networking is about accessing named content and a number of research projects have emerged around this theme. At this point in time the work is not yet ready for standardization in the IETF. Instead, the formation of an IRTF research group has been proposed and more details are available on the IRTF DISCUSS mailing list [irtf-discuss].
Architectural Guidelines:


Participants suggested developing an architectural write-up about what can be done with the IETF protocols we have today and how these different elements may be combined to offer an explanation for the broader community. This would be a task for the Internet Architecture Board (IAB). An example of prior work that serves as input is [I-D.baker-ietf-core].
Network Management:


While this topic did not make it onto the workshop agenda it was considered useful to start a discussion about how to implement network management protocols, such as Network Configuration Protocol (NETCONF), on smart objects. The following position papers may be useful as a starting point for further discussions [Ersue], [Schoenwaelde]. An IETF draft is also available [I-D.hamid-6lowpan-snmp-optimizations].
Congestion Control:


When smart objects transmit sensor readings to some server on the Internet then these protocol interactions often carry a small amount of data and happen infrequently, although regularly. With the work on new application protocols, like the CoAP [I-D.ietf-core-coap], the question of congestion control mechanism should be used at the underlying transport protocol or by the application protocol itself. [I-D.eggert-core-congestion-control] is a starting point for the CoAP protocol but further work is needed.
Data Models:


Standardization of application layer protocols is important but does not ensure that, for example, a light switch and a light bulb are able to interact with each other. One area where participants saw the need for further work was in the area of data models. While prior IETF standardization work on, for example, location [GEOPRIV] can be re-used the question was raised where the IETF should focus their standardization efforts on since domain expertise in each area will be needed. As potential example energy pricing was discussed, with an example provided by [I-D.jennings-energy-pricing].
Discovery:


Additional extensions to developed discovery protocols (such as mDNS) may be needed for the smart object environment.
Home Networking:


Home network architectures should take into account the possibility of smart objects and dedicated subnetworks focusing on smart objects. A new working group, Home Networking (HOMENET) [FUN], has been proposed after the workshop to look at this topic.
Routing:


Changing radio conditions and link fluctuation may lead to the need for re-numbering. Workshop participants argued that work should be started on the multi-link subnetworks to mitigate this problem, i.e., to extend the notion of a subnet to be able to span multiple links. With the publication of RFC 4903 [RFC4903] the Internet Architecture Board had looked into this topic already and provided pros and cons.

The applicability of specific routing protocols designed for smart object networks needs further investigation. The ROLL working group is chartered with the task of constructing an applicability document for the RPL protocol, for instance.

5. Security Considerations

The workshop discussions covered a range of potential engineering activities, each with its own security considerations. As the IETF community begins to pursue specific avenues arising out of this workshop, addressing relevant security requirements will be crucial.

As described in this report part of the agenda was focused on the discussion of security, see Section 3.3.

6. Acknowledgements

We would like to thank all the participants for their position papers. The authors of the position papers were invited to the workshop.

Big thanks to Elwyn Davies for helping us to fix language bugs. We would also like to thank Andrei Robachevsky for his review comments.

Additionally, we would like to thank Ericsson and Nokia Siemens Networks for their financial support.

7. IANA Considerations

This document does not require actions by IANA.

