Dynamic Packet Duplication for Industrial URLLC
Abstract
:1. Introduction
- Enhanced Mobile BroadBand (eMBB): this service category is an evolution of traditional mobile broadband, with higher data rates (up to 20 Gbps) and bandwidth. It is similar to the traditional use of networks by users, such as web browsing or streaming multimedia content.
- Massive Machine-Type Communications (mMTC): this service category covers massive connection of devices, with a sporadic and lower volume of data exchange over the network. It is mainly focused on the Internet of Things (IoT).
- Ultra-Reliable and Low Latency Communications (URLLC): this service category aims to cover critical communications, where short messages are exchanged with requirements of lower latency and higher reliability. The latency requirement varies from 1 to 15 ms, depending on the application itself. However, in 5G, it is expected to reach a maximum latency of 1 ms with a reliability target of for a packet size of 32 bytes at the user plane [2].
2. Background
2.1. Industrial Networks
2.1.1. Wireless Connectivity in Industry
2.1.2. Critical Applications in Industry 4.0
- Rearrangeable modules in production lines [23]: traditionally, production lines have been made up of static modules that perform specific operations. These modules, each controlled by a Programmable Logic Controller (PLC), are interconnected via wired to the Manufacturing Execution System (MES). By enabling the mobility of these modules, new combinations of elements into new types of production lines are possible.
- Automated Guided Vehicle (AGV) [24]: it is common that vehicles driven by workers perform tasks, such as moving stocks and supplies in factories. In the Industry 4.0 paradigm, due to the customization of production, these kinds of movements increase exponentially. It is harder to provide supplies in batches; therefore, smaller vehicles are required with an increase in the number and variety of trips. To achieve this without increasing the workload, AGVs do this without the need for human drivers.
- Drones [25]: drones are a new category of vehicle that enables novel possibilities in factories. Applications, such as emergency assistance, surveillance or rapid point-to-point delivery can be highly optimized with these vehicles.
- Autonomous robots [26]: robots have been extensively adopted in industry since commercial variants have been available. Nevertheless, early iterations of robotics technologies were limited in the number of tasks that they could perform and depended strongly on operators programming them correctly. Currently, AI and ML, along with Simultaneous Location and Mapping (SLAM) and navigation technologies, are enabling novel functionalities on robots that are much more autonomous and perform tasks that were previously reserved for workers.
- Connected workers solutions [27]: the development of consumer electronics in the last years has had a higher impact in the professional area. Gadgets, such as Augmented Reality (AR) glasses, tablets, haptic interfaces and sensors have shown a productivity boost in factories.
2.2. 5G Multi-Connectivity Overview
Packet Duplication for URLLC
3. Proposal
3.1. System Description
- KPI monitor: collects the KPIs from the radio interface at regular intervals.
- S-KPI monitor: reads the information from the URLLC device and measures the latency.
- Training data collector: joins the data generated by the KPI monitor and the S-KPI monitor. The data joined is used as input to train the ML model.
- Estimator: performs the task of estimating the S-KPI. The primary inputs are the current KPIs (such as SINR, MCS, HARQ feedback etc.) as measured by the MN. The output is the estimation of the E2E latency. This module also has a secondary input that consist in the estimation model extracted from the ML.
3.2. KPI to S-KPI Mapping
- Signal to Interference plus Noise Ratio (SINR): includes all the usable signals in the computation. It is used by some vendors to better determine the CQI to adapt the modulation.
- Modulation index: indicates the modulation index used from the table of the MCS when performing a packet transmission. A higher index selects a more efficient modulation, with a higher spectral efficiency and code rate. Otherwise, a lower modulation index selects a more robust modulation, with a lower spectral efficiency and code rate.
- Reception Success: indicates if a packet has been decoded successfully at the receiver or not. This is used to determine if a packet has suffered a HARQ retransmission, since the NACK message is indicated by the receiver to the base station.
3.3. Random Forests
- Number of decision trees: this establishes the number of trees that constitutes the forest. This must be chosen in relation to the input dataset to avoid overhead.
