| Roll | A. Brandt |
| Internet-Draft | Sigma Designs |
| Intended status: Standards Track | E. Baccelli |
| Expires: July 10, 2018 | INRIA |
| R. Cragie | |
| ARM Ltd. | |
| P. van der Stok | |
| Consultant | |
| January 6, 2018 |
Applicability Statement: The use of the RPL protocol suite in Home Automation and Building Control
draft-ietf-roll-applicability-home-building-10
The purpose of this document is to provide guidance in the selection and use of protocols from the RPL protocol suite to implement the features required for control in building and home environments.
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 https://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 July 10, 2018.
Copyright (c) 2018 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 (https://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.
The primary purpose of this document is to give guidance in the use of the Routing Protocol for Low power and lossy networks (RPL) protocol suite in two application domains:
The guidance is based on the features required by the requirements documents "Home Automation Routing Requirements in Low-Power and Lossy Networks" [RFC5826] and "Building Automation Routing Requirements in Low-Power and Lossy Networks" [RFC5867] respectively. The Advanced Metering Infrastructure is also considered where appropriate. The applicability domains distinguish themselves in the way they are operated, their performance requirements, and the most likely network structures. An abstract set of distinct communication paradigms is then used to frame the applicability domains.
Home automation and building automation application domains share a substantial number of properties:
The differences between the two application domains mostly appear in commissioning, maintenance and the user interface, which do not typically affect routing. Therefore, the focus of this applicability document is on reliability, timeliness, and local routing.
The Routing Over Low power and Lossy networks (ROLL) working group has specified a set of routing protocols for Low-Power and Lossy Networks (LLN) [RFC6550]. This applicability text describes a subset of those protocols and the conditions under which the subset is appropriate and provides recommendations and requirements for the accompanying parameter value ranges.
In addition, an extension document has been produced specifically to provide a solution for reactive discovery of point-to-point routes in LLNs [RFC6997]. The present applicability document provides recommendations and requirements for the accompanying parameter value ranges.
A common set of security threats are described in [RFC7416]. The applicability statements complement the security threats document by describing preferred security settings and solutions within the applicability statement conditions. This applicability statement recommends lighter weight security solutions appropriate for home and building environments and indicates why these solutions are appropriate.
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].
Additionally, this document uses terminology from [RFC6997], [I-D.ietf-roll-trickle-mcast], [RFC7102], [IEEE802.15.4], and [RFC6550].
Applicable requirements are described in [RFC5826] and [RFC5867]. A survey of the application field is described in [BCsurvey].
The considered network diameter is limited to a maximum diameter of 10 hops and a typical diameter of 5 hops, which captures the most common cases in home automation and building control networks.
This document does not consider the applicability of Routing Protocol for Low-Power and Lossy Networks (RPL)-related specifications for urban and industrial applications [RFC5548], [RFC5673], which may exhibit significantly larger network diameters.
The use of communications networks in buildings is essential to satisfy energy saving regulations. Environmental conditions of buildings can be adapted to suit the comfort of the individuals present inside. Consequently when no one is present, energy consumption can be reduced. Cost is the main driving factor behind deployment of wireless networking in buildings, especially in the case of retrofitting, where wireless connectivity saves costs incurred due to cabling and building modifications.
A typical home automation network is comprised of less than 100 nodes. Large building deployments may span 10,000 nodes but to ensure uninterrupted service of light and air conditioning systems in individual zones of the building, nodes are typically organized in sub-networks. Each sub-network in a building automation deployment typically contains tens to hundreds of nodes, and for critical operations may operate independently from the other sub-networks.
The main purpose of the home or building automation network is to provide control over light and heating/cooling resources. User intervention via wall controllers is combined with movement, light and temperature sensors to enable automatic adjustment of window blinds, reduction of room temperature, etc. In general, the sensors and actuators in a home or building typically have fixed physical locations and will remain in the same home or building automation network.
People expect an immediate and reliable response to their presence or actions. For example, a light not switching on after entry into a room may lead to confusion and a profound dissatisfaction with the lighting product.
Monitoring of functional correctness is at least as important as timely responses. Devices typically communicate their status regularly and send alarm messages notifying a malfunction of controlled equipment or network.
In building control, the infrastructure of the building management network can be shared with the security/access, the Internet Protocol (IP) telephony, and the fire/alarm networks. This approach has a positive impact on the operation and cost of the network; however, care should be taken to ensure that the availability of the building management network does not become compromised beyond the ability for critical functions to perform adequately.
In homes, the entertainment network for audio/video streaming and gaming has different requirements, where the most important requirement is the need for high bandwidth not typically needed for home or building control. It is therefore expected that the entertainment network in the home will mostly be separate from the control network, which also lessens the impact on availability of the control network
In general, the home automation network or building control network consists of wired and wireless sub-networks. In large buildings especially, the wireless sub-networks can be connected to an IP backbone network where all infrastructure services are located, such as Domain Name System (DNS), automation servers, etc.