8. References

, ", ", ", ", ", "
[RFC5582] Schulzrinne, H., "Location-to-URL Mapping Architecture and Framework", RFC 5582, September 2009.
[RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC Text on Security Considerations", BCP 72, RFC 3552, July 2003.
[RFC4101] Rescorla, E., IAB, "Writing Protocol Models", RFC 4101, June 2005.
[RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J. and H. Levkowetz, "Extensible Authentication Protocol (EAP)", RFC 3748, June 2004.
[RFC2743] Linn, J., "Generic Security Service Application Program Interface Version 2, Update 1", RFC 2743, January 2000.
[RFC2222] Myers, J.G., "Simple Authentication and Security Layer (SASL)", RFC 2222, October 1997.
[RFC4903] Thaler, D., "Multi-Link Subnet Issues", RFC 4903, June 2007.
[I-D.baker-ietf-core] Baker, F and D Meyer, "Internet Protocols for the Smart Grid", Internet-Draft draft-baker-ietf-core-15, April 2011.
[I-D.kivinen-ipsecme-ikev2-minimal] Kivinen, T, "Minimal IKEv2", Internet-Draft draft-kivinen-ipsecme-ikev2-minimal-00, February 2011.
[I-D.hamid-6lowpan-snmp-optimizations] Schoenwaelder, J, Mukhtar, H, Joo, S and K Kim, "SNMP Optimizations for Constrained Devices", Internet-Draft draft-hamid-6lowpan-snmp-optimizations-03, October 2010.
[I-D.eggert-core-congestion-control] Eggert, L, "Congestion Control for the Constrained Application Protocol (CoAP)", Internet-Draft draft-eggert-core-congestion-control-01, January 2011.
[I-D.jennings-energy-pricing] Jennings, C and B Nordman, "Communication of Energy Price Information", Internet-Draft draft-jennings-energy-pricing-01, July 2011.
[I-D.ietf-core-coap] Shelby, Z, Hartke, K, Bormann, C and B Frank, "Constrained Application Protocol (CoAP)", Internet-Draft draft-ietf-core-coap-08, October 2011.
[I-D.ietf-roll-rpl] Winter, T, Thubert, P, Brandt, A, Clausen, T, Hui, J, Kelsey, R, Levis, P, Pister, K, Struik, R and J Vasseur, "RPL: IPv6 Routing Protocol for Low power and Lossy Networks", Internet-Draft draft-ietf-roll-rpl-19, March 2011.
[RFC3561] Perkins, C., Belding-Royer, E. and S. Das, "Ad hoc On-Demand Distance Vector (AODV) Routing", RFC 3561, July 2003.
[I-D.routing-architecture-iot] Hui, J and J Vasseur, "Routing Architecture in Low-Power and Lossy Networks (LLNs)", Internet-Draft draft-routing-architecture-iot-00, March 2011.
[RFC5867] Martocci, J., De Mil, P., Riou, N. and W. Vermeylen, "Building Automation Routing Requirements in Low-Power and Lossy Networks", RFC 5867, June 2010.
[RFC5826] Brandt, A., Buron, J. and G. Porcu, "Home Automation Routing Requirements in Low-Power and Lossy Networks", RFC 5826, April 2010.