- Bootstrap: this parameter decides how each tree is built independently. If it is not activated, the complete dataset is used for each tree. Otherwise, the initial dataset is divided into subsets of dataset for each tree.
- Division criterion: this defines the quality of a split according to the condition set in the node. The most used criteria for regression are the squared error and absolute error.
- Maximum leaves per tree: this sets the maximum depth of the tree in the forest.
- Maximum samples to split: this determines the maximum number of samples to consider in order to choose the condition that determines the split.
- Minimum samples to split: this determines the minimum samples needed to consider a new split.
3.4. Implementation Considerations
- Implement the data collection and ML stages in the network core. The main advantage is the availability of large datasets that add diversity to the final model. Another advantage of this method is that cloud computing resources can be used better. This is even more important when looking ahead to future 6G networks, where network elements in the core network for ML and AI are envisioned.
- Implement everything in the network edge. In this case, to gain diversity, a Federated Learning (FL) mechanism can be used to share model parameters between different agents. FL is the collaborative learning, which trains and updates the model through the joint effort of multiple servers that are deployed in a decentralized manner within the network.
4. Tests
4.1. Simulation Scenario
- The packet processing from MAC to PHY layer is fixed at two slots. This is a delay between the control/data acquisition from the RLC layer by the MAC layer and the moment at which the data is available to go over the air.
- The transport block decode latency is set to 100 microseconds at UE and gNB. It is a delay between the data acquisition from the air by the PHY layer and the moment at which the data block is available to process at the MAC layer.
- The processing delay needed to decode Downlink Control Information (DCI) and decode downlink data is set to 0 slots.
- The processing delay needed from the end of downlink data reception to the earliest possible start of the corresponding ACK/NACK transmission is set to 1 slot.
4.2. KPIs Recollection
5. Results and Discussion
5.1. Prediction Results
5.2. Packet Duplication Results
6. Conclusions
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Conflicts of Interest
References
- Lasi, H.; Fettke, P.; Kemper, H.G.; Feld, T.; Hoffmann, M. Industry 4.0. Bus. Inf. Syst. Eng. 2014, 6, 239–242. [Google Scholar] [CrossRef]
- 3GPP. TR 38.913, Study on Scenarios and Requirements for Next Generation Access Technologies; V14.3.0, Rel-14. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=2996 (accessed on 28 September 2021).
- 5G Americas. New Services & Applications with 5G Ultra-Reliable Low Latency Communications; 5G Americas White Paper, November 2018. Available online: https://www.5gamericas.org/wp-content/uploads/2019/07/5G_Americas_URLLLC_White_Paper_Final__updateJW.pdf (accessed on 21 December 2021).
- Zaidi, A.A.; Baldemair, R.; Tullberg, H.; Bjorkegren, H.; Sundstrom, L.; Medbo, J.; Kilinc, C.; Da Silva, I. Waveform and Numerology to Support 5G Services and Requirements. IEEE Commun. Mag. 2016, 54, 90–98. [Google Scholar] [CrossRef]
- Segura, D.; Khatib, E.J.; Munilla, J.; Barco, R. 5G Numerologies Assessment for URLLC in Industrial Communications. Sensors 2021, 21, 2489. [Google Scholar] [CrossRef] [PubMed]