The wireless sub-network can be configured according to any of the following topologies:
Many of the nodes are battery-powered and may be sleeping nodes which wake up according to clock signals or external events.
In a building control network, for a large installation with multiple border routers, sub-networks often overlap both geographically and from a wireless coverage perspective. Due to two purposes of the network, (i) direct control and (ii) monitoring, there may exist two types of routing topologies in a given sub-network: (i) a tree-shaped collection of routes spanning from a central building controller via the border router, on to destination nodes in the sub-network; and/or (ii) a flat, un-directed collection of intra-network routes between functionally related nodes in the sub-network.
The majority of nodes in home and building automation networks are typically class 0 devices [RFC7228], such as individual wall switches. Only a few nodes (such as multi-purpose remote controls) are more expensive Class 1 devices, which can afford more memory capacity.
Traffic may enter the network originating from a central controller or it may originate from an intra-network node. The majority of traffic is light-weight point-to-point control style; e.g. Put-Ack or Get-Response. There are however exceptions. Bulk data transfer is used for firmware update and logging, where firmware updates enter the network and logs leave the network. Group communication is used for service discovery or to control groups of nodes, such as light fixtures.
Often, there is a direct physical relation between a controlling sensor and the controlled equipment. For example the temperature sensor and room controller are located in the same room sharing the same climate conditions. Consequently, the bulk of senders and receivers are separated by a distance that allows one-hop direct path communication. A graph of the communication will show several fully connected subsets of nodes. However, due to interference, multipath fading, reflection and other transmission mechanisms, the one-hop direct path may be temporally disconnected. For reliability purposes, it is therefore essential that alternative n-hop communication routes exist for quick error recovery. (See Appendix B for motivation.)
Looking over time periods of a day, the networks are very lightly loaded. However, bursts of traffic can be generated by e.g. incessant pushing of the button of a remote control, the occurrence of a defect, and other unforeseen events. Under those conditions, the timeliness must nevertheless be maintained. Therefore, measures are necessary to remove any unnecessary traffic. Short routes are preferred. Long multi-hop routes via the border router, should be avoided whenever possible.
Group communication is essential for lighting control. For example, once the presence of a person is detected in a given room, lighting control applies to that room only and no other lights should be dimmed, or switched on/off. In many cases, this means that a multicast message with a 1-hop and 2-hop radius would suffice to control the required lights. The same argument holds for Heating, Ventilating, and Air Conditioning (HVAC) and other climate control devices. To reduce network load, it is advisable that messages to the lights in a room are not distributed any further in the mesh than necessary based on intended receivers.
An example of an office surface is shown in [office-light], and the current use of wireless lighting control products is shown in [occuswitch].
Whilst air conditioning and other environmental-control applications may accept response delays of tens of seconds or longer, alarm and light control applications may be regarded as soft real-time systems. A slight delay is acceptable, but the perceived quality of service degrades significantly if response times exceed 250 ms. If the light does not turn on at short notice, a user may activate the controls again, thus causing a sequence of commands such as Light{on,off,on,off,..} or Volume{up,up,up,up,up,...}. In addition the repetitive sending of commands creates an unnecessary loading of the network, which in turn increases the bad responsiveness of the network.
This paradigm translates to many sources sending messages to the same sink, sometimes reachable via the border router. As such, source-sink (SS) traffic can be present in home and building networks. The traffic may be generated by environmental sensors (often present in a wireless sub-network) which push periodic readings to a central server. The readings may be used for pure logging, or more often, processed to adjust light, heating and ventilation. Alarm sensors may also generate SS style traffic. The central server in a home automation network will be connected mostly to a wired network segment of the home network, although it is likely that cloud services will also be used. The central server in a building automation network may be connected to a backbone or be placed outside the building.
With regards to message latency, most SS transmissions can tolerate worst-case delays measured in tens of seconds. Fire detectors, however, represent an exception; For example, special provisions with respect to the location of the Fire detectors and the smoke dampers need to be put in place to meet the stringent delay requirements measured in seconds.
This paradigm translates to a number of devices expressing their interest for a service provided by a server device. For example, a server device can be a sensor delivering temperature readings on the basis of delivery criteria, like changes in acquisition value or age of the latest acquisition. In building automation networks, this paradigm may be closely related to the SS paradigm given that servers, which are connected to the backbone or outside the building, can subscribe to data collectors that are present at strategic places in the building automation network. The use of PS will probably differ significantly from installation to installation.
This paradigm translates to a device transferring data to another device often connected to the same sub-network. Peer-to-peer (P2P) traffic is a common traffic type in home automation networks. Most building automation networks rely on P2P traffic, described in the next paragraph. Other building automation networks rely on P2P control traffic between controls and a local controller box for advanced group control. A local controller box can be further connected to service control boxes, thus generating more SS or PS traffic.
