TOC 
6LoWPAN Working GroupE. Kim
Internet-DraftETRI
Intended status: InformationalD. Kaspar
Expires: May 12, 2011Simula Research Laboratory
 C. Gomez
 Tech. Univ. of Catalonia/i2CAT
 C. Bormann
 Universität Bremen TZI
 November 8, 2010


Problem Statement and Requirements for 6LoWPAN Routing
draft-ietf-6lowpan-routing-requirements-08

Abstract

6LoWPANs are formed by devices that are compatible with the IEEE 802.15.4 standard. However, neither the IEEE 802.15.4 standard nor the 6LoWPAN format specification define how mesh topologies could be obtained and maintained. Thus, it should be considered how 6LoWPAN formation and multi-hop routing could be supported.
This document provides the problem statement and design space for 6LoWPAN routing. It defines the routing requirements for 6LoWPAN networks, considering the low-power and other particular characteristics of the devices and links. The purpose of this document is not to recommend specific solutions, but to provide general, layer-agnostic guidelines about the design of 6LoWPAN routing, which can lead to further analysis and protocol design. This document is intended as input to groups working on routing protocols relevant to 6LoWPAN, such as the IETF ROLL WG.

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 May 12, 2011.

Copyright Notice

Copyright (c) 2010 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.  Problem Statement
2.  Terminology
3.  Design Space
    3.1.  Reference Network Model
4.  Scenario Considerations and Parameters for 6LoWPAN Routing
5.  6LoWPAN Routing Requirements
    5.1.  Support of 6LoWPAN Device Properties
    5.2.  Support of 6LoWPAN Link Properties
    5.3.  Support of 6LoWPAN Network Characteristics
    5.4.  Support of Security
    5.5.  Support of Mesh Under Forwarding
6.  Security Considerations
7.  IANA Considerations
8.  Acknowledgements
9.  References
    9.1.  Normative References
    9.2.  Informative References
§  Authors' Addresses




 TOC 

1.  Problem Statement

6LoWPANs are formed by devices that are compatible with the IEEE 802.15.4 standard [IEEE802.15.4] (IEEE Computer Society, “IEEE Std. 802.15.4-2006 (as amended),” 2007.). Most of the LoWPAN devices are distinguished by their low bandwidth, short range, scarce memory capacity, limited processing capability and other attributes of inexpensive hardware. The characteristics of nodes participating in LoWPANs are assumed to be those described in the 6LoWPAN problem statement [RFC4919] (Kushalnagar, N., Montenegro, G., and C. Schumacher, “IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals,” August 2007.), and the IPv6 over IEEE 802.15.4 [RFC4944] (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.) document which has specified how to carry IPv6 packets over IEEE 802.15.4 and similar networks. Whereas IEEE 802.15.4 distinguishes two types of devices called full-function devices (FFD) and reduced-function devices (RFDs), this distinction is based on some MAC layer features that are not always in use. Hence, the distinction is not made in this document. Nevertheless, some 6LoWPAN nodes may limit themselves to the role of hosts only, whereas other 6LoWPAN nodes may take part in routing. This host/router distinction can correlate with the processing and storage capabilities of the device and power available in a similar way to the idea of RFDs and FFDs.

IEEE 802.15.4 networks support star and mesh topologies. However, neither the IEEE 802.15.4 standard nor the 6LoWPAN format specification ([RFC4944] (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.)) define how mesh topologies could be obtained and maintained. Thus, 6LoWPAN formation and multi-hop routing can be supported either below the IP layer (the adaptation layer or LLC) or the IP layer. (Note that in the IETF, the term "routing" usually, but not always [RFC5556] (Touch, J. and R. Perlman, “Transparent Interconnection of Lots of Links (TRILL): Problem and Applicability Statement,” May 2009.), refers exclusively to the formation of paths and the forwarding at the IP layer. In this document we distinguish the layer at which these services are performed by the terms "Route Over" and "Mesh Under". See Section 2 (Terminology) and Section 3 (Design Space).) A number of IP routing protocols have been developed in various IETF working groups. However, these existing routing protocols may not satisfy the requirements of multi-hop routing in 6LoWPANs, for the following reasons:

These properties create new challenges on obtaining robust and reliable routing within LoWPANs.

The 6LoWPAN problem statement document ("6LoWPAN Problems and Goals" [RFC4919] (Kushalnagar, N., Montenegro, G., and C. Schumacher, “IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals,” August 2007.)) briefly mentions four requirements on routing protocols:

(a) low overhead on data packets

(b) low routing overhead

(c) minimal memory and computation requirements

(d) support for sleeping nodes considering battery saving

These four high-level requirements describe the basic requirements for 6LoWPAN routing. Based on the fundamental features of 6LoWPAN, more detailed routing requirements are presented in this document, which can lead to further analysis and protocol design.

Considering the problems above, detailed 6LoWPAN routing requirements must be defined. Application-specific features affect the design of 6LoWPAN routing requirements and the corresponding solutions. However, various applications can be profiled by similar technical characteristics, although the related detailed requirements might differ (e.g., a few dozens of nodes in a home lighting system need appropriate scalability for its applications, while millions of nodes for a highway infrastructure system also need appropriate scalability).

This routing requirements document states the routing requirements of 6LoWPAN applications in general, providing examples for different cases of routing. It does not imply a single routing solution to be favorable for all 6LoWPAN applications and there is no requirement of different routing protocols to run simultaneously.



 TOC 

2.  Terminology

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in [RFC2119] (Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels,” March 1997.).

Readers are expected to be familiar with all the terms and concepts that are discussed in "IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals" (Kushalnagar, N., Montenegro, G., and C. Schumacher, “IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals,” August 2007.) [RFC4919], and "Transmission of IPv6 Packets over IEEE 802.15.4 Networks" (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.) [RFC4944].

This specification makes use of the terminology defined in the "Neighbor Discovery for 6LoWPAN" [I‑D.ietf‑6lowpan‑nd] (Shelby, Z., Chakrabarti, S., and E. Nordmark, “Neighbor Discovery Optimization for Low-power and Lossy Networks,” October 2010.) . In addition, this specification defines a Mesh node as a device which is capable of routing data below the IP layer.



 TOC 

3.  Design Space

Apart from a wide variety of conceivable routing algorithms for 6LoWPAN, it is possible to perform routing in the IP layer, using a Route Over approach or below IP, as defined by the 6LoWPAN format document [RFC4944] (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.), using the Mesh Under approach (see Figure 1 (Mesh Under (left) and Route Over routing (right))).

The Route Over approach relies on IP routing and therefore supports routing over possibly various types of interconnected links.
Note: The ROLL WG is now working on Route Over approaches for Low power and Lossy Networks (LLNs), not specifically for 6LoWPAN. This document focuses on 6LoWPAN-specific requirements; it may be used in conjunction with the more application-oriented requirements defined by the ROLL WG.

