Flexibilis Deterministic Switch (FDS)
With Flexibilis Deterministic Switch (FDS), it is possible to combine deterministic Ethernet with seamless redundancy, thus making the network very predictable and reliable.
FDS is a 10Mbps/100Mbps/1Gbps Ethernet Layer-2 switch implementing Time-sensitive Networking (TSN), including the IEEE 802.1CB redundancy protocol. It also employs IEEE 1588 v2 Precision Time Protocol. FDS is an IP core meant to be used in programmable hardware environments (FPGA).
Time-sensitive Networking (TSN)
TSN provides fully deterministic real-time communication over Ethernet. Ethernet is widely used and growing in popularity because it’s low-cost, well standardized and easily accessible. It is also scalable and offers a wide range of bandwidth.
However, regular Ethernet is not enough for many systems, because they require real-time and safety-critical communication. This includes industrial automation, where machines must share their data for synchronization, control, analysis and optimization. The same goes for energy production, where networking is needed for example for substation connectivity. Reliable communication is also necessary in automotive industry especially now that vehicles are becoming more and more automatized. Regular Ethernet does not fulfil the real-time communication requirements of these systems, because they need certainty about the communication latency times.
Time-sensitive Networking is the solution: it guarantees the communication latency times. Also network redundancy is often needed and most notable of the current redundancy protocols are IEEE 802.1CB and High-availability Seamless Redundancy (HSR). They provide redundancy with no single point of failure and zero time to recovery in case of a failure. The network is fully operational even during maintenance, as any device can be disconnected and replaced without breaking the network connectivity.
Because of TSN and redundancy protocols Ethernet becomes very attractive option for systems requiring real-time and safety-critical communication. Indeed, the usage of Ethernet is rapidly increasing in automotive, industrial automation and energy production. Ethernet and TSN will play a key role in the Industrial IoT and Industry 4.0. As machines and devices become connected, data will be more easily accessible – those who don’t take advantage of it will be left behind.
TSN is a set of IEEE 802 sub-standards. Of these standards, Flexibilis supports the following (some of the standards are still drafts):
- IEEE 802.1AS
- IEEE 802.1Q-2014, including e.g. IEEE 802.1BA, IEEE 802.1Qav, IEEE 802.1Qci
- IEEE 802.1CB
Traffic Policing and Shaping
TSN standard guarantees communication latency times with traffic policing and shaping methods combined with resource reservation. In traffic policing and shaping, traffic bandwidth is measured and if it exceeds the configured rate, actions are taken to force the rate back to where it should be. Traffic shaper can delay frames while policer simply drops the excess traffic.
Traffic policer provides means for limiting the rate of incoming data stream(s) by dropping frames exceeding the configured rate. The policer needs to know the frame length to compare if it fits with the configured traffic rate.
Traffic shaper limits the transmit rate to comply with configured maximum allowed transmit rate. Shaping consumes more hardware resources than policing because it requires buffering the frames. Frames that exceed the configured rate are delayed increasing the latency they experience. The rate at which frames can be transmitted from certain priority queue can be configured, so traffic shaping can affect in which order the priority queues are emptied. Also shaper may drop the frames if buffers are full.
Flexibilis Deterministic Switch (FDS) has eight priority queues for every output interface. Queues are first-in first-out type. The priority of a frame is determined during inbound processing and the forwarding core places the frame to the right priority queue. If traffic shaping is not active, the queues are emptied in priority order and frames from high priority queues are sent first.
Read more about the other protocols:
What is High-availability Seamless Redundancy (HSR)?
What is Parallel Redundancy Protocol (PRP)?
What is Precision Time Protocol (PTP)?
FDS is scalable from three to eight ports and has up to 1 Gigabit speed. Thanks to its scalability, the single IP core can be easily used both in low as well as high-end devices based on FPGA technology.
FDS has been validated using sophisticated methods that guarantee the quality of the IP – for example HW accelerated simulation. The ready-made reference systems enable rapid and predictable product development cycle while still benefiting from the flexibility of FPGA technology.
Standard features include:
- From 3 to 8 triple-speed (10Mbps/100Mbps/1Gbps) ports, full-duplex operation on all ports
- Traffic policing and shaping (IEEE 802 TSN standards)
- Quality of Services (QoS) with priority tagging, packet filtering and eight priority queues per port
- IEEE 802.1CB Seamless Redundancy
- HSR, High-availability Seamless Redundancy (IEC 62439-3 Clause 5)
- Time and frequency synchronization using IEEE1588-2008 Precision Time Protocol v2 and IEEE 802.1AS
- IEEE1588v2 End-to-end one-step Transparent Clock Functionality
- IEEE1588v2 Peer-to-peer Transparent Clock support functions
- Ethernet packet forwarding at wire-speed, non-blocking
- Cut-through and store-and-forward operation
- Port-based VLAN and VLAN tagging, max number of VLANs is 4096
- All of the ports can be either copper or fiber Ethernet interfaces, or connected to other FPGA blocks
- Interface options – MII and GMII (RMII, RGMII, SGMII, 1000BASE-X and 100BASE-FX with separate adapters)
- Register interface for accessing control and status registers
For more information, please contact Flexibilis.