3 hours course on IEEE and IETF protocols introducing the 6TiSCH architecture and the RPL routing protocol. Course given at telecom Bretagne on Feb 12th 2014
So let’s take a look at the characteristics on an LLN. What is it that makes an LLNAs mentioned before the devices are highly constrained, so we need to keep the state in each device to a minimum – for example we would not want the entire Link State database in every SensorIt is critical than an LLN uses the minimum amount of energyThere is a various array of traffic patterns - multipoint and point2pointProbably the most important thing to note is LLNs operate over networks with very restricted frame sizesWe can’t have giant packets running over poor quality linksThe routing protocol has to be really efficient – we can’t have it catering for every event/situation possible.
Wespeak of iot but most of the time thisisreally about ITOT, thatis the integrationof OT and IT.Benefits of convergence. 40 yearsago, OT = OsTralia. In convergence isindustrialthat’scalled indus internet.Cheerscale => IPv6Evolution thatis happening. Eg HART-IP . Startingnaturally in the plant net, then control net thendevices. We are creating a cross fire to accelerate the adoption.130 millions ports, 25% Ethernet. Split of H1 (FF, wireless) and H2 (ethernet) fieldbuses .The convergence of IT and OT technologies, aka the Industrial Internet, represents a multibillion opportunity for IT vendors.Our strategy to promote that convergence includes pushing IPv6 to the industrial M2M endpoints to enable end-to-end connectivity over deterministic wireless networks that are being developed at IEEE to emulate the characteristics of incumbent OT networks This work will benefit IOTG (CIBU)IOTG has already started to incorporate partner solutions (1552S, 1552WU) but in a fragmented market. This project approach will converge and promote an international standard solution (IPv6, 6TiSCH, CoAP, 802.15.4e based). The project leverages standards to drive IPv6 at the endpoints (that we do not own), which in turn justifies the introduction of Cisco equipment in the factory
• Direct-sequence spread spectrum (DSSS) modulation technique - this makes the ISA100.11a signal look like noise to other wireless systems.• Spatial diversity - Two field access points receive transmission from the field instrument.• Frequency diversity - Frequency hopping over the available channels in the bandwidth of the device's transmitting frequency.• Dynamic power control - Reduces possible interference with other wireless networks.• Channel black listing and adaptive channel hopping - Avoids congested channels. • Implementation of IEEE 802.15.4-2006 - Proven to coexist in very congested environments. • Careful management of the ISA100.11a wireless network implementation.
RPL builds a routing topology in the form of a DAG – a Directed Acyclic Graph (tree vsDAGs)It is a distance vector protocol that is proactive – it can build alternate paths during topology setup – rather than reactive where we rely on control plane messages after the failure to figure out the alternate path DV was chosen because the size of a link state DB would be way too largeHistorically, a number of interesting research initiatives on routing in WSN,Main focus on algorithms … a bit less on architectureMost work assuming the use of MAC addresses – L2 “routing” (mesh-under)Support of multiple PHY/MAC is a MUST: IEEE 802.15.4, LP Wifi, PLC (number of flavors), …Now … if what you want is a layered architecture supporting multiple PHY/MAC, there aren’t that many options …IP !RPL has been designed for lossy links including PLC such as P1901.2, which are of the utmost important for example for AMI in Europe. Sincethese links do exhibit similar properties in terms of instability, BER, ... RPL is well suited for these networks too. I had to fight a bit when we first determineROLL's charter to make sure that non RF links such as PLC were part of the charter (strictly speaking we are independent of the L3 of course but this was to insist onthat decoupling and point out that RPL was a routing protocol for LLN).
RPL builds a routing topology in the form of a DAG – a Directed Acyclic Graph (tree vsDAGs)It is a distance vector protocol that is proactive – it can build alternate paths during topology setup – rather than reactive where we rely on control plane messages after the failure to figure out the alternate path DV was chosen because the size of a link state DB would be way too largeHistorically, a number of interesting research initiatives on routing in WSN,Main focus on algorithms … a bit less on architectureMost work assuming the use of MAC addresses – L2 “routing” (mesh-under)Support of multiple PHY/MAC is a MUST: IEEE 802.15.4, LP Wifi, PLC (number of flavors), …Now … if what you want is a layered architecture supporting multiple PHY/MAC, there aren’t that many options …IP !RPL has been designed for lossy links including PLC such as P1901.2, which are of the utmost important for example for AMI in Europe. Sincethese links do exhibit similar properties in terms of instability, BER, ... RPL is well suited for these networks too. I had to fight a bit when we first determineROLL's charter to make sure that non RF links such as PLC were part of the charter (strictly speaking we are independent of the L3 of course but this was to insist onthat decoupling and point out that RPL was a routing protocol for LLN).
A local RPLInstanceID is autoconfigured by the node that owns the DODAGID and it MUST be unique for that DODAGID. The DODAGID used to configure the local RPLInstanceID MUST be a reachable IPv6 address of the node, and MUST be used as an endpoint of all communications within that local instance.
RPL builds a routing topology in the form of a DAG – a Directed Acyclic Graph (tree vsDAGs)It is a distance vector protocol that is proactive – it can build alternate paths during topology setup – rather than reactive where we rely on control plane messages after the failure to figure out the alternate path DV was chosen because the size of a link state DB would be way too largeHistorically, a number of interesting research initiatives on routing in WSN,Main focus on algorithms … a bit less on architectureMost work assuming the use of MAC addresses – L2 “routing” (mesh-under)Support of multiple PHY/MAC is a MUST: IEEE 802.15.4, LP Wifi, PLC (number of flavors), …Now … if what you want is a layered architecture supporting multiple PHY/MAC, there aren’t that many options …IP !RPL has been designed for lossy links including PLC such as P1901.2, which are of the utmost important for example for AMI in Europe. Sincethese links do exhibit similar properties in terms of instability, BER, ... RPL is well suited for these networks too. I had to fight a bit when we first determineROLL's charter to make sure that non RF links such as PLC were part of the charter (strictly speaking we are independent of the L3 of course but this was to insist onthat decoupling and point out that RPL was a routing protocol for LLN).