The Mesh Under approach performs the multi-hop communication below the IP link. The most significant consequence of Mesh Under mechanism is that the characteristics of IEEE 802.15.4 directly affect the 6LoWPAN routing mechanisms, including the use of 64-bit (or 16-bit short) link layer addresses instead of IP addresses. A 6LoWPAN would therefore be seen as a single IP link.

Most statements in this document consider both the Route Over and Mesh Under cases.



Figure 1 (Mesh Under (left) and Route Over routing (right)) shows the place of 6LoWPAN routing in the entire network stack.

 +---------------------------+  +-----------------------------+
 |      Application Layer    |  |      Application Layer      |
 +---------------------------+  +-----------------------------+
 | Transport Layer (TCP/UDP) |  |  Transport Layer (TCP/UDP)  |
 +---------------------------+  +-----------------------------+
 |     Network Layer (IPv6)  |  |  Network       +---------+  |
 +---------------------------+  |  Layer         | Routing |  |
 |  6LoWPAN                  |  |  (IPv6)        +---------+  |
 |  Adaptation               |  +-----------------------------+
 |  Layer       +----------+ |  |  6LoWPAN Adaptation Layer   |
 +--------------| Routing* |-+  +-----------------------------+
 | 802.15.4 MAC +----------+ |  |        802.15.4 MAC         |
 +---------------------------+  +-----------------------------+
 |         802.15.4 PHY      |  |        802.15.4 PHY         |
 +---------------------------+  +-----------------------------+
  * Here, 'Routing' is not equivalent to IP routing,
    but includes the functionalities of path computation and
    forwarding under the IP layer.
    The term 'Routing' is used in the figure in order to
    illustrate which layer handles path computation and
    packet forwarding in Mesh Under compared to Route Over.

 Figure 1: Mesh Under (left) and Route Over routing (right) 

In order to avoid packet fragmentation and the overhead for reassembly, routing packets should fit into a single IEEE 802.15.4 physical frame and application data should not be expanded to an extent that they no longer fit.



 TOC 

3.1.  Reference Network Model

For multi-hop communication in 6LoWPAN, when a Route Over mechanism is in use, all routers (i.e. 6LoWPAN Border Routers (6LBRs) and 6LoWPAN Routers (6LRs)) perform IP routing within the stub network (see Figure 2 (An example of a Route Over LoWPAN)). In this case, the link-local scope covers the set of nodes within symmetric radio range of a node.

When a LoWPAN follows the Mesh Under configuration, the 6LBR is the only IPv6 router in the LoWPAN (see Figure 3 (An example of a Mesh Under LoWPAN)). This means that the IPv6 link-local scope includes all nodes in the LoWPAN. For this, a Mesh Under mechanism MUST be provided to support multi-hop transmission.



     h   h
    /    |                     6LBR: 6LoWPAN Border Router
6LBR -- 6LR --- 6LR --- h       6LR: 6LoWPAN Router
        / \                       h: Host
       h  6LR --- h
           |
          / \
       6LR - 6LR -- h
 Figure 2: An example of a Route Over LoWPAN 



     h   h
    /    |                    6LBR: 6LoWPAN Border Router
6LBR --- m --- m --- h           m: Mesh Node
        / \                      h: Host
       h   m --- h
           |
          / \
         m - m -- h
 Figure 3: An example of a Mesh Under LoWPAN 

Note than in both Mesh Under and Route Over networks, there is no expectation of topologically based address assignment in the 6LoWPAN. Instead, addresses are typically assigned based on the EUI-64 addresses assigned at manufacturing time to nodes, or based on a (from a topological point of view) more or less random process assigning 16-bit MAC addresses to individual nodes. Within a 6LoWPAN, there is therefore no opportunity for aggregation or summarization of IPv6 addresses beyond the sharing of (one or more) common prefixes.

Not all devices that are in radio range of each other need to be part of the same LoWPAN. When multiple LoWPANs are formed with globally unique IPv6 addresses in the 6LoWPANs, and device (a) of LoWPAN [A] wants to communicate with device (b) of LoWPAN [B], the normal IPv6 mechanisms will be employed. For Route Over, the IPv6 address of (b) is set as the destination of the packets, and the devices perform IP routing to the 6LBR for these outgoing packets. For Mesh Under, there is one IP hop from a device (a) to the 6LBR of [A], no matter how many radio hops they are apart from each other. This, of course, assumes the existence of a Mesh Under routing protocol in order to reach the 6LBR. Note that a default route to the 6LBR could be inserted into the 6LoWPAN routing system for both Route Over and Mesh Under.



 TOC 

4.  Scenario Considerations and Parameters for 6LoWPAN Routing

IP-based LoWPAN technology is still in its early stage of development, but the range of conceivable usage scenarios is tremendous. The numerous possible applications of sensor networks make it obvious that mesh topologies will be prevalent in LoWPAN environments and robust routing will be a necessity for expedient communication. Research efforts in the area of sensor networking have put forth a large variety of multi-hop routing algorithms [refs.bulusu] (Bulusu, N. and S. Jha, “Wireless Sensor Networks,” July 2005.). Most related work focuses on optimizing routing for specific application scenarios, which can be realized using several models of communication, including the following ones [refs.cctc] (Lu, J., Valois, F., Dohler, M., and D. Barthel, “Quantifying Organization by Means of Entropy,” 2008.):

Depending on the topology of a LoWPAN and the application(s) running over it, different types of routing may be used. However, this document abstracts from application-specific communication and describes general routing requirements valid for overall routing in LoWPANs.

The following parameters can be used to describe specific scenarios in which the candidate routing protocols could be evaluated.

