Embedded Files
One of the most challenging tasks that autonomous network control systems need to face is to provision end-to-end packet flows spanning several domains. With the advent of time sensitive networking, the co-existence of services and infrastructure with and without time sensitive capabilities adds more complexity to the end-to-end flow management. In particular, avoiding unproper provisioning decisions is key to guarantee stringent quality of service requirements. In this demo, we will showcase an integrated control plane for provisioning flows spanning through domains with and without TSN capabilities. The decision making process will be supported by a digital twin in charge of estimating expected performance indicators before path provisioning acceptance.
Provisioning Time-Sensitive (TS) flows on dedicated infrastructures requires just calculating the time-slot dedicated to that service, which is then allocated in the superframes used by the schedulers running in the different nodes in the network. In such scenario, there is no impact on the flows already being served by the network as every TS flow has dedicated resources supporting it. However, in heterogeneous scenarios supporting TS and non-TS flows, the impact on existing non-TS flows can be very severe, and needs to be analyzed. In this paper, we overview a provisioning process that includes the use of a Network Flow Scheduler for reserving time slots for TS flows and a Digital Twin to accurately estimate the performance of the new and already deployed non-TS flows.
Time-Sensitive Networking (TSN) standards provide scheduling and traffic shaping mechanisms to ensure the coexistence of Time-Sensitive (TS) and non-TS traffic classes on the same network infrastructure. Nonetheless, much effort is still needed on the operation of such TSN capable network infrastructure to ensure that the required performance of the different flows, defined in terms of key performance indicators, can be met once the flows are deployed in the network. In this paper, we focus on such aspects and propose a solution involving network-wide scheduling for TS flows, as well as performance estimation for non-TS flows. Specifically, a control plane architecture especially designed for provisioning TS and non-TS flows is proposed. The architecture integrates: i) a TS Flow Scheduler Planner for defining the scheduling of requested TS flows along a path so as to meet their required performance; and ii) a Network Digital Twin to estimate the performance of requested and already established non-TS flows. Differently from standardized time-aware schedulers, per-TS flow queues are assumed so as to guarantee minimal jitter. Efficient algorithms are proposed so the provisioning of flows can be carried out with high accuracy and short time. Simulation results for heterogeneous scenarios demonstrate the feasibility and efficiency of the proposed control plane architecture, as well as point out the limitations of current time-synchronization mechanisms when high-speed interfaces are considered.
This paper presents a novel radio frame design for wireless links that allows efficient scheduling of multiple time-sensitive flows with bounded delay and arbitrarily small packet jitter requirements. The proposed design is compatible with the IEEE 802.1 standards for Time Sensitive Networks (TSN) and permits the efficient integration of wireless links into newly arrived wired TSN infrastructure. With this aim, we have reformulated the window reservation mechanism introduced in standard 802.1Qbv to deal with the variable transmission rate of wireless links. The new radio frame is organized in a sequence of partially-overlapped windows. If the intended flow does not use its window completely, the unconsumed time can be scheduled to other flows in overlapped windows. The benefits of this approach are evaluated numerically in an illustrative industrial scenario showing substantial gain in terms of both reduced control cycle (flows’ periodicity) and enhanced transmission throughput (number of admitted flows).
6G systems will need to make faster and more reliable decisions at the network edge to surpass the specifications outlined for eXtreme URLLC (xURLLC) services. One fundamental requirement concerning connectivity is determinism. Currently, a solution for incorporating determinism in Layer 2 is enabled by TSN, while in Layer 3 it is targeted by DetNet. Moreover, one possible option for forwarding in the case of DetNet is precisely the utilization of TSN as a data plane. Hence, there could potentially exist a DetNet/TSN domain in real deployments. Both standards present techniques to enhance network reliability through packet replication and elimination, as FRER in the case of TSN and PREOF in the case of DetNet. This paper provides a comparison of both techniques.
Time-Sensitive Networking (TSN) standards provide scheduling and traffic shaping mechanisms to ensure the coexistence of Time-Sensitive (TS) and non-TS traffic classes on the same network infrastructure. Nonetheless, much effort is still needed on the operation of such TSN capable network infrastructure to ensure that the required performance of the different flows, defined in terms of key performance indicators, can be met once the flows are deployed in the network. In this paper, we focus on such aspects and propose a solution involving not only packet schedulers in the data plane, but also network-wide scheduling for TS flows, as well as performance estimation for non-TS flows.
We demonstrate that joint orchestration of TSN and optical network domains in support of IIoT applications reduces the TSN blocking by four orders of magnitude and the usage of high priority queues by a 28-100%.
To achieve high reliability for Wi-Fi, we propose to use network coding (NC) as a proactive inter-frame (packet-level) redundancy technique at the medium access control (MAC) layer, which has the advantage of low latency and high spectral efficiency compared to the existing retransmission and repetition techniques. In this paper, we explore practical ways to integrate NC in Wi-Fi systems and conduct over-the-air experiments in an office environment using a Wi-Fi-based platform, where NC is implemented as a software layer. The experiment results show that NC can achieve up to two orders of magnitude reliability gain over the baseline repetition scheme with the same spectral efficiency. When interferences exist in Wi-Fi transmissions, but the packet erasures are sufficiently uncorrelated, using NC can achieve very high reliability. When the correlation increases the performance gain of NC degrades, but it still maintains a significant advantage over repetition.