- Pedersen, K.; Pocovi, G.; Steiner, J.; Maeder, A. Agile 5G Scheduler for Improved E2E Performance and Flexibility for Different Network Implementations. IEEE Commun. Mag. 2018, 56, 210–217. [Google Scholar] [CrossRef] [Green Version]
- Jacobsen, T.; Abreu, R.; Berardinelli, G.; Pedersen, K.; Mogensen, P.; Kovacs, I.Z.; Madsen, T.K. System Level Analysis of Uplink Grant-Free Transmission for URLLC. In Proceedings of the 2017 IEEE Globecom Workshops (GC Wkshps), Singapore, 4–8 December 2017; pp. 1–6. [Google Scholar]
- Khatib, E.J.; Wassie, D.A.; Berardinelli, G.; Rodriguez, I.; Mogensen, P. Multi-Connectivity for Ultra-Reliable Communication in Industrial Scenarios. In Proceedings of the 2019 IEEE 89th Vehicular Technology Conference (VTC2019-Spring), Kuala Lumpur, Malaysia, 28 April–1 May 2019; pp. 1–6. [Google Scholar]
- Mahmood, N.H.; Lopez, M.; Laselva, D.; Pedersen, K.; Berardinelli, G. Reliability Oriented Dual Connectivity for URLLC services in 5G New Radio. In Proceedings of the 2018 15th International Symposium on Wireless Communication Systems (ISWCS), Lisbon, Portugal, 28–31 August 2018; pp. 1–6. [Google Scholar]
- Rayavarapu, S.M.; Amuru, S.D.; Kiran, K. Dynamic Control of Packet Duplication in 5G-NR Dual Connectivity Architecture. In Proceedings of the 2020 International Conference on COMmunication Systems NETworkS (COMSNETS), Bengaluru, India, 7–11 January 2020; pp. 539–542. [Google Scholar]
- Huang, Y.; Hou, Y.T.; Lou, W. A Deep-Learning-based Link Adaptation Design for eMBB/URLLC Multiplexing in 5G NR. In Proceedings of the 2021 IEEE Conference on Computer Communications (INFOCOM), Vancouver, BC, Canada, 10–13 May 2021; pp. 1–10. [Google Scholar]
- Rao, J.; Vrzic, S. Packet Duplication for URLLC in 5G: Architectural Enhancements and Performance Analysis. IEEE Netw. 2018, 32, 32–40. [Google Scholar] [CrossRef]
- Belogaev, A.; Khorov, E.; Krasilov, A.; Shmelkin, D.; Tang, S. Conservative Link Adaptation for Ultra Reliable Low Latency Communications. In Proceedings of the 2019 IEEE International Black Sea Conference on Communications and Networking (BlackSeaCom), Sochi, Russia, 3–6 August 2019; pp. 1–5. [Google Scholar]
- Khan, J.; Jacob, L. Link Adaptation for Multi-connectivity Enabled 5G URLLC: Challenges and Solutions. In Proceedings of the 2021 International Conference on COMmunication Systems and NETworkS (COMSNETS), Bangalore, India, 5–9 February 2021; pp. 148–152. [Google Scholar]
- Centenaro, M.; Laselva, D.; Steiner, J.; Pedersen, K.; Mogensen, P. Resource-Efficient Dual Connectivity for Ultra-Reliable Low-Latency Communication. In Proceedings of the 2020 IEEE 91st Vehicular Technology Conference (VTC2020-Spring), Antwerp, Belgium, 25–28 May 2020; pp. 1–5. [Google Scholar]
- Zhao, Q.; Paris, S.; Veijalainen, T.; Ali, S. Hierarchical Multi-Objective Deep Reinforcement Learning for Packet Duplication in Multi-Connectivity for URLLC. In Proceedings of the 2021 Joint European Conference on Networks and Communications 6G Summit (EuCNC/6G Summit), Porto, Portugal, 8–11 June 2021; pp. 142–147. [Google Scholar]
- Lorca, J.; Solana, B.; Barco, R.; Herrera-Garcia, A.; Palacios, D.; Fortes, S.; Demestichas, P.; Kosmatos, E.; Georgakopoulos, A.; Stavroulaki, V.; et al. Deliverable D2.1: Scenarios, KPIs, Use Cases and Baseline System Evaluation. Technical Report, E2E-Aware Optimizations and Advancements for Network Edge of 5G New Radio (ONE5G). 2017. Available online: https://one5g.eu/wp-content/uploads/2017/12/ONE5G_D2.1_finalversion.pdf (accessed on 28 September 2021).