a.
Network Properties:
  • Number of Devices, Density and Network Diameter:
    These parameters usually affect the routing state directly (e.g. the number of entries in a routing table or neighbor list). Especially in large and dense networks, policies must be applied for discarding "low-quality" and stale routing entries in order to prevent memory overflow.
  • Connectivity:
    Due to external factors or programmed disconnections, a LoWPAN can be in several states of connectivity; anything in the range from "always connected" to "rarely connected". This poses great challenges to the dynamic discovery of routes across a LoWPAN.
  • Dynamicity (including mobility):
    Location changes can be induced by unpredictable external factors or by controlled motion, which may in turn cause route changes. Also, nodes may dynamically be introduced into a LoWPAN and removed from it later. The routing state and the volume of control messages may heavily depend on the number of moving nodes in a LoWPAN and their speed, as well as how quickly and frequently environmental characteristics influencing radio propagation change.
  • Deployment:
    In a LoWPAN, it is possible for nodes to be scattered randomly or to be deployed in an organized manner. The deployment can occur at once, or as an iterative process, which may also affect the routing state.
  • Spatial Distribution of Nodes and Gateways:
    Network connectivity depends on the spatial distribution of the nodes and on other factors, such as device number, density and transmission range. For instance, nodes can be placed on a grid, or randomly located in an area (as can be modeled by a bidimensional Poisson distribution), etc. Assuming a random spatial distribution, an average of 7 neighbors per node are required for approximately 95% network connectivity (10 neighbors per node are needed for 99% connectivity)[refs.Kuhn] (Kuhn, F., Wattenhofer, R., and A. Zollinger, “Worst-Case Optimal and Average-Case Efficient Ad-Hoc Geometric Routing,” 2003.). In addition, if the LoWPAN is connected to other networks through infrastructure nodes called gateways, the number and spatial distribution of gateways affects network congestion and available data rate, among others.
  • Traffic Patterns, Topology and Applications:
    The design of a LoWPAN and the requirements on its application have a big impact on the network topology and the most efficient routing type to be used. For different traffic patterns (point-to-point, multipoint-to-point, point-to-multipoint) and network architectures, various routing mechanisms have been developed, such as data-centric, event-driven, address-centric, and geographic routing.
  • Classes of Service:
    For mixing applications of different criticality on one LoWPAN, support of multiple classes of service may be required in resource-constrained LoWPANs and may require a new routing protocol functionality.
  • Security:
    LoWPANs may carry sensitive information and require a high level of security support where the availability, integrity, and confidentiality of data are of prime relevance. Secured messages cause overhead and affect the power consumption of LoWPAN routing protocols.
b.
Node Parameters:
  • Processing Speed and Memory Size:
    These basic parameters define the maximum size of the routing state and the maximum complexity of its processing. LoWPAN nodes may have different performance characteristics, queuing strategies and queue buffer sizes.
  • Power Consumption and Power Source:
    The number of battery- and mains-powered nodes and their positions in the topology created by them in a LoWPAN affect routing protocols in their selection of paths that optimize network lifetime.
  • Transmission Range:
    This parameter affects routing. For example, a high transmission range may cause a dense network, which in turn results in more direct neighbors of a node, higher connectivity and a larger routing state.
  • Traffic Pattern:
    This parameter affects routing since highly loaded nodes (either because they are the source of packets to be transmitted or due to forwarding) may contribute to higher delivery delays and may consume more energy than lightly loaded nodes. This applies to both data packets and routing control messages.
c.
Link Parameters: This section discusses link parameters that apply to IEEE 802.15.4 legacy mode (i.e. not making use of improved modulation schemes).
  • Throughput:

    The maximum user data throughput of a bulk data transmission between a single sender and a single receiver through an unslotted IEEE 802.15.4 2.4 GHz channel in ideal conditions is as follows [refs.Latre] (Latre, M., De Mil, P., Moerman, I., Dhoedt, B., and P. Demeester, “Throughput and Delay Analysis of Unslotted IEEE 802.15.4,” May 2006.):
    • 16-bit MAC addresses, unreliable mode: 151.6 kbit/s
    • 16-bit MAC addresses, reliable mode: 139.0 kbit/s
    • 64-bit MAC addresses, unreliable mode: 135.6 kbit/s
    • 64-bit MAC addresses, reliable mode: 124.4 kbit/s


    In the case of 915 MHz band:
    • 16-bit MAC addresses, unreliable mode: 31.1 kbit/s
    • 16-bit MAC addresses, reliable mode: 28.6 kbit/s
    • 64-bit MAC addresses, unreliable mode: 27.8 kbit/s
    • 64-bit MAC addresses, reliable mode: 25.6 kbit/s


    In the case of 868 MHz band:
    • 16-bit MAC addresses, unreliable mode: 15.5 kbit/s
    • 16-bit MAC addresses, reliable mode: 14.3 kbit/s
    • 64-bit MAC addresses, unreliable mode: 13.9 kbit/s
    • 64-bit MAC addresses, reliable mode: 12.8 kbit/s
  • Latency:
    The range of latencies, depending on payload size, of a frame transmission between a single sender and a single receiver through an unslotted IEEE 802.15.4 2.4 GHz channel in ideal conditions are as shown next [refs.Latre] (Latre, M., De Mil, P., Moerman, I., Dhoedt, B., and P. Demeester, “Throughput and Delay Analysis of Unslotted IEEE 802.15.4,” May 2006.). For unreliable mode, the actual latency is provided. For reliable mode, the round-trip-time including transmission of a layer two acknowledgment is provided:
    • 16-bit MAC addresses, unreliable mode: [1.92 ms, 6.02 ms]
    • 16-bit MAC addresses, reliable mode: [2.46 ms, 6.56 ms]
    • 64-bit MAC addresses, unreliable mode: [2.75 ms, 6.02 ms]
    • 64-bit MAC addresses, reliable mode: [3.30 ms, 6.56 ms]

    For the 915 MHz band:
    • 16-bit MAC addresses, unreliable mode: [5.85 ms, 29.35 ms]
    • 16-bit MAC addresses, reliable mode: [8.35 ms, 31.85 ms]
    • 64-bit MAC addresses, unreliable mode: [8.95 ms, 29.35 ms]
    • 64-bit MAC addresses, reliable mode: [11.45 ms, 31.85 ms]

    For the 868 MHz band:
    • 16-bit MAC addresses, unreliable mode: [11.7 ms, 58.7 ms]
    • 16-bit MAC addresses, reliable mode: [16.7 ms, 63.7 ms]
    • 64-bit MAC addresses, unreliable mode: [17.9 ms, 58.7 ms]
    • 64-bit MAC addresses, reliable mode: [22.9 ms, 63.7 ms]

Note that some of the parameters presented in this section may be used as link or node evaluation metrics. However, multi-criteria routing may be too expensive for 6LoWPAN nodes. Rather, various single-criteria metrics are available and can be selected to suit the environment or application.



 TOC 

5.  6LoWPAN Routing Requirements

This section defines a list of requirements for 6LoWPAN routing. An important design property specific to low-power networks is that LoWPANs have to support multiple device types and roles, such as:

Due to these different device types and roles LoWPANs need to consider the following two primary attributes:

These fundamental attributes of LoWPANs affect the design of routing solutions. Whether existing routing specifications are simplified and modified, or new solutions are introduced in order to fit the low-power requirements of LoWPANs, they need to meet the requirements described in the following.



 TOC 

5.1.  Support of 6LoWPAN Device Properties

The general objectives listed in this section should be met by 6LoWPAN routing protocols. The importance of each requirement is dependent on what node type the protocol is running on and what the role of the node is. The following requirements consider the presence of battery-powered nodes in LoWPANs.