[RFC5673] Pister, K., Thubert, P., Dwars, S. and T. Phinney, "Industrial Routing Requirements in Low-Power and Lossy Networks", RFC 5673, October 2009.
[RFC5548] Dohler, M., Watteyne, T., Winter, T. and D. Barthel, "Routing Requirements for Urban Low-Power and Lossy Networks", RFC 5548, May 2009.
[I-D.ietf-roll-protocols-survey] Tavakoli, A, Dawson-Haggerty, S and P Levis, "Overview of Existing Routing Protocols for Low Power and Lossy Networks", Internet-Draft draft-ietf-roll-protocols-survey-07, April 2009.
[CFRG] McGrew (Chair), D., "IRTF Crypto Forum Research Group (CFRG)", http://irtf.org/cfrg , June 2011.
[LWIG] IETF Light-Weight Implementation Guidance (LWIG) Working Group ", http://datatracker.ietf.org/wg/lwig/charter/ , June 2011.
[enroll] IETF Credential and Provisioning Working Group Mailing List ", http://mailman.mit.edu/pipermail/ietf-enroll/ , June 2011.
[RECIPE] Reducing Energy Consumption with Internet Protocols Exploration (RECIPE) Mailing List", https://www.ietf.org/mailman/listinfo/recipe , June 2011.
[FUN] FUture home Networking (FUN) Mailing List", https://www.ietf.org/mailman/listinfo/fun , June 2011.
[GEOPRIV] IETF Geographic Location/Privacy Working Group", http://datatracker.ietf.org/wg/geopriv/ , June 2011.
[Ersue] Ersue, M. and J. Korhonen, "Ersue / Korhonen Smart Object Workshop Position Paper ", IAB Interconnecting Smart Objects with the Internet Workshop, Prague, Czech Republic, http://www.iab.org/wp-content/IAB-uploads/2011/03/Ersue.pdf, March 2011.
[Dolin] Dolin, B., "Application Communications Requirements for ‘The Internet of Things’", IAB Interconnecting Smart Objects with the Internet Workshop, Prague, Czech Republic, http://www.iab.org/wp-content/IAB-uploads/2011/03/Ersue.pdf, March 2011.
[Clausen] Clausen, T. and U. Herberg, "Some Considerations on Routing in Particular and Lossy Environments", IAB Interconnecting Smart Objects with the Internet Workshop, Prague, Czech Republic, http://www.iab.org/wp-content/IAB-uploads/2011/03/Clausen.pdf, March 2011.
[Schoenwaelde] Schoenwaelde, J., Tsou, T. and B. Sarikaya, "Protocol Profiles for Constrained Devices", IAB Interconnecting Smart Objects with the Internet Workshop, Prague, Czech Republic, http://www.iab.org/wp-content/IAB-uploads/2011/03/Schoenwaelder.pdf, March 2011.
[Wasserman] Wasserman, M., "It's Not Easy Being "Green"", IAB Interconnecting Smart Objects with the Internet Workshop, Prague, Czech Republic, http://www.iab.org/wp-content/IAB-uploads/2011/03/Wasserman.pdf, March 2011.
[irtf-discuss] Draft ICN RG Charter on IRTF DISCUSS Mailing List", http://www.ietf.org/mail-archive/web/irtf-discuss/current/msg00041.html , May 2011.

