The information is prepared on the basis of data from the information-analytical system RSF, informative part is represented in the author's edition. All rights belong to the authors, the use or reprinting of materials is permitted only with the prior consent of the authors.

Project titleQueueing systems and admission control algorithms for Ultra-Reliable and Low Latency Communications traffic within 5G+ Industrial Deployments

AffiliationPeoples’ Friendship University of Russia named after Patrice Lumumba,

Research area 09 - ENGINEERING SCIENCES, 09-601 - Theory, development methods and operational efficiency of technical systems

Keywords5G+, New Radio (NR), mobile networks, Industrial Deployment, Ultra-Reliable and Low Latency Communications (URLLC), Enhanced Mobile Broadband (eMBB), resource sharing, Network Slicing (NS), priority scheduling, queueing systems, Quality of Service (QoS), performance measure, computational algorithm

Annotation
The innovations seen nowadays in Industrial IoT, aka industrial Internet of Things (IoT), are the consequence of the ongoing fourth industrial revolution, namely Industry 4.0. In fifth generation (5G) wireless networks, IoT use cases include smart transportation systems, smart grids, healthcare, smart metering, public safety surveillance, industrial automation, and remote manufacturing. With characteristically extremely high quality of service (QoS) requirements on wireless access points, industrial automation is one of the most important use cases. Reconfiguring the industry’s vertical processes around automation is indeed a mandatory task to enable Industry 4.0 digital transformation and, thus, make daily operations more productive, efficient, and faster. This inextricably involves low-cost and low-energy applications while maximizing the reliability of available resources. In the end, this will obviously lead to far greater competitiveness among the industry verticals. 5G New Radio (5G NR) technology offers users very new industrial automation applications such as mobile robot co-management, time synchronization, positioning, augmented reality services for personnel, and telepresence-based service. Systems controlling moving parts of production equipment typically generate low-speed traffic but require ultra-reliable low-latency communication (URLLC). At the same time, enhanced mobile broadband (eMBB) is required for video surveillance or mobile robots. Thus, 5G NR base station (BS) must simultaneously support eMBB and URLLC services, which have fundamentally different quality of service (QoS) requirements. Mechanisms to exclusively support eMBB or URLLC services on NR BSs in the millimeter-wave (mmWave) are extensively being researched. However, due to the different QoS requirements of eMBB and URLLC services, i.e., especially the 1ms data transmission delay limit for URLLC services, these cannot be implemented simultaneously within the 5G networks’ framework. It is expected that methods allowing the coexistence of traffic with fundamentally different service requirements will be proposed through the standardization process of 5G+ (5G-Advanced) networks in the coming five years. Note that, implementing simultaneous support of eMBB and URLLC services that generate radically different traffic types in 5G wireless networks is a complex and urgent problem that requires the development of new approaches, algorithms, and models for servicing such traffic. The proposed project aims at developing models and algorithms for latency-critical and reliable traffic service in industrial automation scenarios on 5G+ NR’s wireless systems basis. Note that, within these systems, it is necessary to simultaneously service eMBB and URLLC traffic while satisfying the strict data rate latency and reliability requirements for URLLC services. Moreover, research on this thematic highlighting the advantages and drawbacks of data transmission through 5G+ NR BSs using direct device-to-device (D2D) technology is relevant. On top of that, proposing possible combinations or simultaneous support of these traffic services will be remarkable. Furthermore, developing and analyzing the most effective resource-sharing scenarios to meet the various heterogeneous traffic QoS requirements will be quite interesting. In that logic, one will consider using network slicing (NS) technology and propose scheduling algorithms based on priority levels, service preemption, data rate reduction, and resource reservation. Notice that, there is almost no research carried out on that thematic, not to talk of the networks’ deployment impact on vertical’s industry processes within the NS technology framework. More particularly, research on service’s preemption “intensity” as a result of (i) signal’s propagation paths dynamic blocking, and (ii) terminal devices mobility is practically inexistent. Note that the principal reason behind all this is the sharp movements of devices carrying the various sensors. Moreover, there is no clear information about (i) the resource sharing methods at the wireless access points, i.e., priority scheduling, resource reservation, etc., and (ii) the actual service mechanism characteristics, i.e., data transmission over non-orthogonal multiple access (NOMA) protocols, direct D2D communication, etc., allowing to lower traffic load at the BS, and thus, to improve URLLC traffic service efficiency. On the other hand, questions remain about traffic isolation within NS framework, and they must be answered. In connection with that, the project team faces the task of conducting research with the purpose of achieving 5G+ NR access’ scheme models development and analysis. Through that research, the project team expects to propose and build algorithms for traffic coexistence in the context of industrial automation deployment. In that regard, the project team will rely on (i) the mathematical apparatus of queueing and teletraffic theory, and (ii) the stochastic geometry, both having proved themselves quite useful at international and Russian levels.