[R01] 6LoWPAN routing protocols SHOULD allow implementation with small code size and require low routing state to fit the typical 6LoWPAN node capacity. Generally speaking, the code size is bounded by available flash memory size, and the routing table is bounded by RAM size, possibly limiting it to less than 32 entries.

The RAM size of LoWPAN nodes often ranges between 4 kB (2 kB minimum) and 10 kB, and program flash memory normally consists of 48 kB to 128 kB. (e.g., in the current market, MICAz has 128 kB program flash, 4 kB EEPROM, 512 kB external flash ROM; TIP700CM has 48 kB program flash, 10 kB RAM, 1 MB external flash ROM).

Due to these hardware restrictions, code SHOULD fit within a small memory size; no more than 48 kB to 128 kB of flash memory including at least a few tens of KB of application code size. (As a general observation, a routing protocol of low complexity may help achieving the goal of reducing power consumption, improves robustness, requires lower routing state, is easier to analyze, and may be less prone to security attacks.)

In addition, operation with limited amounts of routing state (such as routing tables and neighbor lists) SHOULD be maintained since some typical memory sizes preclude storing state of a large number of nodes. For instance, industrial monitoring applications may need to support at maximum 20 hops [RFC5673] (Pister, K., Thubert, P., Dwars, S., and T. Phinney, “Industrial Routing Requirements in Low-Power and Lossy Networks,” October 2009.). Small networks can be designed to support a smaller number of hops. While the need for this is highly dependent on the network architecture, there should be at least one mode of operation that can function with 32 forwarding entries or less.

[R02] 6LoWPAN routing protocols SHOULD cause minimal power consumption by the efficient use of control packets (e.g., minimize expensive IP multicast which causes link broadcast to the entire LoWPAN) and by the efficient routing of data packets.

One way of battery lifetime optimization is by achieving a minimal control message overhead. Compared to functions such as computational operations or taking sensor samples, radio communications is by far the dominant factor of power consumption [refs.SmartDust] (Pister, K. and B. Boser, “Smart Dust: Wireless Networks of Millimeter-Scale Sensor Nodes,” .). Power consumption of transmission and/or reception depends linearly on the length of data units and on the frequency of transmission and reception of the data units [refs.Shih] (Shih, E., “Physical Layer Driven Protocols and Algorithm Design for Energy-Efficient Wireless Sensor Networks,” July 2001.).

The energy consumption of two example RF controllers for low-power nodes is shown in [refs.Hill] (Hill, J., “System Architecture for Wireless Sensor Networks,” .). The TR1000 radio consumes 21 mW when transmitting at 0.75 mW, and 15 mW on reception (with a receiver sensitivity of -85 dBm). The CC1000 consumes 31.6 mW when transmitting 0.75 mW, and 20 mW for receiving (with a receiver sensitivity of -105 dBm). The power endurance under the concept of an idealized power source is explained in [refs.Hill] (Hill, J., “System Architecture for Wireless Sensor Networks,” .). Based on the energy of an idealized AA battery, the CC1000 can transmit for approximately 4 days straight or receive for 9 consecutive days. Note that availability for reception consumes power as well.

As multicast may cause flooding in the LoWPAN, a 6LoWPAN routing protocol SHOULD minimize the control cost by multicasting routing packets.

Control cost of routing protocols in low power and lossy networks is discussed in more detail in [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,” April 2009.).



 TOC 

5.2.  Support of 6LoWPAN Link Properties

6LoWPAN links have the characteristics of low data rate and possibly high loss rates. The routing requirements described in this section are derived from the link properties.

[R03] 6LoWPAN routing protocol control messages SHOULD NOT exceed a single IEEE 802.15.4 frame size in order to avoid packet fragmentation and the overhead for reassembly.

In order to save energy, routing overhead should be minimized to prevent fragmentation of frames. Therefore, 6LoWPAN routing should not cause packets to exceed the IEEE 802.15.4 frame size. This reduces the energy required for transmission, avoids unnecessary waste of bandwidth, and prevents the need for packet reassembly. As calculated in RFC4944 [RFC4944] (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.), the maximum size of a 6LoWPAN frame, in order not to cause fragmentation, is 81 octets. This may imply the use of semantic fragmentation and/or algorithms that can work on small increments of routing information.

[R04] The design of routing protocols for LoWPANs must consider the fact that packets are to be delivered with sufficient probability according to application requirements.

Requirements on successful end-to-end packet delivery ratio (where delivery may be bounded within certain latency) vary depending on applications. In industrial applications, some non-critical monitoring applications may tolerate successful delivery ratio of less than 90% with hours of latency; in some other cases, a delivery ratio of 99.9% is required [RFC5673] (Pister, K., Thubert, P., Dwars, S., and T. Phinney, “Industrial Routing Requirements in Low-Power and Lossy Networks,” October 2009.). In building automation applications, application layer errors must be below 0.01% [RFC5867] (Martocci, J., De Mil, P., Riou, N., and W. Vermeylen, “Building Automation Routing Requirements in Low-Power and Lossy Networks,” June 2010.).

Successful end-to-end delivery of packets in an IEEE 802.15.4 mesh depends on the quality of the path selected by the routing protocol and on the ability of the routing protocol to cope with short-term and long-term quality variation. The metric of the routing protocol strongly influences performance of the routing protocol in terms of delivery ratio.

The quality of a given path depends on the individual qualities of the links (including the devices) that compose that path. IEEE 802.15.4 settings affect the quality perceived at upper layers. In particular, in IEEE 802.15.4 reliable mode, if an acknowledgment frame is not received after a given period, the originator retries frame transmission up to a maximum number of times. If an acknowledgment frame is still not received by the sender after performing the maximum number of transmission attempts, the MAC layer assumes the transmission has failed and notifies the next higher layer of the failure. Note that excessive retransmission may be detrimental, see RFC 3819 [RFC3819] (Karn, P., Bormann, C., Fairhurst, G., Grossman, D., Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. Wood, “Advice for Internet Subnetwork Designers,” July 2004.).

[R05] The design of routing protocols for LoWPANs must consider the latency requirements of applications and IEEE 802.15.4 link latency characteristics.

Latency requirements may differ from a few hundreds milliseconds to minutes, depending on the type of application. Real-time building automation applications usually need response times below 500 ms between egress and ingress, while forced entry security alerts must be routed to one or more fixed or mobile user devices within 5 s [RFC5867] (Martocci, J., De Mil, P., Riou, N., and W. Vermeylen, “Building Automation Routing Requirements in Low-Power and Lossy Networks,” June 2010.). Non-critical closed loop applications for industrial automation have latency requirements that can be as low as 100 ms but many control loops are tolerant of latencies above 1 s [RFC5673] (Pister, K., Thubert, P., Dwars, S., and T. Phinney, “Industrial Routing Requirements in Low-Power and Lossy Networks,” October 2009.). In contrast to this, urban monitoring applications allow latencies smaller than the typical intervals used for reporting sensed information; for instance, in the order of seconds to minutes [RFC5548] (Dohler, M., Watteyne, T., Winter, T., and D. Barthel, “Routing Requirements for Urban Low-Power and Lossy Networks,” May 2009.).

