In multi-Unmanned Aerial Vehicles (UAVs) data collection scenarios, it remains a critical challenge to jointly optimize time-sensitive responsiveness and maximize the volume of collected data. To address this issue, a graph reinforcement learning (RL) framework termed STIG-RL is proposed. The framework integrates an attention-based spatio-temporal interactive graph (ASTIG) network to extract efficient graph representations of the UAVuser state. Additionally, a hybrid decision-making strategy that combines model-based and model-free RL is employed to enhance policy efficiency. To further improve learning performance and reduce reliance on real-world experience, a simulated trajectory generation module (SRM) with a state prediction mechanism is incorporated. Extensive experimental evaluations demonstrate that the proposed approach outperforms five baseline algorithms in terms of data collection efficiency and responsiveness.

The accurate probabilistic forecasting of ultra-short-term power generation from distributed photovoltaic (DPV) systems is of great significance for optimizing electricity markets and managing energy on the user side. Existing methods regarding cluster information sharing tend to easily trigger issues of data privacy leakage during information sharing, or they suffer from insufficient information sharing while protecting data privacy, leading to suboptimal forecasting performance. To address these issues, this paper proposes a privacy-preserving deep federated learning method for the probabilistic forecasting of ultra-short-term power generation from DPV systems. Firstly, a collaborative feature federated learning framework is established. For the central server, information sharing among clients is realized through the interaction of global models and features while avoiding the direct interaction of raw data to ensure the security of client data privacy. For local clients, a Transformer autoencoder is used as the forecasting model to extract local temporal features, which are combined with global features to form spatiotemporal correlation features, thereby deeply exploring the spatiotemporal correlations between different power stations and improving the accuracy of forecasting. Subsequently, a joint probability distribution model of forecasting values and errors is constructed, and the distribution patterns of errors are finely studied based on the dependencies between data to enhance the accuracy of probabilistic forecasting. Finally, the effectiveness of the proposed method was validated through real datasets.

In response to the issue of the hydropower consumption of run-of-river hydropower stations in Southwest China, the district cooling system can provide regulation capacity for hydropower utilization and suppress fluctuations caused by the uncertainty of hydropower. The innovative method is to utilize the thermal characteristics of pipelines and buildings, as well as the thermal comfort elasticity to shift the cooling and electricity loads, which helps to consume the surplus hydroelectric power generation. Taking the minimum total cost of coal consumption in thermal power units, hydropower abandonment penalty, and the carbon trading cost as the objective function, models were established for power supply balance constraints, heat transport constraints, and unit output constraints. The hybrid integer linear programming algorithm was used to achieve the low-carbon economic dispatch of the electric-cooling system. The calculation examples indicate that compared to the traditional real-time balance of cooling supply, the comprehensive consideration of thermal characteristics in a cooling system and flexible thermal comfort have a better operational performance. The carbon trading cost, coal consumption cost, and abandoned hydropower rate of a typical day was reduced by 4.25% (approximately CNY 7.55 × 104), 4.47% (approximately CNY 22.23 × 104), and 3.66%, respectively. Therefore, the electric-cooling dispatch model considering the thermal characteristics in cooling networks, building thermal inertia, and thermal comfort elasticity is more conducive to the hydropower utilization of run-of-river stations.

This paper presents a frequency offset estimation algorithm based on TRS (tracking reference signal) for 5G NR system in the UE tracking mode. The algorithm makes use of the 4 TRS resources within 2 consecutive slots in 5G NR FR1 to provide a larger frequency tracking range while maintaining a non-decreasing accuracy with a little more calculation complexity.

Observed time difference of arrival (OTDOA) positioning method of 5G selects the time of arrival (TOA) due to the serving base station signal as the reference, and subtracts the TOAs of other observation base stations from the reference to obtain the reference signal time difference (RSTD), each RSTD value corresponds to a hyperbola equation, on which the user equipment (UE) is assumed to be located. The performance of TOA measurement directly affects the OTDOA positioning accuracy. In wireless channel, the optimal accuracy of TOA is obtained under line of sight (LOS) condition, where the signal propagates directly from the source to the receiver. However, for 5G UE in urban areas, the multipath of non-Los caused by obstacles can cause deep signal fading, greatly reducing measurement accuracy. TOA of the first path in multipath environment needs to be accurately measured as it is closest to the actual TOA. To address this issue, the classic correlation algorithm for the first path detection has been proposed as an effective method. This paper is based on correlation algorithm and combines it with convolutional neural networks (CNN) to further improve the TOA measurement accuracy under low signal-to-noise ratio (SNR). Simulation results show that this proposed method effectively improve the OTDOA positioning accuracy of 5G UE in multipath channel.

This paper presents a resource optimization method for the 2-step random access (RA) procedure in the 5 th generation new radio (5G NR). In 2-step RA, the MsgA includes a sequence as a preamble on the physical random access channel (PRACH) and uplink payload on the physical uplink shared channel (PUSCH). The mapping between PRACH slots and PUSCH slots is specified by system configuration parameters. The network reserves the corresponding PUSCH resource fixedly, regardless of the random access occasion (RO) occupancy rate and the uplink data traffic request. It deprives the uplink scheduling resource efficiency in cases of some RO unoccupied. To achieve higher utilization efficiency, this paper studies the reusing of the PUSCH slots and occasions fixed to RACH in MsgA. By detecting the preambles in RO before the corresponding PUSCH slots, the network acquires the occupancy of the RO. Then the PUSCH occasion (PO) in reserved PUSCH can be released, which conforms to the RO detection result. The release of the previously reserved PUSCH occasions can be rescheduled and reused by the user equipment (UE) for uplink data transmission. Hence more wireless resources can be scheduled in contrast to the existing allocation of uplink share channel in 2-step type random access. It suggests that using the method improves the efficiency of resource utilization.