- Aijaz, A. Packet Duplication in Dual Connectivity Enabled 5G Wireless Networks: Overview and Challenges. IEEE Commun. Stand. Mag. 2019, 3, 20–28. [Google Scholar] [CrossRef] [Green Version]
- Herrera-Garcia, A.; Fortes, S.; Baena, E.; Mendoza, J.; Baena, C.; Barco, R. Modeling of Key Quality Indicators for End-to-End Network Management: Preparing for 5G. IEEE Veh. Technol. Mag. 2019, 14, 76–84. [Google Scholar] [CrossRef]
- Hu, Y.C.; Patel, M.; Sabella, D.; Sprecher, N.; Young, V. Mobile Edge Computing—A key technology towards 5G. ETSI White Pap. 2015, 11, 1–16. [Google Scholar]
- Hasan, S.; Ben-David, Y.; Bittman, M.; Raghavan, B. The Challenges of Scaling WISPs. In Proceedings of the 2015 Annual Symposium on Computing for Development (DEV’15). Association for Computing Machinery, New York, NY, USA, 1–2 December 2015; pp. 3–11. [Google Scholar]
- Rostami, A. Private 5G Networks for Vertical Industries: Deployment and Operation Models. In Proceedings of the 2019 IEEE 2nd 5G World Forum (5GWF), Dresden, Germany, 30 September–2 October 2019; pp. 433–439. [Google Scholar]
- Weyer, S.; Schmitt, M.; Ohmer, M.; Gorecky, D. Towards Industry 4.0—Standardization as the crucial challenge for highly modular, multi-vendor production systems. IFAC-PapersOnLine 2015, 48, 579–584. [Google Scholar] [CrossRef]
- Mehami, J.; Nawi, M.; Zhong, R.Y. Smart automated guided vehicles for manufacturing in the context of Industry 4.0. Procedia Manuf. 2018, 26, 1077–1086. [Google Scholar] [CrossRef]
- Fernández-Caramés, T.M.; Blanco-Novoa, O.; Froiz-Míguez, I.; Fraga-Lamas, P. Towards an Autonomous Industry 4.0 Warehouse: A UAV and Blockchain-Based System for Inventory and Traceability Applications in Big Data-Driven Supply Chain Management. Sensors 2019, 19, 2394. [Google Scholar] [CrossRef] [PubMed] [Green Version]
- Gonzalez, A.G.; Alves, M.V.; Viana, G.S.; Carvalho, L.K.; Basilio, J.C. Supervisory control-based navigation architecture: A new framework for autonomous robots in industry 4.0 environments. IEEE Trans. Ind. Inform. 2017, 14, 1732–1743. [Google Scholar] [CrossRef]
- Paelke, V. Augmented reality in the smart factory: Supporting workers in an industry 4.0. environment. In Proceedings of the 2014 IEEE Emerging Technology and Factory Automation (ETFA), Barcelona, Spain, 16–19 September 2014; pp. 1–4. [Google Scholar]
- Mogensen, R.S.; Rodriguez, I.; Berardinelli, G.; Fink, A.; Marcker, R.; Markussen, S.; Raunholt, T.; Kolding, T.; Pocovi, G.; Barbera, S. Implementation and Trial Evaluation of a Wireless Manufacturing Execution System for Industry 4.0. In Proceedings of the IEEE 90th Vehicular Technology Conference (VTC2019-Fall), Honolulu, HI, USA, 22–25 September 2019; pp. 1–7. [Google Scholar]
- 3GPP. TS 36.300, Evolved Universal Terrestrial Radio Access (E-UTRA) and Evolved Universal Terrestrial Radio Access Network (E-UTRAN); Overall Description; Stage 2; V12.10.0, Rel-12. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=2430 (accessed on 17 September 2021).
- Rosa, C.; Pedersen, K.; Wang, H.; Michaelsen, P.-H.; Barbera, S.; Malkamäki, E.; Henttonen, T.; Sébire, B. Dual connectivity for LTE small cell evolution: Functionality and performance aspects. IEEE Commun. Mag. 2016, 54, 137–143. [Google Scholar] [CrossRef]
- 3GPP. TS 37.340, Evolved Universal Terrestrial Radio Access (E-UTRA) and NR; Multi-Connectivity; Stage 2; V16.0.0, Rel-16. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=3198 (accessed on 17 September 2021).