The range of latencies of a frame transmission between a single sender and a single receiver through an ideal unslotted IEEE 802.15.4 2.4 GHz channel is between 2.46 ms and 6.02 ms in 64 bit MAC address unreliable mode and 2.20 ms to 6.56 ms in 64 bit address reliable mode. The range of latencies of 868 MHz band is from 11.7 ms to 63.7 ms, depending on the address type and reliable/unreliable mode used. Note that the latencies may be larger than that depending on channel load, MAC layer settings procedure, and reliable/unreliable mode choice. Note that other MAC approaches than the legacy 802.15.4 may be used (e.g. TDMA). Duty cycling may further affect latency (see [R08]). Depending on the routing path chosen and the network diameter, multiple of these hops may contribute to the end-to-end latency that application experience.

Note that a tradeoff exists between [R05] and [R04].

[R06] 6LoWPAN routing protocols SHOULD be robust to dynamic loss caused by link failure or device unavailability either in the short term (ca. 30 ms), due to RSSI variation, interference variation, noise and asynchrony, or in the long term, due to a depleted power source, hardware breakdown, operating system misbehavior, etc.

An important trait of 6LoWPAN devices is their unreliability due to limited system capabilities, and also because they might be closely coupled to the physical world with all its unpredictable variation. In harsh environments, LoWPANs easily suffer from link failure. Collision or link failure easily increases send and receive queues and can lead to queue overflow and packet losses.

For home applications, where users expect feedback after carrying out actions (such as handling a remote control while moving around), routing protocols must converge within 2 seconds if the destination node of the packet has moved and must converge within 0.5 seconds if only the sender has moved [RFC5826] (Brandt, A., Buron, J., and G. Porcu, “Home Automation Routing Requirements in Low-Power and Lossy Networks,” April 2010.). The tolerance of the recovery time can vary depending on the application, however, the routing protocol must provide the detection of short-term unavailability and long-term disappearance. The routing protocol has to exploit network resources (e.g. path redundancy) to offer good network behavior despite of node failure.

Different routing protocols may exhibit different scaling characteristics with respect to the recovery/convergence time and the computational resources to achieve recovery after a convergence, hence see also R01/R10.

[R07] 6LoWPAN routing protocols SHOULD be designed to correctly operate in the presence of link asymmetry.

Link asymmetry occurs when the probability of successful transmission between two nodes is significantly higher in one direction than in the other one. This phenomenon has been reported in a large number of experimental studies and it is expected that 6LoWPANs will exhibit link asymmetry.



 TOC 

5.3.  Support of 6LoWPAN Network Characteristics

6LoWPANs can be deployed in different sizes and topologies, adhere to various models of mobility, be exposed to various levels of interference, etc. In any case, LoWPANs must maintain low energy consumption. The requirements described in the following subsection are derived from the network attributes of 6LoWPANs.

[R08] The design of 6LoWPAN routing protocols SHOULD take into account that some nodes may be unresponsive during certain time intervals due to periodic hibernation.

Many nodes in LoWPAN environments might periodically hibernate (i.e. disable their transceiver activity) in order to save energy. Therefore, routing protocols must ensure robust packet delivery despite nodes frequently shutting off their radio transmission interface. Feedback from the lower IEEE 802.15.4 layer may be considered to enhance the power-awareness of 6LoWPAN routing protocols.

CC1000-based nodes must operate at a duty cycle of approximately 2% to survive for one year from idealized AA battery power source [refs.Hill] (Hill, J., “System Architecture for Wireless Sensor Networks,” .). For home automation purposes, it is suggested that the devices have to maximize the sleep phase with a duty cycle lower than 1% [RFC5826] (Brandt, A., Buron, J., and G. Porcu, “Home Automation Routing Requirements in Low-Power and Lossy Networks,” April 2010.), while in building automation applications, batteries must be operational for at least 5 years when the sensing devices are transmitting data (e.g. 64 bytes) once per minute [RFC5867] (Martocci, J., De Mil, P., Riou, N., and W. Vermeylen, “Building Automation Routing Requirements in Low-Power and Lossy Networks,” June 2010.).

Dependent on the application in use, packet rates may range from one per second to one per day or beyond. Routing protocols may take advantage of knowledge about the packet transmission rate and utilize this information in calculating routing paths. In many IEEE 802.15.4 deployments, and in other wireless low-power technologies, forwarders are mains-powered devices (and hence do not need to sleep). However, it cannot be assumed that all forwarders are mains-powered. A routing protocol that addresses this case SHOULD provide a mode in which power consumption is a metric. In addition, using nodes in power-saving modes for forwarding may increase delay and reduce packet delivery probability, which in this case also should be available as an input into the path computation.

[R09] The metric used by 6LoWPAN routing protocols SHOULD provide some flexibility with respect to the inputs provided by the lower layers and other measures to optimize path selection considering energy balance and link qualities.

In homes, buildings, or infrastructure, some nodes will be installed with mains power. Such power-installed nodes MUST be considered as relay points for a prominent role in packet delivery. 6LoWPAN routing protocols MUST know the power constraints of the nodes.

Simple hop-count-only mechanisms may be inefficient in 6LoWPANs. There is a Link Quality Indication (LQI), or/and RSSI from IEEE 802.15.4 that may be taken into account for better metrics. The metric to be used (and its goal) may depend on applications and requirements.

The numbers in Figure 4 (An example network) represent the Link Delivery Ratio (LDR) of each pair of nodes. There are studies that show a piecewise linear dependence between LQI and LDR [refs.Chen] (Chen, B., Muniswamy-Reddy, K., and M. Welsh, “Ad-Hoc Multicast Routing on Resource-Limited Sensor Nodes,” 2006.).



                                  0.6
                               A-------C
                                \     /
                             0.9 \   / 0.9
                                  \ /
                                   B
 Figure 4: An example network 

In this simple example, there are two options in routing from node A to node C, with the following features:

A.
Path AC:
  • (1/0.6) = 1.67 avg. transmissions needed for each packet (confirmed link layer delivery with retransmissions and negligible ACK loss have been assumed)
  • one-hop path
  • good in energy consumption and end-to-end latency of data packets, bad in delivery ratio (0.6)
  • bad in probability of route reconfigurations
B.
Path ABC:
  • (1/0.9)+(1/0.9) = 2.22 avg. transmissions needed for each packet (under the same assumptions as above)
  • two-hop path
  • bad in energy consumption and end-to-end latency of data packets, good in delivery ratio (0.81)

If energy consumption of the network must be minimized, path AC is the best (this path would be chosen based on a hop count metric). However, if the delivery ratio in that case is not sufficient, the best path is ABC (it would be chosen by an LQI based metric). Combinations of both metrics can be used.