P2P traffic is typically generated by remote controls and wall controllers which push control messages directly to light or heat sources. P2P traffic has a stringent requirement for low latency since P2P traffic often carries application messages that are invoked by humans. As mentioned in Section 2.2.1 application messages should be delivered within a few hundred milliseconds - even when connections fail momentarily.
This paradigm translates to a device sending a message as many times as there are destination devices. Peer-to-multipeer (P2MP) traffic is common in home and building automation networks. Often, a thermostat in a living room responds to temperature changes by sending temperature acquisitions to several fans and valves consecutively. This paradigm is also closely related to the PS paradigm in the case where a single server device has multiple subscribers.
This paradigm translates to a device sending a message to many destinations in one network transfer invocation. Multicast is well suited for lighting where a presence sensor sends a presence message to a set of lighting devices. Multicast increases the probability that the message is delivered within the strict time constraints. The recommended multicast algorithm (e.g. [I-D.ietf-roll-trickle-mcast]) assures that messages are delivered to ALL intended destinations.
In the case of the SS paradigm applied to a wireless sub-network to a server reachable via a border router, the use of RPL [RFC6550] in non-storing mode is appropriate. Given the low resources of the devices, source routing will be used from the border router to the destination in the wireless sub-network for messages generated outside the mesh network. No specific timing constraints are associated with the SS type messages so network repair does not violate the operational constraints. When no SS traffic takes place, it is good practice to load only RPL code enabling P2P mode of operation [RFC6997] to reduce the code size and satisfy memory requirements.
P2P-RPL [RFC6997] is required for all P2P and P2MP traffic taking place between nodes within a wireless sub-network (excluding the border router) to assure responsiveness. Source and destination devices are typically physically close based on room layout. Consequently, most P2P and P2MP traffic is 1-hop or 2-hop traffic. Appendix A explains why P2P-RPL is preferable to RPL for this type of communication. Appendix B explains why reliability measures such as multi-path routing are necessary even when 1-hop communication dominates.
Additional advantages of P2P-RPL for home and building automation networks are, for example: Section 4.1.2.
Due to the limited memory of the majority of devices, P2P-RPL SHOULD be deployed with source routing in non-storing mode as explained in
Multicast with Multicast Protocol for Low power and Lossy Networks (MPL) [I-D.ietf-roll-trickle-mcast] is preferably deployed for N-cast over the wireless network. Configuration constraints that are necessary to meet reliability and timeliness with MPL are discussed in Section 4.1.7.
This document applies to [IEEE802.15.4] and [G.9959] which are adapted to IPv6 by the adaption layers [RFC4944] and [RFC7428]. Other layer-2 technologies, accompanied by an "IP over Foo" specification, are also relevant provided there is no frame size issue, and there are link layer acknowledgements.
The above mentioned adaptation layers leverage on the compression capabilities of [RFC6554] and [RFC6282]. Header compression allows small IP packets to fit into a single layer 2 frame even when source routing is used. A network diameter limited to 5 hops helps to achieve this even while using source routing.
Dropped packets are often experienced in the targeted environments. Internet Control Message Protocol (ICMP), User Datagram Protocol (UDP) and even Transmission Control Protocol (TCP) flows may benefit from link layer unicast acknowledgments and retransmissions. Link layer unicast acknowledgments SHOULD be enabled when [IEEE802.15.4] or [G.9959] is used with RPL and P2P-RPL.
Several features required by [RFC5826], [RFC5867] challenge the P2P paths provided by RPL. Appendix A reviews these challenges. In some cases, a node may need to spontaneously initiate the discovery of a path towards a desired destination that is neither the root of a DAG, nor a destination originating Destination Advertisement Object (DAO) signalling. Furthermore, P2P paths provided by RPL are not satisfactory in all cases because they involve too many intermediate nodes before reaching the destination.
P2P-RPL [RFC6997] SHOULD be used in home automation and building control networks, as point-to-point style traffic is substantial and route repair needs to be completed within seconds. P2P-RPL provides a reactive mechanism for quick, efficient and root-independent route discovery/repair. The use of P2P-RPL furthermore allows data traffic to avoid having to go through a central region around the root of the tree, and drastically reduces path length [SOFT11] [INTEROP12]. These characteristics are desirable in home and building automation networks because they substantially decrease unnecessary network congestion around the root of the tree.
When more reliability is required, P2P-RPL enables the establishment of multiple independent paths. For 1-hop destinations this means that one 1-hop communication and a second 2-hop communication take place via a neighbouring node. Such a pair of redundant communication paths can be achieved by using MPL where the source is a MPL forwarder, while a second MPL forwarder is 1 hop away from both the source and the destination node. When the source multicasts the message, it may be received by both the destination and the 2nd forwarder. The 2nd forwarder forwards the message to the destination, thus providing two routes from sender to destination.