This paper deals with the integration of OPC UA over Wireless / Wired (Hybrid) Networks with Time Sensitive Networking (TSN) capabilities. The paper overviews the current state of middleware protocols typically built over Ethernet, namely OPC UA and DDS. Focusing on OPC UA, the paper proposes a HW/SW device architecture, based on the SHARP platform to enable the integration of OPC UA and Hybrid TSN. The integration includes both the common OPC UA Client/Server mechanism and the recently standardized OPC UA PubSub. The platform is used to build a HW testbed to assess the performance of OPC UA over Hybrid TSN. The results demonstrate that the network can provide seamless application interoperability and that can satisfy the traffic needs in a typical industrial setup. However, the paper also highlights that there is still significant research to be done to achieve deep integration between OPC UA and Hybrid TSN.
Considerable research and standardization efforts are being made to support time-sensitive traffic, e.g., generated by applications like Industry 4.0 and 5G fronthaul, on packet networks. This paper focuses on analyzing the impact of conveying time-sensitive traffic in operators’ networks when such traffic is mixed with best-effort traffic. In particular, extensions to a continuous G/G/1/k queue model are proposed to evaluate two different Ethernet technologies, synchronous and asynchronous, supporting time-sensitive flows in terms of their influence on the performance of best-effort traffic.
Accurate delay estimation is one of the enablers of future network connectivity services. If such connectivity services require isolation (slicing), such delay estimation should not be limited to a maximum value defined in the Service Level Agreement, but to a finer-grained description of the expected delay in the form of, e.g., a continuous function of the load. Obtaining accurate end-to-end (e2e) delay modeling is even more challenging in a multi-operator (Multi-AS) scenario, where the provisioning of e2e connectivity services is provided across heterogeneous multi-operator (Multi-AS or just domains) networks. In this work, we propose a collaborative environment, where each domain models intra-domain delay components of inter-domain paths and share those models with a broker system providing the e2e connectivity services. The broker, in turn, models the delay of inter-domain links based on e2e monitoring and the received intra-domain models.
Network Services automation requires predictable Quality of Service (QoS) performance, measured in terms of throughput, delay and jitter, to allow making proactive decisions. QoS is typically guaranteed by overprovisioning capacity dedicated to the packet connection, which increases costs for customers and network operators, especially when the traffic generated by the users and/or the virtual functions highly varies over the time. This paper presents the PILOT methodology for modeling the performance of packet connections during commissioning testing in terms of throughput, delay and jitter. PILOT runs in a sandbox domain and constructs a scenario where an efficient traffic flow simulation environment, based on the CURSA-SQ model, is used to generate large amounts of data for Machine Learning (ML) model training and validation. The simulation scenario is tuned using real measurements of the connection obtained from a set of active probes.
In general, the availability of an accurate machine learning (ML) model plays a particularly important role in the development of new networking solutions and is one of the main drivers for the development of 5G and beyond networking. Although an option is to update the model once inaccurate data is detected, such approach requires high computational effort, specially once the data history is large. In this paper, we propose an approach that
combines a traffic prediction model based on Long Short-Term Memory (LSTM) with an analysis module for dynamic connection capacity allocation. Once the model is generated, re-training can be triggered after inaccuracies are detected by the analysis module. Illustrative numerical results show the benefits from the proposed decision-based re-training approach to reduce the number of re-training rounds while maintaining model accuracy.
The provisioning of time sensitive end-to-end services in future 6G networks imposes multiple technical challenges, spanning from the data plane to the control and orchestration planes. In particular, the automation of the provisioning and maintenance of connectivity services with deterministic constraints over multiple technology/administrative domains requires control and orchestration solutions able to assure the strict time service requirements. In line with that, the paper investigates some requirements imposed to the control and orchestration planes and it also shows potential enabling architectures for end-to-end service guarantees.
The deployment of beyond 5G and 6G network infrastructures will enable highly dynamic services requiring stringent Quality of Service (QoS). Supporting such combinations in today's transport networks will require high flexibility and automation to operate near real-time and reduce overprovisioning. Many solutions for autonomous network operation based on Machine Learning require a global network view, and thus need to be deployed at the Software-Defined Networking (SDN) controller. In consequence, these solutions require implementing control loops, where algorithms running in the controller use telemetry measurements collected at the data plane to make decisions that need to be applied at the data plane. Such control loops fit well for provisioning and failure management purposes, but not for near real-time operation because of their long response times. In this paper, we propose a distributed approach for autonomous near-real-time flow routing with QoS assurance. Our solution brings intelligence closer to the data plane to reduce response times; it is based on the combined application of Deep Reinforcement Learning (DRL) and Multi-Agent Systems (MAS) to create a distributed collaborative network control plane. Node agents ensure QoS of traffic flows, specifically end-to-end delay, while minimizing routing costs by making distributed routing decisions. Algorithms in the centralized network controller provide the agents with the set of routes that can be used for each traffic flow and give freedom to the agents to use them during operation. Results show that the proposed solution is able to ensure end-to-end delay under the desired maximum and greatly reduce routing costs. This performance is achieved in dynamic scenarios without previous knowledge of the traffic profile or the background traffic, for single domain and multidomain networks.
Page updated
Report abuse