The metric also affects the probability of route reconfiguration. Route reconfiguration, which may be triggered by packet losses, may require transmission of routing protocol messages. It is possible to use a metric aimed at selecting the path with low route reconfiguration rate by using LQI as an input to the metric. Such a path has good properties, including stability and low control message overhead.

Note that a tradeoff exists between [R09] and [R01].

[R10] 6LoWPAN routing protocols SHOULD be designed to achieve both scalability from a few nodes to maybe millions of nodes and minimality in terms of used system resources.

A LoWPAN may consist of just a couple of nodes (for instance in a body-area network), but may also contain much higher numbers of devices (e.g. monitoring of a city infrastructure or a highway). For home automation applications it is envisioned that the routing protocol must support 250 devices in the network [RFC5826] (Brandt, A., Buron, J., and G. Porcu, “Home Automation Routing Requirements in Low-Power and Lossy Networks,” April 2010.), while routing protocols for metropolitan-scale sensor networks must be capable of clustering a large number of sensing nodes into regions containing on the order of 10^2 to 10^4 sensing nodes each [RFC5548] (Dohler, M., Watteyne, T., Winter, T., and D. Barthel, “Routing Requirements for Urban Low-Power and Lossy Networks,” May 2009.). It is therefore necessary that routing mechanisms are designed to be scalable for operation in various network sizes. However, due to a lack of memory size and computational power, 6LoWPAN routing might limit forwarding entries to a small number, such as at maximum 32 routing table entries. Specially in large networks, the routing mechanism MUST be designed in such a way that the number of routers be smaller than the number of hosts.

[R11] The procedure of route repair and related control messages SHOULD NOT harm overall energy consumption from the routing protocols.

Local repair improves throughput and end-to-end latency, especially in large networks. Since routes are repaired quickly, fewer data packets are dropped, and a smaller number of routing protocol packet transmissions are needed since routes can be repaired without source initiated Route Discovery [refs.Lee] (Lee, S., Belding-Royer, E., and C. Perkins, “Scalability Study of the Ad Hoc On-Demand Distance-Vector Routing Protocol,” March 2003.). One important consideration here may be to avoid premature energy depletion, even in case that impairs other requirements.

[R12] 6LoWPAN routing protocols SHOULD allow for dynamically adaptive topologies and mobile nodes. When supporting dynamic topologies and mobile nodes, route maintenance should keep in mind the goal of a minimal routing state and routing protocol message overhead.

Topological node mobility may be the result of physical movement and/or of a changing radio environment; making it very likely that mobility needs to be handled even in a network with physically static nodes. 6LoWPAN does not make use of a separate protocol to maintain connectivity to moving nodes but expects the routing protocol to handle it.

In addition, some nodes may move from one 6LoWPAN to another and are expected to become functional members of the latter 6LoWPAN in a limited amount of time.

Building monitoring applications, for instance, have a number of requirements with respect to recovery and settling time for mobility that range between 5 and 20 seconds (section 5.3.1 of [RFC5867] (Martocci, J., De Mil, P., Riou, N., and W. Vermeylen, “Building Automation Routing Requirements in Low-Power and Lossy Networks,” June 2010.)). For more interactive applications such as used in home automation systems, where users are giving input and expect instant feedback, mobility requirements are also stricter and, for moves within a network, a convergence time below 0.5 seconds is commonly required (section 3.2 of [RFC5826] (Brandt, A., Buron, J., and G. Porcu, “Home Automation Routing Requirements in Low-Power and Lossy Networks,” April 2010.)). In industrial environments, where mobile equipment such as cranes move around, the support of vehicular speeds of up to 35 km/h are required to be supported by the routing protocol [RFC5673] (Pister, K., Thubert, P., Dwars, S., and T. Phinney, “Industrial Routing Requirements in Low-Power and Lossy Networks,” October 2009.). Currently, 6LoWPANs are not normally being used for such a fast mobility, but dynamic association and disassociation MUST be supported in 6LoWPAN.

There are several challenges that should be addressed by a 6LoWPAN routing protocol in order to create robust routing in dynamic environments:

[R13] A 6LoWPAN routing protocol SHOULD support various traffic patterns: point-to-point, point-to-multipoint, and multipoint-to-point, while avoiding excessive multicast traffic in a LoWPAN.

6LoWPANs often have point-to-multipoint or multipoint-to-point traffic patterns. Many emerging applications include point-to-point communication as well. 6LoWPAN routing protocols should be designed with the consideration of forwarding packets from/to multiple sources/destinations. Current documents of the ROLL working group explain that the workload or traffic pattern of use cases for LoWPANs tends to be highly structured, unlike the any-to-any data transfers that dominate typical client and server workloads. In many cases, exploiting such structure may simplify difficult problems arising from resource constraints or variation in connectivity.



 TOC 

5.4.  Support of Security

The routing requirement described in this subsection allows secure transmission of routing messages. As in traditional networks, routing mechanisms in 6lowpan present another window from which, an attacker might disrupt and significantly degrade the 6lowpan overall performance. Attacks against unsecure routing aim mainly to contaminate WPAN networks with false routing information resulting in routing inconsistencies. A malicious node can also snoop packets and then launch replay attacks on the 6lowpan nodes. These attacks can cause harm especially when the attacker is a high-power device, such as laptop. It can also easily drain 6lowpan devices batteries by sending broadcast messages, redirecting routes etc.

[R14] 6LoWPAN routing protocols MUST support confidentiality, authentication and integrity services as required for secure delivery of control messages.

A general set of requirements that may apply to these services can be found in [I‑D.ietf‑karp‑threats‑reqs] (Lebovitz, G., Bhatia, M., and R. White, “The Threat Analysis and Requirements for Cryptographic Authentication of Routing Protocols' Transports,” October 2010.).

Security is very important for designing robust routing protocols, but it should not cause significant transmission overhead. The security aspect, however, seems a bit tradeoff in the 6lowpan since security is always a costly function. 6lowpan poses unique challenges to which, traditional security techniques cannot be applied directly. For example, public key cryptography primitives are typically avoided (as being too expensive) as are relatively heavyweight conventional encryption methods.