To provide more reliability with multiple paths, P2P-RPL can maintain two independent P2P source routes per destination, at the source. Good practice is to use the paths alternately to assess their existence. When one P2P path has failed (possibly only temporarily), as described in Appendix B, the alternative P2P path can be used without discarding the failed path. The failed P2P path, unless proven to work again, can be safely discarded after a timeout (typically 15 minutes). A new route discovery is done when the number of P2P paths is exhausted due to persistent link failures.
P2P-RPL SHOULD be used in home automation and building control networks. Its reactive discovery allows for low application response times even when on-the-fly route repair is needed. Non-storing mode SHOULD be used to reduce memory consumption in repeaters with constrained memory when source routing is used.
An important constraint on the application of RPL is the presence of sleeping nodes.
For example, in a stand-alone network, the master node (or coordinator) providing the logical layer-2 identifier and unique node identifiers to connected nodes may be a remote control which returns to sleep once new nodes have been added. Due to the absence of the border router, there may be no global routable prefixes at all. Likewise, there may be no authoritative always-on root node since there is no border router to host this function.
In a network with a border router and many sleeping nodes, there may be battery powered sensors and wall controllers configured to contact other nodes in response to events and then return to sleep. Such nodes may never detect the announcement of new prefixes via multicast.
In each of the above mentioned constrained deployments, a link layer node (e.g. coordinator or master) SHOULD assume the role of authoritative root node, transmitting unicast Router Advertisement (RA) messages with a Unique Local Address (ULA) prefix information option to nodes during the joining process to prepare the nodes for a later operational phase, where a border router is added.
A border router SHOULD be designed to be aware of sleeping nodes in order to support the distribution of updated global prefixes to such sleeping nodes.
When operating P2P-RPL on a stand-alone basis, there is no authoritative root node maintaining a permanent RPL Direction-Oriented Directed Acyclic Graph (DODAG). A node MUST be able to join at least one RPL instance, as a new, temporary instance is created during each P2P-RPL route discovery operation. A node MAY be designed to join multiple RPL instances.
Non-storing mode MUST be used to cope with the extremely constrained memory of a majority of nodes in the network (such as individual light switches).
Nodes send DAO messages to establish downward paths from the root to themselves. DAO messages are not acknowledged in networks composed of battery operated field devices in order to minimize the power consumption overhead associated with path discovery. The DAO messages build up a source route because the nodes MUST be in non-storing mode.
If devices in LLNs participate in multiple RPL instances and DODAGs, both the RPLInstance ID and the DODAGID SHOULD be included in the DAO.
Expected Transmission Count (ETX) is the RECOMMENDED metric. [RFC6551] provides other options.
Packets from asymmetric and/or unstable channels SHOULD be deleted at layer 2.
Objective Function 0 (OF0) MUST be the Objective Function. Other Objective Functions MAY be used when dictated by circumstances.
Since P2P-RPL only creates DODAGs on a temporary basis during route repair or route discovery, there is no need to repair DODAGs.
For SS traffic, local repair is sufficient. The accompanying process is known as poisoning and is described in Section 8.2.2.5 of [RFC6550]. Given that the majority of nodes in the building do not physically move around, creating new DODAGs should not happen frequently.
Commercial lighting deployments may have a need for multicast to distribute commands to a group of lights in a timely fashion. Several mechanisms exist for achieving such functionality; [I-D.ietf-roll-trickle-mcast] is the RECOMMENDED protocol for home and building deployments. This section relies heavily on the conclusions of [RT-MPL].
At reception of a packet, the MPL forwarder starts a series of consecutive trickle timer intervals, where the first interval has a minimum size of Imin. Each consecutive interval is twice as long as the former with a maximum value of Imax. There is a maximum number of intervals given by max_expiration. For each interval of length I, a time t is randomly chosen in the period [I/2, I]. For a given packet, p, MPL counts the number of times it receives p during the period [0, t] in a counter c. At time t, MPL re-broadcasts p when c < k, where k is a predefined constant with a value k > 0.
The density of forwarders and the frequency of message generation are important aspects to obtain timeliness during control operations. A high frequency of message generation can be expected when a remote control button is incessantly pressed, or when alarm situations arise.
Guaranteeing timeliness is intimately related to the density of the MPL routers. In ideal circumstances the message is propagated as a single wave through the network, such that the maximum delay is related to the number of hops times the smallest repetition interval of MPL. Each forwarder that receives the message passes the message on to the next hop by repeating the message. When several copies of a message reach the forwarder, it is specified that the copy need not be repeated. Repetition of the message can be inhibited by a small value of k. To assure timeliness, the value of k should be chosen high enough to make sure that messages are repeated at the first arrival of the message in the forwarder. However, a network that is too dense leads to a saturation of the medium that can only be prevented by selecting a low value of k. Consequently, timeliness is assured by choosing a relatively high value of k but assuring at the same time a low enough density of forwarders to reduce the risk of medium saturation. Depending on the reliability of the network channels, it is advisable to choose the network such that at least 2 forwarders per hop repeat messages to the same set of destinations.