Appendix A. Program Committee

The following persons are responsible for the organization of the associated workshop and are responsible also for this event: Jari Arkko, Hannes Tschofenig, Bernard Aboba,Carsten Bormann, David Culler, Lars Eggert, JP Vasseur, Stewart Bryant, Adrian Farrel, Ralph Droms, Geoffrey Mulligan, Alexey Melnikov, Peter Saint-Andre, Marcelo Bagnulo, Zach Shelby, Isidro Ballesteros Laso, Fred Baker, Cullen Jennings, Manfred Hauswirth, and Lukas Kencl.

Appendix B. Published Workshop Material

Information about the workshop can be found at the IAB webpage: http://www.iab.org/about/workshops/smartobjects/

71 position papers were submitted to the workshop:

  1. Deployment Experience with Low Power Lossy Wireless Sensor Networks by C. Adjih, E. Baccelli, P. Jacquet, P. Minet, M. Philipp, and G. Wittenburg
  2. CoAP improvements to meet embedded device hardware constraints by Davide Barbieri
  3. Unified Device Networking Protocols for Smart Objects by Daniel Barisic and Anton Pfefferseder
  4. Simplified neighbour cache implementation in RPL/6LoWPAN by Dominique Barthel
  5. Home Control in a consumer’s perspective by Anders Brandt
  6. Authoring Place-based Experiences with an Internet of Things: Tussles of Expressive, Operational, and Participatory Goals by Jeff Burke
  7. A Dynamic Distributed Federated Approach for the Internet of Things by Diego Casado Mansilla, Juan Ramón Velasco Pérez, and Mario López-Ramos
  8. Quickly interoperable Internet of Things using simple transparent gateways by Angelo P. Castellani, Salvatore Loreto, Nicola Bui, and Michele Zorzi
  9. Position Paper of the Brno University of Technology Department of Telecommunications by Vladimir Cervenka, Lubomir Mraz, Milan Simek, Karel Pavlata
  10. Secure Access to IOT Network: An Application-based Group Key Approach by Samita Chakrabarti, and Wassim Haddad
  11. Domain-Insulated Autonomous Network Architecture (DIANA) by W. Chun
  12. Yet Another Definition on Name, Address, ID, and Locator (YANAIL) by W. Chun
  13. The Challenge of Mobility in Low Power Networks by E. Davies
  14. If the routing protocol is so smart, why should neighbour discovery be so dumb? by Nicolas Déjean
  15. Making Smart Objects IPv6 Ready: Where are we? by M. Durvy and M. Valente
  16. Position Paper on “Interconnecting Smart Objects with the Internet” by Mehmet Ersue, and Jouni Korhonen
  17. The Real-time Enterprise: IoT-enabled Business Processes by Stephan Haller, and Carsten Magerkurth
  18. Making Internet-Connected Objects readily useful by Manfred Hauswirth, Dennis Pfisterer, Stefan Decker
  19. Some Considerations on Routing in Particular and Lossy Environments by Thomas Clausen, and Ulrich Herberg
  20. A Security Protocol Adaptation Layer for the IP-based Internet of Things by René Hummen, Tobias Heer, and Klaus Wehrle
  21. Simplified SIP Approach for the Smart Object by Ryota Ishibashi, Takumi Ohba, Arata Koike, and Akira Kurokawa
  22. Mobility support for the small and smart Future Internet devices by Antonio J. Jara, and Antonio F. G. Skarmeta
  23. The Need for Efficient Reliable Multicast in Smart Grid Networks by J. Jetcheva, D. Popa, M.G. Stuber, and H. Van Wyk
  24. Lightweight Cryptography for the Internet of Things by Masanobu Katagi, and Shiho Moriai
  25. Thoughts on Reliability in the Internet of Things by James Kempf, Jari Arkko, Neda Beheshti, and Kiran Yedavalli
  26. IKEv2 and Smart Objects by Tero Kivinen
  27. Position Paper on Thing Name Service (TNS) for the Internet of Things (IoT) by Ning Kong, and Shuo Shen
  28. Connecting Smart Objects to Wireless WANs by Suresh Krishnan
  29. Towards an Information-Centric Internet with more Things by D. Kutscher, and S. Farrell
  30. Application of 6LoWPAN for the Real-Time Positioning of Manufacturing Assets by Jaacán Martinez and José L. M. Lastra
  31. Lighting interface to wireless network by Jaroslav Meduna
  32. Social-driven Internet of Connected Objects by Paulo Mendes
  33. Protocols should go forward that are required by non IP based protocols by Tsuyoshi Momose
  34. Web services for Wireless Sensor and Actuator Networks by Guido Moritz