Consequently, it becomes questionable whether the 6lowpan devices can support IPsec as it is. While IPsec is mandatory with IPv6, considering the power constraints and limited processing capabilities of IEEE 802.15.4 capable devices, IPsec is computationally expensive; Internet key exchange (IKEv2) messaging described in RFC5996 [RFC5996] (Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, “Internet Key Exchange Protocol Version 2 (IKEv2),” September 2010.) will not work well in 6lowpans as we want to minimize the amount of signaling in these networks. IPsec supports AH for authenticating the IP header and ESP for authenticating and encrypting the payload. The main issues of using IPsec are two-fold: (1) processing power and (2) key management. Since these tiny 6lowpan devices do not process huge number of data or communicate with many different nodes, it is not well understood if complete implementation of SADB, policy-database and dynamic key-management protocol are appropriate for these small battery powered devices.

Bandwidth is a very scarce resource in 6lowpan environments. The fact that IPsec additionally requires another header (AH or ESP) in every packet makes its use problematic in 6lowpan environments. IPsec requires two communicating peers to share a secret key that is typically established dynamically with the Internet Key Exchange (IKEv2) protocol. Thus, it has an additional packet overhead incurred by IKEv2 packets exchange.

Given existing constraints in 6lowpan environments, IPsec may not be suitable to use in such environments, especially that 6lowpan node may not be able to operate all IPsec algorithms on its own capability. Thus, 6lowpan may need to define its own keying management method(s) that requires minimum overhead in packet-size and in number of signaling messages exchange. IPsec will provide authentication and confidentiality between end-nodes and across multiple lowpan- links, and may be useful only when two nodes want to apply security to all exchanged messages. However, in most cases, the security may be requested at the application layer as needed, while other messages can flow in the network without security overhead.

Security threats within LoWPANs may be different from existing threat models in ad-hoc network environments. If IEEE 802.15.4 security is not used, Neighbor Discovery (ND) in IEEE 802.15.4 links is susceptible to threats. These include NS/NA spoofing, malicious router, default router killed, good router goes bad, spoofed redirect, replay attacks and remote ND DoS [RFC3756] (Nikander, P., Kempf, J., and E. Nordmark, “IPv6 Neighbor Discovery (ND) Trust Models and Threats,” May 2004.). However, if IEEE 802.15.4 security is used, no other protection is needed for ND, as long as none of the nodes becomes compromised, because the Corporate Intranet Model of RFC 3756 can be assumed [I‑D.ietf‑6lowpan‑nd] (Shelby, Z., Chakrabarti, S., and E. Nordmark, “Neighbor Discovery Optimization for Low-power and Lossy Networks,” October 2010.).

Bootstrapping may also impose additional threats. For example, a malicious node can obtain initial configuration information in order to appear as a legitimate node and then carry out various types of attacks. Such a node can also keep legitimate nodes busy by broadcasting authentication/join requests. One option for mitigating such threats is the use of mutual authentication schemes based on the use of pre-shared keys [refs.Ikram] (Ikram, M., “A Simple Lightweight Authentic Bootstrapping Protocol for IPv6-based Low Rate Wireless Personal Area Networks (6LoWPANs),” June 2009.).

The IEEE 802.15.4 MAC provides an AES-based security mechanism. Routing protocols may define how this mechanism (in conjunction with IP security whenever available) can be used to obtain the intended security, either for the routing protocol alone or in conjunction with the security used for the data. Byte overhead of the mechanism, which depends on the security services selected, must be considered. In the worst case in terms of overhead, the mechanism consumes 21 bytes of MAC payload.

The IEEE 802.15.4 MAC security is typically supported by crypto hardware even in very simple chips that will be used in a 6LoWPAN. Even if the IEEE 802.15.4 MAC security mechanisms are not used, this crypto hardware is usually available for use by application code running on these chips. A security protocol outside IEEE 802.15.4 MAC security SHOULD therefore provide a mode of operation that is covered by this crypto hardware.

IEEE 802.15.4 does not specify protection for acknowledgement frames. Since the sequence numbers of data frames are sent in the clear, an adversary can forge an acknowledgement for each data frame. This weakness can be combined with targeted jamming to prevent delivery of selected packets. In consequence, IEEE 802.15.4 acknowledgements cannot be relied upon. In applications that require high security, the routing protocol must not exploit feedback from acknowledgements (e.g. to keep track of neighbor connectivity, see [R16]).



 TOC 

5.5.  Support of Mesh Under Forwarding

One LoWPAN may be built as one IPv6 link. In this case, Mesh Under forwarding mechanisms must be supported. The requirements described in this subsection allow optimization and correct operation of routing solutions taking into account the specific features of the Mesh Under configuration.

[R15] For Mesh Under, the forwarding mechanisms MUST support 16-bit short and 64-bit extended MAC addresses.

[R16] In order to perform discovery and maintenance of neighbors (i.e., neighborhood discovery as opposed to ND-style neighbor discovery), LoWPAN Nodes SHOULD avoid sending separate "Hello" messages. Instead, link-layer mechanisms (such as acknowledgments) MAY be utilized to keep track of active neighbors.

Reception of an acknowledgement after a frame transmission may render unnecessary the transmission of explicit Hello messages, for example. In a more general view, any frame received by a node may be used as an input to evaluate the connectivity between the sender and receiver of that frame.

[R17] If the routing protocol functionality includes enabling IP multicast, then it MAY employ structure in the network for efficient distribution in order to minimize link layer broadcast.



 TOC 

6.  Security Considerations

Security issues are described in Section 5.4. The security considerations in RFC 4919 [RFC4919] (Kushalnagar, N., Montenegro, G., and C. Schumacher, “IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals,” August 2007.), RFC 4944 [RFC4944] (Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” September 2007.) and RFC 4593 [RFC4593] (Barbir, A., Murphy, S., and Y. Yang, “Generic Threats to Routing Protocols,” October 2006.) apply as well.

The use of wireless links renders a 6LoWPAN susceptible to attacks like any other wireless network. In outdoor 6LoWPANs, the physical exposure of the nodes allows an adversary to capture, clone or tamper with these devices. In ad-hoc 6LoWPANs that are dynamic in both their topology and node memberships, a static security configuration does not suffice. Spoofed, altered, or replayed routing information might occur while multihopping could delay the detection and treatment of attacks.

This specification expects that the link layer is sufficiently protected, either by means of physical or IP security for the backbone link or with MAC sublayer cryptography. However, link-layer encryption and authentication may not be sufficient to provide confidentiality, authentication, integrity, and freshness to both data and routing protocol packets. Time synchronization, self-organization and secure localization for multi-hop routing are also critical to support.