There are no rules about selecting forwarders for MPL. In buildings with central management tools, the forwarders can be selected, but in the home is not possible to automatically configure the forwarder topology at the time of writing this document.
RPL MAY use unsecured messages to reduce message size. If there is a single node that uses unsecured RPL messages, link-layer security MUST be present.(see Section 7). If RPL is used with secured messages [RFC6550], the following RPL security parameter values SHOULD be used:
[RFC6997] MUST be used to accommodate P2P traffic, which is typically substantial in home and building automation networks.
Assigned IP addresses MUST be routable and unique within the routing domain [RFC5889].
No particular requirements exist for layer 2 but for the ones cited in the IP over Foo RFCs. (See Section 2.3)
Not applicable
Not applicable
Not applicable
Not applicable
The following sections describe the recommended parameter values for P2P-RPL and Trickle.
Trickle is used to distribute network parameter values to all nodes without stringent time restrictions. The recommended Trickle parameter values are:
When a node sends a changed DIO, this is an inconsistency and forces the receiving node to respond within Imin. So when something happens which affects the DIO, the change is ideally communicated to a node, n hops away, within n times Imin. Often, dependent on the node density, packets are lost, or not sent, leading to larger delays.
In general we can expect DIO changes to propagate within 1 to 3 seconds within the envisaged networks.
When nothing happens, the DIO sending interval increases to 4.37 minutes, thus drastically reducing the network load. When a node does not receive DIO messages during more than 10 minutes it can safely conclude the connection with other nodes has been lost.
This section discusses the P2P-RPL parameters.
P2P-RPL [RFC6997] provides the features requested by [RFC5826] and [RFC5867]. P2P-RPL uses a subset of the frame formats and features defined for RPL [RFC6550] but may be combined with RPL frame flows in advanced deployments.
The recommended parameter values for P2P-RPL are:
MPL is used to distribute values to groups of devices. Using MPL, based on the Trickle algorithm, timeliness should also be guaranteed. A deadline of 200 ms needs to be met when human action is followed by an immediately observable action such as switching on lights. The deadline needs to be met in a building where the number of hops from seed to destination varies between 1 and 10.
When the network is heavily loaded, MAC delays contribute significantly to the end to end delays when MPL intervals between 10 to 100 ms are used to meet the 200 ms deadline. It is possible to set the number of buffers in the MAC to 1 and set the number of Back-off repetitions to 1. The number of MPL repetitions compensates for the reduced probability of transmission per MAC invocation [RT-MPL].
In addition, end to end delays and message losses are reduced, by adding a real-time layer between MPL and MAC to throw away the earliest messages (exploiting the MPL message numbering) and favour the most recent ones.
This section proposes values for the Trickle parameters used by MPL for the distribution of packets that need to meet a 200 ms deadline. The probability of meeting the deadline is increased by (1) choosing a small Imin value,(2) reducing the number of MPL intervals thus reducing the load, and (3) reducing the number of MPL forwarders to also reduce the load.
The consequence of this approach is that the value of k can be larger than 1 because network load reduction is already guaranteed by the network configuration.
Under the condition that the density of MPL repeaters can be limited, it is possible to choose low MPL repeat intervals (Imin) connected to k values such that k>1. The minimum value of k is related to:
Within the first MPL interval a limited number, q, of messages can be transmitted. Assuming a 3 ms transmission interval, q is given by q = Imin/3. Assuming that at most q message copies can reach a given forwarder within the first repeat interval of length Imin, the related MPL parameter values are suggested in the following sections.
The recommended value is Imin = 10 to 50 ms.
When Imin is chosen much smaller, the interference between the copies leads to significant losses given that q is much smaller than the number of repeated packets. With much larger intervals the probability that the deadline will be met decreases with increasing hop count.
The recommended value is Imax = 100 to 400 ms.
The value of Imax is less important than the value of max_expiration. Given an Imin value of 10 ms, the 3rd MPL interval has a value of 10*2*2 = 40 ms. When Imin has a value of 40 ms, the 3rd interval has a value of 160 ms. Given that more than 3 intervals are unnecessary, the Imax does not contribute much to the performance.
Other parameters are the k parameter and the max_expiration parameter.
k > q (see condition above). Under this condition and for small Imin, a value of k=2 or k=3 is usually sufficient to minimize the losses of packets in the first repeat interval.
max_expiration = 2 - 4. Higher values lead to more network load while generating copies which will probably not meet their deadline.
At this moment it is not clear how homenets will be managed. Consequently it is not clear which tools will be used and which parameters must be exposed for management.
In building control, management is mandatory. It is expected that installations will be managed using the set of currently available tools(including IETF tools like Management Information Base (MIB) modules, NETCONF modules, Dynamic Host Configuration Protocol (DHCP) and others) with large differences between the ways an installation is managed.
