- OTN transport provides excellent performance in traditional transport use cases, but standard OTN has disadvantages that hamper its use in networks carrying 5G, IoT, and private line traffic.
- Huawei’s OTN offering aims to future-proof OTN and promote its use in a wider set of use cases.
With operators facing limitations in flexibility, granularity, and traffic differentiation in their OTN networks, Huawei is introducing extensions to the technology – called Liquid OTN. It aims to improve OTN’s applicability to traffic types like 5G transport, IoT, private lines, and AR/VR, but also with a view toward making the networks more flexible and amenable to automation.
OTN, as a technology, has proven its worth in optical transport and features several important advantages that led a number of leading telcos to adopt it as their technology of choice in transport switching.
Namely, it provides the following:
- Hard isolation of transmitted carrier signals, preventing one service from affecting others.
- High-quality transmission and stable architecture, with zero packet loss and zero congestion.
- Deterministic latency, which can be managed, controlled, predicted, and monetized.
These characteristics were designed into OTN to make it suitable for proving telco-grade performance and reliability in transport (primarily backbone and metro) networks, but also make it well suited for carrying services that require high reliability and stable service parameters like latency, for example.
However, OTN has other features that need to be improved upon, in order to make it more suitable as a customer-reaching interface and extend it into the access networks.
- OTN has low granularity; the minimum size of the service pipeline (ODU0) is 1.25G, making it unsuitable for carrying a high number of low-bandwidth services.
- Another consequence of low service granularity is poor resource utilization; ODU0, for example, when used to carry a 100 Mbps service, utilizes only 10% of the available bandwidth.
- OTN’s multi-level encapsulation and mapping produces deterministic latency, but it may not be low enough, especially for 5G/uRLLC traffic types.
- Bandwidth adjustment processes are slow and unwieldy; services need to be torn down and then set up again with every bandwidth adjustment.
Huawei aims to remedy these shortcomings with its Liquid OTN, which features simplified architecture, much higher service granularity, simplified mapping, and increased bandwidth adjustment flexibility.
Liquid OTN Benefits and Implementation
To achieve these service characteristics, Liquid OTN introduces unified grooming via a new architectural element (the optical service unit, or OSU). The resulting architecture should, according to Huawei, be able to reduce the switching footprint and power consumption by 70% and 50%, respectively; provide hard slices in multiples of 2 Mbps; offer 70% lower single-site latency; and allow for bandwidth adjustment with ‘near zero’ impairment.
However, to gain the new features, operators will need to re-architect their networks – though not all the network elements will need to be changed at the same time. In the first phase, Huawei envisions coexistence of Liquid OTN and OTN in the same networks, by using extended cross-connect boards to connect Liquid OTN tributaries via the existing OTN network. In the second phase, it proposes wholly new Liquid OTN networks to replace the current OTN.
Huawei’s Liquid OTN architecture is a needed innovation in optical transport, and it can increase the attractiveness of OTN outside of its traditional backbone and transport use cases, by extending high-quality optical connections to the network edge. However, introducing Liquid OTN requires new network investment and assumes operators’ willingness to expand OTN services, instead of migrating to purely packet-base service offerings or deploying an existing SDH migration solution. It also is a proprietary technology at present, although Huawei (along with Chinese Tier-1 operators) is likely to propose some, or all, of the Liquid OTN characteristics in future OTN standards.