Goal-oriented (GO) semantic communication facilitates scaling modern networks with growing real-time traffic generated within networked Cyber Physical Systems (CPSs). Network resource management in GO communication prioritizes data effectiveness for the application goal. This implies reducing network resources allocated to low-priority information. Existing GO approaches often lack generalization, because they tailor particular network schemes to particular applications. In the current work, we propose a practical GO scheme operating in the transport layer (TL) middleware, i.e., not requiring specific hardware or network structure. Using Reinforcement Learning (RL), the proposed GO RL TL captures the potential contribution of the currently sampled observed state to the real-time CPS process evolution at the remote monitor. Together with the network congestion level, the state’s effect on the application goal determines whether the distributed sensors deploying GO RL TL agents accept corresponding packets into the network or discard them. The offline environment for training uses the real data traces of traffic patterns and application dynamics. The model generalizes to arbitrary network and application setups present in traces by learning the corresponding inter-dependencies from data. The extensive hardware tests witness the adaptability of the proposed GO RL TL, as well as its superiority in application performance compared to competitors. GO RL TL improves remote estimation mean-squared error by $20 \%$ to $100 \%$ in static network conditions, and by $\sim 30 \%$ in the dynamic setup.

Evolving communication technologies and standards enable novel IoT use cases, such as real-time edge or fog assistance for distributed sensing and control in smart environments. Synchronizing the system state at centralized management entities with distributed components is a network resource-hungry task. Semantic communication is envisioned to cope with rapid network traffic increase and associated performance degradation by optimizing resource utilization for the application goal. Identification (ID) codes are a novel semantic technology that reduces network traffic for remote synchronization by representing a potentially multidimensional system state with a short tag. This work exploits ID codes in the intelligent transportation context. The vehicles in a truck platoon use ID codes to synchronize their distributed control decisions with the centralized edge server. We present one of the first frameworks demonstrating the advantages of using ID codes in practical scenarios. We show that the truck platoon management utilizing ID codes is$\sim 40 {\%}$more efficient than a fully distributed control scenario and never lags behind a fully centralized control. Moreover, ID codes reduce network traffic by$\sim 4$times compared to the latter.

Networked control systems (NCSs) are an example of task-oriented communication systems, where the purpose of communication is real-time control of processes over a network. In the context of NCSs, with the processes sending their state measurements to the remote controllers, the deterioration of control performance due to the network congestion can be partly mitigated by shaping the traffic injected into the network at the transport layer (TL). In this work, we conduct an extensive performance evaluation of selected TL protocols and show that existing approaches from communication and control theories fail to deliver sufficient control performance in realistic network scenarios. Moreover, we propose a new semantic-aware TL policy, which uses the process state information to filter the most relevant updates and the network state information to prevent delays due to network congestion. The proposed mechanism is shown to outperform all the considered TL protocols with respect to control performance.

Goal-oriented communication is a promising approach to tailor the network resource management algorithms to the needs of particular applications, thus enhancing the efficiency of resource utilization and boosting the application performance. In the context of distributed cyber-physical systems and networked control systems, the design of a control-aware transport layer (TL) represents a realistic approach for goaloriented communications since it can be integrated into generic control setups without making assumptions on particular hardware or network technologies and deliver enhanced end-to-end performance. This demo showcases the application performance of different TL schemes used for communication between the sensors and the controllers monitoring and actuating inverted pendulums, i.e., multi-dimensional plants. The nodes of the control loops are realized with Zolertia Re-Mote devices, and multiple control loops communicate over the shared wireless network using IEEE 802.15.4 standard. We use the demonstration testbed to compare the performance of conventional, state-of-the-art, and novel goal-oriented TL schemes by observing the emulated dynamics of inverted pendulums.

In Wireless Networked Control Systems (WNCSs), the feedback control loops are closed over a wireless communication network. The proliferation of WNCSs requires efficient network resource management mechanisms since the control performance is significantly affected by the impairments caused by network limitations. In conventional communication networks, the amount of transmitted data is one of the key performance indicators. In contrast, in WNCSs, the efficiency of the network is measured by its ability to facilitate control applications, and the data transmission rate should be limited to avoid network congestion. In this work, we consider an experimental setup where multiple control loops share a wireless communication network. Our testbed comprises up to five control loops that include Zolertia Re-Mote devices implementing IEEE 802.15.4 standard. We propose a novel relevance- and network-aware transport layer (TL) scheme for WNCSs. The proposed scheme admits the most important measurements for the control process into the network while considering current network conditions. Moreover, we propose a mechanism for the scheme parameters adaptation in dynamic scenarios with unknown network statistics. Unlike the conventional TL mechanisms failing to provide adequate control performance due to either congestion in the network or inefficient utilization of available resources, our method prevents network congestion while keeping the control performance high. We argue that relevance- and network-awareness are critical components of network protocol design to avoid control performance degradation in practice.