Expected results
The project aims at developing models and algorithms for latency-critical and reliable traffic service in industrial automation scenarios based on 5G+ NR systems. To implement industrial automation scenarios in 5G+ NR wireless networks, it is necessary to develop new data transmission strategies and propose efficient algorithms to simultaneously service traffic generated by (i) systems controlling moving parts of industrial equipment, i.e., URLLC traffic; and (ii) mobile robots or video surveillance systems, i.e., eMBB traffic. Expected results: (I) Models based on signal’s propagation paths dynamic blocking and terminal devices mobility for 5G+ NR systems’ deployment considering environmental conditions and production’s machines characteristics. These models will be developed considering the specifics of light and heavy industries. More precisely, will be built (i) a “transparency” model for automated machines; (ii) a model for signal’s propagation paths dynamic blocking with consideration for the automated machines mobility characteristics; and (iii) a model for mobile devices carried by automated machines in 5G+ NR industrial deployments. An abstract model for the 5G+ NR communication channel will be developed considering (i) the location of machines in industrial deployments and (ii) the communication’s preemption process due to device blockage and mobility constraints. (II) Base model for simultaneous support of URLLC and eMBB traffic service. An analysis will be provided on existing scheme models for accessing radio resources of wireless networks with an accent on possible simultaneous support of different traffic service types. Performance measures for simultaneous support of URLLC and eMBB traffic service will be proposed. Upon this, a base model will be developed and an analysis of the simultaneous’ support of URLLC and eMBB traffic service characteristics at 5G+ NR BSs will be provided considering potential communication preemption due to device blockage and mobility constraints. (III) URLLC and eMBB traffic service models at the data link layer of 5G+ NR technology An analysis will be provided on methods for accessing 5G+ NR resources with an accent on resource sharing, priority scheduling, and resource reservation algorithms. Upon this, one will develop and provide a comparative analysis of models based on URLLC and eMBB traffic service at the data link layer of 5G+ NR with consideration for priority scheduling, resource reservation, and non-orthogonal access. (IV) Models and algorithms for simultaneous support of URLLC and eMBB traffic service at 5G+ NR BSs within NS framework. Questions about traffic isolation within NS framework will be answered. Models and algorithms for simultaneous support of URLLC and eMBB traffic service at 5G+ NR BSs will be developed within NS framework with particular attention on bandwidth reservation priority scheduling methods. (V) Models for simultaneous support of URLLC and eMBB traffic service in wireless interfaces with complex servicing mechanisms at various levels. Based on the analytical review of the literature, algorithms will be developed for joint servicing of traffic with low delays and broadband access of the 5G+ NR BS, allowing us to consider the advantages of using (i) systems with non-orthogonal multiple access NOMA, (ii) mechanisms for partial adaptation of the speed of devices to available resources when servicing traffic systems, (iii) mechanisms for offloading URLLC traffic to direct connections. A shared service model for URLLC and eMBB traffic using NOMA non-orthogonal multiple access will be developed. A model of joint service of URLLC and eMBB traffic with a partial adaptation of the speed of devices to the available system resources will be developed, an algorithm for selecting devices using the eMBB service, the quality of service on which can be reduced, will be proposed. Models and algorithms for offloading URLLC traffic to direct D2D connections between devices will be developed, considering three options: full, partial, or no control by the 5G+ NR BS. The amount of signal interference and delays associated with the coordination of the terminals will be estimated. (VI) Comparative analysis of service characteristics of URLLC and eMBB traffic within the proposed methods and determination of recommendations for joint service of these types of traffic in industrial deployments of 5G+ NR networks. The result of the project will be a set of optimal scenarios for industrial deployment of 5G+ NR systems, obtained based on a comparative analysis of the developed models, which is a set of algorithms and schemes for sharing radio frequencies by systems that control moving parts of industrial equipment and mobile robots or video surveillance systems. For each of the schemes, a description of its advantages and disadvantages will be presented depending on the implementation conditions - the features of the chosen data transmission technology, as well as the radio resource management method - priority or reservation. This will make it possible to calculate and evaluate the quality indicators of simultaneous service of traffic with low delays and broadband access, which are necessary for the implementation of industrial automation systems in fifth-generation wireless networks.