This section refers to the security considerations of [RFC6997], [RFC6550], [I-D.ietf-roll-trickle-mcast], and the counter measures discussed in sections 6 and 7 of [RFC7416].
Communications network security is based on providing integrity protection and encryption to messages. This can be applied at various layers in the network protocol stack based on using various credentials and a network identity.
The credentials which are relevant in the case of RPL are: (i) the credential used at the link layer in the case where link layer security is applied (see Section 7.1) or (ii) the credential used for securing RPL messages. In both cases, the assumption is that the credential is a shared key. Therefore, there MUST be a mechanism in place which allows secure distribution of a shared key and configuration of network identity. Both MAY be done using: (i) pre-installation using an out-of-band method, (ii) delivered securely when a device is introduced into the network or (iii) delivered securely by a trusted neighbouring device. The shared key MUST be stored in a secure fashion which makes it difficult to be read by an unauthorized party.
This document mandates that a layer-2 mechanism be used during initial and incremental deployment. Please see the following sections.
Wireless mesh networks are typically secured at the link layer in order to prevent unauthorized parties from accessing the information exchanged over the links. It is good practice to create a network of nodes which share the same keys for link layer security and exclude nodes sending unsecured messages. With per-message data origin authentication, it is possible to prevent unauthorized nodes joining the mesh.
At initial deployment the network is secured by consecutively securing nodes at the link layer, thus building a network of secured nodes. The Protocol for carrying Authentication for Network Access (PANA) [RFC5191] [RFC6345] with an Extensible Authentication Protocol (EAP) provides a framework for network access and delivery of common link keys. Several versions of EAP exist. ZigBee specifies the use of EAP-TLS [RFC5216] (see section 5 of [ZigBeeIP]. Wi-SUN HAN (Home Area Network) uses EAP-PSK [RFC4764] (see section 5.6 of [WI-SUN]), which also looks promising for building control at this moment.
This document does not specify a multicast security solution. Networks deployed with this specification will depend upon layer-2 security to prevent outsiders from sending multicast traffic. It is recognized that this does not protect this control traffic from impersonation by already trusted devices. This is an area for a future specification.
For building control an installer will probably use an installation tool that establishes a secure communication path with the joining node. It is recognized that the recommendations for initial deployment of Section 7 and Section 7.1 do not cover all building requirements such as selecting the node-to-secure independent of network topology.
In the home, nodes can be visually inspected by the home owner and a simple procedure, e.g. pushing buttons simultaneously on an already secured device and an unsecured joining device is usually sufficient to ensure that the unsecured joining device is authenticated and configured securely, and paired appropriately.
This recommendation is in line with the countermeasures described in section 6.1.1 of [RFC7416].
Normally, the network remains secure by not allowing the addition of new nodes. If a new node needs to be added to the network, the network is usually configured to allow the new node to join via an assisting node in the manner described in Section 7.1. If an existing node becomes lost, it is usually possible to re-key all other existing nodes to isolate the lost node to ensure that, should it be found again, it has to re-join as if it were a new node.
Refer to the security considerations of [RFC6997].
The routing of MPL is determined by the enabling of the interfaces for specified Multicast addresses. The specification of these addresses can be done via a Constrained Application Protocol (CoAP) application as specified in [RFC7390]. An alternative is the creation of a MPL MIB and use of Simple Network Management Protocol (SNMP)v3 [RFC3411] or equivalent techniques to specify the Multicast addresses in the MIB. The application of security measures for the specification of the multicast addresses assures that the routing of MPL packets is secured.
This section follows the structure of section 7, "RPL security features" of [RFC7416], where a thorough analysis of security threats and proposed counter measures relevant to RPL and MPL are done.
In accordance with section 7.1 of [RFC7416], "Confidentiality features", a secured RPL protocol implements payload protection, as explained in Section 7 of this document. The attributes key-length and life-time of the keys depend on operational conditions, maintenance and installation procedures.
Section 7.1 and Section 7.2 of this document recommend link-layer measures to assure integrity in accordance with section 7.2 of [RFC7416], "Integrity features".
The provision of multiple paths recommended in section 7.3 "Availability features" of [RFC7416] is also recommended from a reliability point of view. Randomly choosing paths MAY be supported.
Key management discussed in section 7.4, "Key Management" of [RFC7416], is not standardized and discussions continue.
Section 7.5, "Considerations on Matching Application Domain Needs" of [RFC7416] applies as such.
Application and transport protocols used in home and building automation domains are expected to mostly consist in CoAP over UDP, or equivalents. Typically, UDP is used for IP transport to keep down the application response time and bandwidth overhead. CoAP is used at the application layer to reduce memory footprint and bandwidth requirements.
No considerations for IANA pertain to this document.