  35. Connecting BT-LE sensors to the Internet using Ipv6 by Markus Isomäki, Kanji Kerai, Jari Mutikainen, Johanna Nieminen, Basavaraj Patil, Teemu Savolainen, and Zach Shelby
  36. Position Paper by Bruce Nordman by Bruce Nordman
  37. Issues and Challenges in Provisioning Keys to Smart Objects by Yoshihiro Ohba, and Subir Das
  38. Challenges and Solutions of Secure Smart Environments by Eila Ovaska and Antti Evesti
  39. Research Experiences about Internetworking Mechanisms to Integrate Embedded Wireless Networks into Traditional Networks by José F. Martínez, Iván Corredor, and Miguel S. Familiar
  40. Scalable Video Coding in Networked Environment by Naeem Ramzan, Tomas Piatrik, and Ebroul Izquierdo
  41. Challenges in Opportunistic Networking by Mikko Pitkänen, and Teemu Kärkkäinen
  42. Position Statement by Neeli R. Prasad, and Sateesh Addepalli
  43. A Gateway Architecture for Interconnecting Smart Objects to the Internet by Akbar Rahman, Dorothy Gellert, Dale Seed
  44. Routing Challenges and Directions for Smart Objects in Future Internet of Things by T. A. Ramrekha, E. Panaousis, and C. Politis
  45. 6LoWPAN Extension for IPsec by Shahid Raza, Thiemo Voigt, and Utz Roedig
  46. Connected Vehicle as Smart Object(s) by Ryuji Wakikawa
  47. Problem and possible approach of development and operation of Smart Objects by Shoichi Sakane
  48. Connecting Passive RFID Tags to the Internet of Things by Sandra Dominikus, and Joern-Marc Schmidt
  49. Protocol Profiles for Constrained Devices by Juergen Schoenwaelde, Tina Tsou, and Behcet Sarikaya
  50. Multihoming for Sensor Networks by J. Soininen
  51. Internet Objects for Building Control by Peter van der Stok, and Nicolas Riou
  52. Optimal information placement for Smart Objects by Shigeya Suzuki
  53. Multi-National Initiative for Cloud Computing in Health Care (MUNICH) by Christoph Thuemmler
  54. The time of the hourglass has elapsed by Laurent Toutain, Nicolas Montavont, and Dominique Barthel
  55. Sensor and Actuator Resource Architecture by Vlasios Tsiatsis, Jan Höller, and Richard Gold
  56. IT’S NOT EASY BEING “GREEN” by Margaret Wasserman
  57. Trustworthy Wireless Industrial Sensor Networks by Markus Wehner, Thomas Bartzsch, Dirk Burggraf, Sven Zeisberg, Alexis Olivereau, and Oualha Nouha
  58. Interconnecting smart objects through an overlay networking architecture by Anastasios Zafeiropoulos, Athanassios Liakopoulos and Panagiotis Gouvas
  59. Building trust among Virtual Interconnecting Smart Objects in the Future Internet by Theodore Zahariadic, Helen C. Leligou, Panagiotis Trakadas, and Mischa Dohler
  60. Experience and Challenges of Integrating Smart Devices with the Mobile Internet by Zhen Cao, and Hui Deng
  61. The “mesh-under” versus “route over” debate in IP Smart Objects Networks by JP Vasseur, and Jonathan Hui
  62. Identification and Authentication of IoT Devices by Alper Yegin
  63. Security Challenges For the Internet of Things by Tim Polk, and Sean Turner
  64. Application Communications Requirements for ‘The Internet of Things’ by Bob Dolin
  65. Interoperability Concerns in the Internet of Things by Jari Arkko
  66. Privacy in Ubiquitous Computing by Ivan Gudymenko, and Katrin Borcea-Pfitzmann
  67. The 10 Laws of Smart Object Security Design by Hannes Tschofenig, and Bernard Aboba
  68. Position Paper on “In-Network Object Cloud” Architecture and Design Goals by Alex Galis, and Stuart Clayman
  69. Temptations and Difficulties of Protocols for Smart Objects and the Internet by Alexandru Petrescu
  70. The Internet of Things – Challenge for a New Architecture from Problems by Gyu Myoung Lee, and Noel Crespi
  71. Garrulity and Fluff by Carsten Bormann, and Klaus Hartke

These papers can be retrieved from: http://www.iab.org/about/workshops/smartobjects/papers/

The slides are available for download at the following webpage: http://www.iab.org/about/workshops/smartobjects/agenda.html

Detailed meeting minutes are published here: http://www.iab.org/about/workshops/smartobjects/minutes.html

Authors' Addresses

Hannes Tschofenig Nokia Siemens Networks Linnoitustie 6 Espoo, 02600 Finland Phone: +358 (50) 4871445 EMail: Hannes.Tschofenig@gmx.net URI: http://www.tschofenig.priv.at
Jari Arkko Ericsson Jorvas, 02420 Finland EMail: jari.arkko@piuha.net