Annotation of the results obtained in 2022
- A model of line-of-sight dynamic blocking from device to BS and direct communication between devices in industrial deployments of communication networks. The model is based on the methods of photogrammetry, integral and stochastic geometry. An important feature of the model is that it allows considering the specifics of objects that block the path of radio wave propagation. These include periodic actions performed by the devices, as well as the geometric positions of the devices themselves, which differ from the standard assumptions about the Poisson point process. - A admission control scheme model for traffic with different requirements to the quality of service [1]. The model is described as a QS with K incoming flows and can be parameterized according to the specifics of the wireless communication channel. Any of the K flows can match both streaming and elastic traffic. The concept of network slicing is used to implement shared service. Within the framework of this concept, several possible schemes for accessing network resources are considered, based on reservation/priority strategies with and without interruption of service. The process of eMBB and URLLC traffic coexistence is considered, i.e. case with two incoming flows (K=2) and different QoS requirements. Computational algorithms have been developed to calculate the main performance measures of the model. An analytical form of the stationary probability distribution is obtained for the admission scheme without interruption; for the admission scheme with service interruption a numerical solution of the system of equilibrium equations is carried out. A comparative analysis of the performance guarantees provided to each traffic type by multiplexing through the same NR radio interface is carried out. The results of the analysis showed that the performance trade-off is dictated by the proposed traffic load of the highest priority requests: (i) when the load is light or medium, a mixed strategy based on reservation and prioritization outperforms the full reservation mechanism; (ii) full reservation provides better traffic isolation under congestion conditions. A mixed strategy gives an advantage in terms of blocking probability for traffic with short requests, such as URLLC or mMTC. Whereas the full reservation scheme, in terms of both blocking probability and service interruption probability, works better for flexible eMBB traffic. An important feature is that the mixed strategy can improve resource utilization by up to 95%, which is 10-15% higher than the full reservation scheme, while providing isolation between traffic types under high load conditions. - Methods for implementing the slicing network concept used to develop mathematical models for the joint servicing of different traffic types are studied. An admission control scheme model has been developed [2,4,5]. Users get services with a guaranteed bit rate, i.e., streaming traffic was considered. The model is described as a QS with S incoming flows, each of which is served in one of S slices. To fulfill the key requirements for slice isolation, a strategy of priority access control to network resources with service interruption was used. A feature of the model is that the admission control scheme is based both on the implementation of shared access to a part of the available resources and on service prioritization based on interruption. To implement the interrupt mechanism, each of the S slices is assigned a priority in service. The introduction of priorities, as well as the absence of an individual zone in which requests of other types of s' are not allowed, distinguishes the proposed admission control scheme from the classical non-fully accessible scheme with ceilings in the common part of the resource. At the first stage of research, a model with three slices [2] and services that have the same requirements to QoS was developed. To calculate the stationary probability distribution of the system and key performance measures, a numerical solution of the system of equilibrium equations was carried out using the developed software. Further, a model with an arbitrary number of slices S and services that impose different requirements to QoS is developed [3,4]. An interruption vector function is obtained to calculate the number of slice s' serviced requests, which must be interrupted to admit one request in slice s. In [4], detailed examples of calculating the service interruption vector function in a network model with two and three slices are presented. According to these examples the features of the introduced priority control are shown. A comparative analysis of the model performance measures with the characteristics of admission control scheme based on the resource reservation mechanism was carried out. Namely, we analyzed the average number of serviced requests in slice s and in the system, the probabilities of direct admission and blocking, the average number of occupied resources in the system, the average number of occupied guaranteed resources in slice s. The analysis showed that the proposed admission control scheme allows increasing the network resources usage up to 15% [3], and also [4]: (i) efficiency in the range of small loads in the use of BS physical resources and slice capacities; (ii) significant efficiency gains while supporting service delivery for more number of slices. - The age of information studying has been started [5]. This metric is an extremely important parameter for the analysis of communication between devices in industrial deployments. It is an assessment of the timeliness of delivery of status updates to the receiving device from the transmitting device. Reference: 1. Ivanova Daria, Yves Adou, Ekaterina Markova, Yuliya Gaidamaka, and Konstantin Samouylov. 2023. "Mathematical Framework for Mixed Reservation- and Priority-Based Traffic Coexistence in 5G NR Systems" Mathematics 11, no. 4: 1046. 2. Adou Y., Markova E., Gaidamaka Y.V. (2022). Construction and Analysis of a Queueing Model with Service Prioritization for 5G Systems with Customizable Network Slice Instances. In: Vishnevskiy, V.M., Samouylov, K.E., Kozyrev, D.V. (eds) Distributed Computer and Communication Networks: Control, Computation, Communications. DCCN 2022. Lecture Notes in Computer Science, vol. 13766, pp. 41-53. Springer, Cham.. 3. Yves Adou, Ekaterina Markova and Yuliya Gaidamaka Modeling and Analyzing Preemption-Based Service Prioritization in 5G Networks Slicing Framework, Future Internet 2022, 14(10), 299. 4. K. Y. B. Adou, E. V. Markova, Yu. V. Gaidamaka, S. Ya. Shorgin Preemption-based prioritization scheme for network resources slicing in 5G systems // INFORMATICS AND APPLICATIONS. 2023. volume 17, issue 1, pp. 96 –106. 5. Zhbankova E.A., Markova E.V., Gaidamaka Yu.V. Analysis of the AoI indicator in the industrial deployment of IoT // XI Conference with international participation "Information and Telecommunication Technologies and Mathematical Modeling of High-Tech Systems" ITTM-2023 (2023, Moscow): Materials. – M.: RUDN University. - 2023. - pp. 68–73.