Software-Defined Radio Access Networks (SD-RANs) foster the concepts of programmability and flexibility, which are vital for next generation cellular networks. However, SD-RANs render network management and orchestration very challenging. Indeed, related works indicate that when thousands of connected devices are spread across the underlying network, SD-RAN approaches with a single controller become deficient and exhibit undesired behavior. Despite this, state-of-the-art research papers lack concrete solutions and evaluations with respect to throughput predictability, where the latter is jeopardized by irregularities in the SD-RAN control plane, specifically in realistic testbeds. In order to overcome the aforementioned issues, in this work, we presentDelphi: a novel platform that provides both analytical and experimental methods to achieve our goal, which is computing the maximum achievable throughput in SD-RAN environments. Analyzing the results provided byDelphi, we can capture the impact of the SD-RAN control plane on throughput. Moreover, we can design important guidelines as to which policy to choose given objectives such as throughput maximization or robustness. Providing a platform for SD-RAN evaluations based on open-source components,Delphienables new avenues for research in the mobile network community. Focusing on FlexRAN SD-RAN controller for our initial results, overall, our findings show that when the number of Base Stations (BSs) and User Equipment (UEs) in the network increases beyond 5000, due to non-timely received control packets for the maximum Channel Quality Indicator (maxCQI) policy the overall throughput decreases by more than 20%.

Next generation radio access networks (RANs) en-vision softwarization and programmability as the main tools to provide the quality of service (QoS) requirements of emerging applications. Consequently, software-defined radio access networks (SD-RANs) have gained increased traction as a technology to foster network management and alleviate orchestration. While there exist SD-RAN architecture concepts both with single and multiple SD-RAN controllers, currently developed prototypes only include a single controller. Such a design may be sufficient for a low number of managed devices, for instance below 50. When the number of devices increases beyond 300, the controller performance deteriorates. A distributed control plane provides a solution, but renders the management in the control plane complex and incurs additional overhead, for instance control handover. In this way, both single controller and distributed control plane approaches may have a negative impact on a user’s QoS. Yet, proper evaluations are missing and therefore the performance remains unclear. In order to investigate the effect of SD-RAN control plane on the user performance, in this work, we provide an extensive evaluation based on a 5G simulator, compliant with 3GPP standardization, as well as measurements with open-source SD-RAN controllers. Based on our simulator, we are able to demystify the user QoS depending on the control plane design choices. Our results demonstrate that having a distributed control plane with control handovers improves the user performance by at least 20% in terms of throughput, 5x regarding the packet loss ratio and 140% in terms of delay compared to a single controller approach. This confirms that the benefits of multiple controllers surpass the overhead caused by more complicated management.

The volume of traffic transmitted over cellular networks is growing significantly every year. To improve LTE network throughput, 3GPP has proposed License-Assisted Access (LTE-LAA), which allows utilizing the unlicensed 5GHz spectrum already used by Wi-Fi and other technologies. To enable fair resource sharing, LTE-LAA adopts CSMA/CA-based channel access with binary exponential backoff similar to Wi-Fi, However, because of the LTE-LAA technology constraints, the LTE-LAA base station may start data transmission only at rather rare licensed carrier slot boundaries. The essential question, what the base station should do between the end of the backoff procedure and the closest licensed slot boundary, is left implementation dependent. In the paper, we have developed analytical models to study the performance of different LTE-LAA implementations and analyze channel sharing fairness between LTE-LAA and Wi-Fi networks in the coexistence scenario. Moreover, we investigate how the performance of LTE-LAA channel access can be improved using New Radio flexible numerology, which is one of the main features of currently developed 5G systems.

Growing requirements for cellular network capacity made 3GPP consortium introduce License-Assisted Access (LAA), which implies unlicensed spectrum exploitation. Since these frequencies are already occupied by other technologies, such as Wi-Fi, it is vital to assure fair and efficient sharing of the unlicensed bands by LAA and Wi-Fi devices. Hence, LAA uses Listen Before Talk, the principle of which is similar to CSMA/CA implemented in Wi-Fi. Specifically, similar to Wi-Fi, an LAA base station performs the backoff procedure before every transmission attempt. However, the base station can only start LAA data transmission at licensed spectrum slot boundaries, which results in an extremely long vulnerable period from the end of backoff procedure till the slot boundary. During such a vulnerable period other devices, e.g., Wi-Fi stations, can occupy the channel, canceling the LAA transmission. To protect LAA transmissions, many developers propose sending a reservation signal during the whole vulnerable period, which in turn blocks Wi-Fi stations, occupying the channel with the useless signal and resulting in the channel resource sharing unfairness. To solve this problem, this paper designs a novel mechanism, which allows the LAA base station to detect collisions of Wi-Fi transmissions with the reservation signal. An analytical model for various LAA implementations is developed. Using the model, it is shown how LAA performance can be improved with the proposed mechanism without Wi-Fi performance degradation.