This document reflects discussions and remarks from several individuals including (in alphabetical order): Stephen Farrell, Mukul Goyal, Sandeep Kumar, Jerry Martocci, Catherine Meadows, Yoshira Ohba, Charles Perkins, Yvonne-Anne Pignolet, Michael Richardson, Ines Robles, Zach Shelby, and Meral Sherazipour.
RFC editor, please delete this section before publication.
Changes from version 0 to version 1.
Changes from version 1 to version 2.
Changes from version 2 to version 3.
Changes from version 3 to version 4.
Changes from version 4 to version 5.
Changes from version 5 to version 6.
Changes from version 6 to version 6.
| [BCsurvey] | Kastner, W., Neugschwandtner, G., Soucek, S. and H. Newman, "Communication Systems for Building Automation and Control", Proceedings of the IEEE Vol 93, No 6, June 2005. |
| [I-D.ietf-dice-profile] | Tschofenig, H. and T. Fossati, "TLS/DTLS Profiles for the Internet of Things", Internet-Draft draft-ietf-dice-profile-17, October 2015. |
| [I-D.keoh-dice-multicast-security] | Keoh, S., Kumar, S., Garcia-Morchon, O., Dijk, E. and A. Rahman, "DTLS-based Multicast Security in Constrained Environments", Internet-Draft draft-keoh-dice-multicast-security-08, July 2014. |
| [I-D.kumar-dice-dtls-relay] | Kumar, S., Keoh, S. and O. Garcia-Morchon, "DTLS Relay for Constrained Environments", Internet-Draft draft-kumar-dice-dtls-relay-02, October 2014. |
| [I-D.richardson-6tisch--security-6top] | Richardson, M., "6tisch secure join using 6top", Internet-Draft draft-richardson-6tisch--security-6top-05, November 2015. |
| [INTEROP12] | Baccelli, E., Phillip, M., Brandt, A., Valev , H. and J. Buron , "Report on P2P-RPL Interoperability Testing", RR-7864 INRIA Research Report RR-7864, January 2012. |
| [MEAS] | Holtman, K., "Connectivity loss in large scale IEEE 802.15.4 network", Private Communication, November 2013. |
| [occuswitch] | Lighting, Philips., "OccuSwitch wireless", Brochure, http://www.philipslightingcontrols.com/assets/cms/uploads/files/osw/MK_OSWNETBROC_5.pdf, May 2012. |
| [office-light] | Clanton and Associates, ., "A Life Cycle Cost Evaluation of Multiple Lighting Control Strategies", Wireless Lighting Control, http://www.daintree.net/wp-content/uploads/2014/02/clanton_lighting_control_report_0411.pdf, February 2014. |
| [RFC3411] | Harrington, D., Presuhn, R. and B. Wijnen, "An Architecture for Describing Simple Network Management Protocol (SNMP) Management Frameworks", STD 62, RFC 3411, DOI 10.17487/RFC3411, December 2002. |
| [RFC3561] | Perkins, C., Belding-Royer, E. and S. Das, "Ad hoc On-Demand Distance Vector (AODV) Routing", RFC 3561, DOI 10.17487/RFC3561, July 2003. |
| [RFC5889] | Baccelli, E. and M. Townsley, "IP Addressing Model in Ad Hoc Networks", RFC 5889, DOI 10.17487/RFC5889, September 2010. |
| [RFC6345] | Duffy, P., Chakrabarti, S., Cragie, R., Ohba, Y. and A. Yegin, "Protocol for Carrying Authentication for Network Access (PANA) Relay Element", RFC 6345, DOI 10.17487/RFC6345, August 2011. |
| [RFC7228] | Bormann, C., Ersue, M. and A. Keranen, "Terminology for Constrained-Node Networks", RFC 7228, DOI 10.17487/RFC7228, May 2014. |
| [RFC7390] | Rahman, A. and E. Dijk, "Group Communication for the Constrained Application Protocol (CoAP)", RFC 7390, DOI 10.17487/RFC7390, October 2014. |
| [RFC7428] | Brandt, A. and J. Buron, "Transmission of IPv6 Packets over ITU-T G.9959 Networks", RFC 7428, DOI 10.17487/RFC7428, February 2015. |
| [RT-MPL] | van der Stok, P., "Real-Time multicast for wireless mesh networks using MPL", White paper, http://www.vanderstok.org/papers/Real-time-MPL.pdf, April 2014. |
| [RTN2011] | Holtman, K. and P. van der Stok, "Real-time routing for low-latency 802.15.4 control networks", International Workshop on Real-Time Networks; Euromicro Conference on Real-Time Systems, July 2011. |
| [SOFT11] | Baccelli, E., Phillip, M. and M. Goyal, "The P2P-RPL Routing Protocol for IPv6 Sensor Networks: Testbed Experiments", Proceedings of the Conference on Software Telecommunications and Computer Networks, Split, Croatia,, September 2011. |
| [WI-SUN] | ECHONET Lite, ., "Home network Communication Interface for ECHONET Lite (IEEE802.15.4/4 e/4g 920MHz-band Wireless)", Japanese TTC standard JJ-300.10, May 2014. |
| [ZigBeeIP] | ZigBee Alliance, ., "ZigBee IP specification", ZigBee document 095023r34, March 2014. |
The DAG, being a tree structure is formed from a root. If nodes residing in different branches have a need for communicating internally, DAG mechanisms provided in RPL [RFC6550] will propagate traffic towards the root, potentially all the way to the root, and down along another branch [RFC6998]. In a typical example two nodes could reach each other via just two router nodes but in unfortunate cases, RPL may send traffic three hops up and three hops down again. This leads to several undesired phenomena described in the following sections
If many P2P data flows have to move up towards the root to get down again in another branch there is an increased risk of congestion the nearer to the root of the DAG the data flows. Due to the broadcast nature of RF systems any child node of the root is not just directing RF power downwards its sub-tree but just as much upwards towards the root; potentially jamming other MP2P traffic leaving the tree or preventing the root of the DAG from sending P2MP traffic into the DAG because the listen-before-talk link-layer protection kicks in.