Publications

1. Adou Y, Markova E, Gaidamaka Y., Shorgin S. АНАЛИЗ СХЕМЫ ДОСТУПА С ПРЕРЫВАНИЕМ ПРИ НАРЕЗКЕ РАДИОРЕСУРСОВ СЕТИ ПЯТОГО ПОКОЛЕНИЯ Информатика и её применения, Том 17, выпуск 1, с. 96-106 (year - 2023) https://doi.org/10.14357/19922264230113

2. Adou Y., Markova E., Gaidamaka Y. Construction and Analysis of a Queueing Model with Service Prioritization for 5G Systems with Customizable Network Slice Instances Lecture Notes in Computer Science, volume 13766, pp. 41-53 (year - 2023) https://doi.org/10.1007/978-3-031-23207-7_4

3. Adou Y., Markova E., Gaidamaka Yu. Modeling and Analyzing Preemption-Based Service Prioritization in 5G Networks Slicing Framework Future Internet, Volume 14, Issue 10, №299 (year - 2022) https://doi.org/10.3390/fi14100299

4. Ivanova D., Adou Y., Markova E., Gaidamaka Yu., Samouylov K. Mathematical Framework for Mixed Reservation- and Priority-Based Traffic Coexistence in 5G NR Systems Mathematics, Volume 11, Issue 4, №1046. (year - 2023) https://doi.org/10.3390/math11041046

5. Zhbankova E., Markova E., Gaidamaka Yu. Анализ показателя AoI при промышленном развертывании IoT Информационно-телекоммуникационные технологии и математическое моделирование высокотехнологичных систем : материалы Всероссийской конференции с международным участием., c.68-73 (year - 2023)