Preface 3

For background, mobile network services have increasingly migrated to using IP (Internet Protocol) as their network layer, with all 5G wireless services defined as being carried over IP packets. Advantages to providing services using IP-based networks include increased flexibility and the potential bandwidth efficiency gaining from statistical packet multiplexing. MPLS-TP (Multiprotocol Label Switching-Transport Profile) served as the best packet routing protocol for IP-based services within a telecom provider network. MPLS packets are carried over IEEE 802.3 Ethernet for their data link and physical layers. As the bandwidth of mobile services increased and the wireline backbone networks grew, the use of packet routing technology became more challenging. MPLS routing requires terminating the Ethernet layer so that MPLS packet overhead and routing tables can be used to make the forwarding decisions for each packet.

The processing associated with MPLS per-packet forwarding adds latency and requires a significant amount of power consumption relative to circuit-switched time division multiplexing (TDM) technology. This becomes increasingly important as node bandwidth capacity increases. A solution to this issue is to use hybrid approach where the nodes can use a combination of packet and TDM switching. With TDM, user packet flows that share a common set of ingress and egress nodes in the wireline network which can be carried within the same TDM channel. This allows intermediate nodes to use a TDM switch to bypass their packet switch. Consequently, since entire streams are switched/forwarded based on their TDM channel, per-packet processing is unnecessary. An additional benefit of TDM is that client traffic in one TDM channel cannot impact the performance of traffic in a different TDM channel. This feature is sometimes referred to as “hard isolation” to distinguish it from “soft isolation” in which IEEE 802.1 extensions to Ethernet are used to emulate this isolation feature within an Ethernet packet switched network.

Since the MPLS-TP networks use Ethernet for their lower layers, there was incentive to define a new TDM technology built on Ethernet technology rather than use an existing TDM technology with its own separate network management technology. China Mobile Communications Corporation (CMCC) led the way in pioneering such a technology. The umbrella name for this early development was Slicing Packet Network (SPN), where the “slicing” referred to the TDM channels. CMCC, along with several of their equipment vendors, submitted SPN to the ITU-T for standardization. The primary ITU-T group responsible for this type of technology is Question 11 of Study Group 15 (Q11/15), which defines TDM digital signal formats and rates.

I have been actively involved in telecommunications standards development since 1984. During that time, I have never seen a major proposal that was not significantly improved by the process of considering and incorporating input from participants. The Q11/15 work SPN followed this pattern. Although there were a number of interesting and innovative aspects to SPN, some issues and potential areas for improvement were identified during the Q11/15 discussions. After much discussion and hard work, Q11/15 reached consensus on a compromise that created an enhanced version of SPN. While the most important core aspects of SPN were preserved, in order to avoid confusion, SG15 chose the new name “Metro Transport Network” (MTN) for the version standardized in ITU-T Recommendation G.8312. MTN now forms a new generation of transport hierarchy after SDH and OTN, to support 5G wireless networks and beyond. The various aspects of MTN are covered in the ITU-T G.83xx Recommendation family to support 5G wireless networks and beyond, including:

◆G.8310 (2020)-Architecture of the metro transport network

◆G.8312 (2021)-Interfaces for metro transport networks

◆G.8321 (2022)-Characteristics of MTN equipment functional blocks

◆G.8331 (2022)-Metro transport network linear protection

◆G.8350 (2022)-Management and control for metro transport network

◆G.mtn-sync (2023)-Synchronization aspects of metro transport network

The key SPN concepts that were preserved in MTN, either directly or with enhancement, include:

◆Using TDM to provide “hard-isolation” slices of the network bandwidth rather than packet-based soft-isolation.

◆Using the Optical Internetworking Forum (OIF) FlexE Implementation Agreement as the basis for providing a TDM layer with n×5 Gbit/s channels. Each FlexE channel carries an IEEE 802.3 Clause 82 64B/66B block coded PCS stream.

◆Introducing TDM switching of the 64B/66B block streams to avoid going up to the Ethernet MAC or MPLS layer for packet switching.

◆Using IEEE 802.3 Clause 82 idle blocks insertion/removal for rate adapting the SPN client into the FlexE channel.

◆Inserting network path overhead in-band as special control blocks in the Ethernet inter-packet gap (IPG), inserting one block at a time rather than using a multi-block format within the same IPG. The control blocks are identified by a 0xC O code.

◆Defining a mix of high-priority and low-priority path overhead messages, with each OAM block containing the associated message type information.

　　◇ Basic blocks, which are sent with a regular nominal spacing, use a bit-oriented format to carry information requiring frequent repetition (e.g., path error monitoring and defect status).

　　◇ APS information is sent in separate messages/blocks.

　　◇ Other information (e.g., connectivity verification, delay measurement functions and client signal information) are each carried separately using a message-oriented format that is typically spread across multiple blocks.

◆Focus on 1+1 end-to-end linear protection for the MTN path layer.

MTN enhancements to SPN included:

◆Enhancements to FlexE to create the MTN Section layer.

　　◇ Arbitrary n×5 Gbit/s channel rates rather than the limited subset supported by FlexE.

　　◇ Defining LLDP messages to be carried within the FlexE overhead channel in order to provide MTN Section-specific overhead.

◆Modifying the SPN Ordered-set-like block used for MTN Path overhead so that it fully conformed to the IEEE 802.3 Clause 82 format .

　　◇ Allowing requesting IEEE 802.3 to add a note reserving the associated MTN O code.

　　◇ Ensuring transparency of the overhead block to existing Ethernet implementations.

◆Supporting the insertion of Path OAM ordered set blocks according to the full IEEE 802.3 Clause 82 rate adaptation rules rather than only allowing OAM insertion by replacing idle blocks.

◆Deterministic overhead performance, including overhead latency, which also reduced receiver complexity and buffer requirements.

◆Some message coding enhancements to further improve robustness.

One key aspect of MTN is that it is defined to be transparent to existing intermediate SPN nodes. This allows network transition from SPN to MTN by replacing edge nodes without touching intermediate nodes. This topic is documented in ITU-T Supplement G.Sup.69 “Migration of a pre-standard network to a metro transport network”. Network migration is further simplified if the new edge nodes can interoperate with either SPN or MTN nodes on the other side of the connection.

Dr. Han Li is the main author of this book, and I have appreciated and enjoyed working with him during his many years of active involvement in ITU-T SG15. He is one of the main contributors to MTN Recommendations and one of the key people who helped to reach consensus for MTN. Based on his profound knowledge, this book provides sound descriptions of MTN technologies and also introduces many interesting background stories during the recommendation discussion. I believe readers will learn a lot from the book and enjoy the book!

Steve Gorshe, Ph.D

IEEE Life Fellow，Rapporteur for ITU-T Q11/15