Battery-powered nodes originating P2P traffic depend on the route length. Long routes cause source nodes to stay awake for longer periods before returning to sleep. Thus, a longer route translates proportionally (more or less) into higher battery consumption.
The RPL DAG mechanism uses DIO and DAO messages to monitor the health of the DAG. In rare occasions, changed radio conditions may render routes unusable just after a destination node has returned a DAO indicating that the destination is reachable. Given enough time, the next Trickle timer-controlled DIO/DAO update will eventually repair the broken routes, however this may not occur in a timely manner appropriate to the application. In an apparently stable DAG, Trickle-timer dynamics may reduce the update rate to a few times every hour. If a user issues an actuator command, e.g. light on in the time interval between the last DAO message was issued the destination module and the time one of the parents sends the next DIO, the destination cannot be reached. There is no mechanism in RPL to initiate restoration of connectivity in a reactive fashion. The consequence is a broken service in home and building applications.
Experience from the telecom industry shows that if the voice delay exceeds 250ms, users start getting confused, frustrated and/or annoyed. In the same way, if the light does not turn on within the same period of time, a home control user will activate the controls again, causing a sequence of commands such as Light{on,off,off,on,off,..} or Volume{up,up,up,up,up,...}. Whether the outcome is nothing or some unintended response this is unacceptable. A controlling system must be able to restore connectivity to recover from the error situation. Waiting for an unknown period of time is not an option. While this issue was identified during the P2P analysis, it applies just as well to application scenarios where an IP application outside the LLN controls actuators, lights, etc.
Measurements on the connectivity between neighbouring nodes are discussed in [RTN2011] and [MEAS].
The work is motivated by the measurements in literature which affirm that the range of an antenna is not circle symmetric but that the signal strength of a given level follows an intricate pattern around the antenna, and there may be holes within the area delineated by an iso-strength line. It is reported that communication is not symmetric: reception of messages from node A by node B does not imply reception of messages from node B by node A. The quality of the signal fluctuates over time, and also the height of the antenna within a room can have consequences for the range. As function of the distance from the source, three regions are generally recognized: (1) a clear region with excellent signal quality, (2) a region with fluctuating signal quality, (3) a region without reception. In the text below it is shown that installation of meshes with neighbours in the clear region is not sufficient.
[RTN2011] extends existing work by:
Eight nodes were distributed over a surface of 30m2. All nodes are at one hop distance from each other and are situated in the clear region of each other. Each node sends messages to each of its neighbours, and repeats the message until it arrives. The latency of the message was measured over periods of at least a week. It is noticed that latencies longer than a second occurred without apparent reasons, but only during working days and never in the weekends. Bad periods could last for minutes. By sending messages via two paths: (1) one hop path directly, and (2) two hop path via a randomly chosen neighbour, the probability of delays larger than 100 ms decreased significantly.
The conclusion is that even for 1-hop communication between not too distant "Line of Sight" nodes, there are periods of low reception in which communication deadlines of 200 ms are exceeded. It pays to send a second message over a 2-hop path to increase the reliability of timely message transfer.
[MEAS] confirms that temporary bad reception by close neighbours can occur within other types of areas. Nodes were installed on the ceiling in a grid with a distance of 30-50 cm between nodes. 200 nodes were distributed over an area of 10m x 5m. It clearly transpired that with increasing distance the probability of reception decreases. At the same time a few nodes furthest away from the sender had a high probability of message reception, while some close neighbours of the sender did not receive messages. The patterns of clear reception nodes evolved over time.
The conclusion is that even for direct neighbours reception can temporarily be bad during periods of several minutes. For a reliable and timely communication it is imperative to have at least two communication paths available (e.g. two hop paths next to the 1-hop path for direct neighbours).