For secure routing protocol operation, it may be necessary to consider authenticated broadcast (and multicast) and bidirectional link verification. On the other hand, secure end-to-end data delivery can be assisted by the routing protocol. For example, multi-path routing could be considered for increasing security to prevent selective forwarding. However, the challenge is that 6LoWPANs already have high resource constraints, so that 6LBR and LoWPAN nodes may require different security solutions.



 TOC 

7.  IANA Considerations

This document contains no actions for IANA.



 TOC 

8.  Acknowledgements

The authors highly appreciate the authors of "6LoWPAN security analysis" document (draft-daniel-6lowpan-security-analysis-04). Although their security analysis work is not continuous at this moment, the valuable information and text of the docuement is used in Section 5.4 in this docuement, by advice during IESG review procedures. Thanks to the work, the Section 5.4 is well improved. The authors also thank S. Chakrabarti who gave valuable comments for mesh-under requirements and A. Petrescu for significant review.



 TOC 

9.  References



 TOC 

9.1. Normative References

[IEEE802.15.4] IEEE Computer Society, “IEEE Std. 802.15.4-2006 (as amended),” 2007.
[RFC2119] Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels,” BCP 14, RFC 2119, March 1997 (TXT, HTML, XML).
[RFC3756] Nikander, P., Kempf, J., and E. Nordmark, “IPv6 Neighbor Discovery (ND) Trust Models and Threats,” RFC 3756, May 2004 (TXT).
[RFC3819] Karn, P., Bormann, C., Fairhurst, G., Grossman, D., Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. Wood, “Advice for Internet Subnetwork Designers,” BCP 89, RFC 3819, July 2004 (TXT).
[RFC4593] Barbir, A., Murphy, S., and Y. Yang, “Generic Threats to Routing Protocols,” RFC 4593, October 2006 (TXT).
[RFC4919] Kushalnagar, N., Montenegro, G., and C. Schumacher, “IPv6 over Low-Power Wireless Personal Area Networks (6LoWPANs): Overview, Assumptions, Problem Statement, and Goals,” RFC 4919, August 2007 (TXT).
[RFC4944] Montenegro, G., Kushalnagar, N., Hui, J., and D. Culler, “Transmission of IPv6 Packets over IEEE 802.15.4 Networks,” RFC 4944, September 2007 (TXT).
[RFC5548] Dohler, M., Watteyne, T., Winter, T., and D. Barthel, “Routing Requirements for Urban Low-Power and Lossy Networks,” RFC 5548, May 2009 (TXT).
[RFC5673] Pister, K., Thubert, P., Dwars, S., and T. Phinney, “Industrial Routing Requirements in Low-Power and Lossy Networks,” RFC 5673, October 2009 (TXT).


 TOC 

9.2. Informative References

[I-D.ietf-6lowpan-hc] Hui, J. and P. Thubert, “Compression Format for IPv6 Datagrams in 6LoWPAN Networks,” draft-ietf-6lowpan-hc-13 (work in progress), September 2010 (TXT).
[I-D.ietf-6lowpan-nd] Shelby, Z., Chakrabarti, S., and E. Nordmark, “Neighbor Discovery Optimization for Low-power and Lossy Networks,” draft-ietf-6lowpan-nd-14 (work in progress), October 2010 (TXT).
[I-D.ietf-karp-threats-reqs] Lebovitz, G., Bhatia, M., and R. White, “The Threat Analysis and Requirements for Cryptographic Authentication of Routing Protocols' Transports,” draft-ietf-karp-threats-reqs-01 (work in progress), October 2010 (TXT).
[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,” draft-ietf-roll-protocols-survey-07 (work in progress), April 2009 (TXT).
[RFC5556] Touch, J. and R. Perlman, “Transparent Interconnection of Lots of Links (TRILL): Problem and Applicability Statement,” RFC 5556, May 2009 (TXT).
[RFC5826] Brandt, A., Buron, J., and G. Porcu, “Home Automation Routing Requirements in Low-Power and Lossy Networks,” RFC 5826, April 2010 (TXT).
[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 (TXT).
[RFC5996] Kaufman, C., Hoffman, P., Nir, Y., and P. Eronen, “Internet Key Exchange Protocol Version 2 (IKEv2),” RFC 5996, September 2010 (TXT).
[refs.Chen] Chen, B., Muniswamy-Reddy, K., and M. Welsh, “Ad-Hoc Multicast Routing on Resource-Limited Sensor Nodes,” 2006.
[refs.Hill] Hill, J., “System Architecture for Wireless Sensor Networks.”
[refs.Ikram] Ikram, M., “A Simple Lightweight Authentic Bootstrapping Protocol for IPv6-based Low Rate Wireless Personal Area Networks (6LoWPANs),” June 2009.
[refs.Kuhn] Kuhn, F., Wattenhofer, R., and A. Zollinger, “Worst-Case Optimal and Average-Case Efficient Ad-Hoc Geometric Routing,” 2003.
[refs.Latre] Latre, M., De Mil, P., Moerman, I., Dhoedt, B., and P. Demeester, “Throughput and Delay Analysis of Unslotted IEEE 802.15.4,” May 2006.
[refs.Lee] Lee, S., Belding-Royer, E., and C. Perkins, “Scalability Study of the Ad Hoc On-Demand Distance-Vector Routing Protocol,” March 2003.
[refs.Shih] Shih, E., “Physical Layer Driven Protocols and Algorithm Design for Energy-Efficient Wireless Sensor Networks,” July 2001.
[refs.SmartDust] Pister, K. and B. Boser, “Smart Dust: Wireless Networks of Millimeter-Scale Sensor Nodes.”
[refs.bulusu] Bulusu, N. and S. Jha, “Wireless Sensor Networks,” July 2005.
[refs.cctc] Lu, J., Valois, F., Dohler, M., and D. Barthel, “Quantifying Organization by Means of Entropy,” 2008.


 TOC 

Authors' Addresses

  Eunsook Eunah Kim
  ETRI
  161 Gajeong-dong
  Yuseong-gu
  Daejeon 305-700
  Korea
Phone:  +82-42-860-6124
Email:  eunah.ietf@gmail.com
  
  Dominik Kaspar
  Simula Research Laboratory
  Martin Linges v 17
  Fornebu 1364
  Norway
Phone:  +47-6782-8223
Email:  dokaspar.ietf@gmail.com
  
  Carles Gomez
  Tech. Univ. of Catalonia/i2CAT
  Escola Politecnica Superior de Castelldefels
  C/Esteve Terradas, 7
  Castelldefels 08860
  Spain
Phone:  +34-93-413-7206
Email:  carlesgo@entel.upc.edu
  
  Carsten Bormann
  Universität Bremen TZI
  Postfach 330440
  Bremen D-28359
  Germany
Phone:  +49-421-218-63921
Fax:  +49-421-218-7000
Email:  cabo@tzi.org