- Kassim, M.; Rahman, R.A.; Aziz, M.A.A.; Idris, A.; Yusof, M.I. Performance analysis of VoIP over 3G and 4G LTE network. In Proceedings of the 2017 International Conference on Electrical, Electronics and System Engineering (ICEESE), Kanazawa, Japan, 9–10 November 2017; pp. 37–41. [Google Scholar]
- Alderisi, G.; Iannizzotto, G.; Bello, L.L. Towards IEEE 802.1 Ethernet AVB for Advanced Driver Assistance Systems: A preliminary assessment. In Proceedings of the 2012 IEEE 17th International Conference on Emerging Technologies Factory Automation (ETFA 2012), Krakow, Poland, 17–21 September 2012; pp. 1–4. [Google Scholar]
- Geurts, P.; Ernst, D.; Wehenkel, L. Extremely randomized trees. Mach. Learn. 2006, 63, 3–42. [Google Scholar] [CrossRef] [Green Version]
- Breiman, L. Random Forests. Mach. Learn. 2001, 45, 5–32. [Google Scholar] [CrossRef] [Green Version]
- NS-3-A Discrete-Event Network Simulator for Internet Systems. Available online: https://www.nsnam.org/ (accessed on 20 September 2021).
- Patriciello, N.; Lagen, S.; Bojovic, B.; Giupponi, L. An E2E simulator for 5G NR networks. Simul. Model. Pract. Theory 2019, 96, 101933. [Google Scholar] [CrossRef] [Green Version]
- 3GPP. TR 38.901, Study on Channel Model for Frequencies from 0.5 to 100 GHz; V16.1.0, Rel-16. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=3173 (accessed on 20 September 2021).
- 3GPP. TR 21.915, Release Description; Release 15; V15.0.0, Rel-15. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=3389 (accessed on 22 December 2021).
- 3GPP. TS 38.214, NR; Physical Layer Procedures for Data, Release 16; V16.7.0, Rel-16. Available online: https://portal.3gpp.org/desktopmodules/Specifications/SpecificationDetails.aspx?specificationId=3216 (accessed on 22 December 2021).
Parameter | Value |
---|---|
Channel and propagation loss model | 3GPP 38.901 |
System bandwidth | 20 MHz |
Center frequency | 3.7 GHz |
Numerology | 2 |
Scenario | InF-DH |
Transmission direction | Downlink |
Modulation | Adaptive |
Scheduler | Round-Robin |
UE height | 1.5 m |
gNB height | 8 m |
Transmission power | 23 dBm |
Xn interface delay | 100 μs |
MAC to PHY delay | 2 slots |
Transport block decode latency | 100 μs |
HARQ feedback delay | 1 slot |
HARQ retranmission attempts | 1 |
Packet size | 64 bytes |
Packet interval | 10 ms |
S-KPI | False Positive Rate | False Negative Rate | Success Rate |
---|---|---|---|
Latency | 0.0041% | 0.0615% | 99.9849% |
Technique | Latency below Threshold Rate |
---|---|
Single connection | 81.6549% |
Always PD | 95.7891% |
PD via Random Forest | 95.7541% |
PD Technique | Number of Packets Duplicated | Latency below Threshold Rate | Average (Packet) Latency Reduction Rate | PD Reduction |
---|---|---|---|---|
Always PD PD via | 59,940 | 95.7891% | 25.0917% | Not applicable |
Random Forest | 11,376 | 95.7541% | 86.5506% | 81.0211% |
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Segura, D.; Khatib, E.J.; Barco, R. Dynamic Packet Duplication for Industrial URLLC. Sensors 2022, 22, 587. https://doi.org/10.3390/s22020587
Segura D, Khatib EJ, Barco R. Dynamic Packet Duplication for Industrial URLLC. Sensors. 2022; 22(2):587. https://doi.org/10.3390/s22020587
Chicago/Turabian StyleSegura, David, Emil J. Khatib, and Raquel Barco. 2022. "Dynamic Packet Duplication for Industrial URLLC" Sensors 22, no. 2: 587. https://doi.org/10.3390/s22020587