Next Point-of-Attachment Selection Based on Long Short Term Memory Model in Wireless Networks

Yang, Huigyu; Raza, Syed M.; Kim, Moonseong; Le, Duc-Tai; Vo, Vi Van; Choo, Hyunseung

doi:10.1109/imcom48794.2020.9001672

Cited by 6 publications

(5 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The two proposed models are implemented using Keras and CUDA libraries on the hardware environment consisting of Intel i7 CPU, 64GB RAM, and RTX 2080 GPU. The dataset is divided into training and testing data with 70:30 ratio, to provide comprehensive performance evaluation while maintaining enough data for training [8], [53]. All four models are trained for 300 epochs (110 epochs for OMD) using the training data with batch size of 100.…”

Section: Results and Analysesmentioning

confidence: 99%

“…For example, when the sequences ⟨s 1 , s 2 ⟩ and ⟨s 1 , s 2 , ..., s 5 ⟩, where s j is AP identifier at index j, are given with the target length 3, the normalization process results in removal of s 1 , s 2 and the longer sequence is converted to multiple subsequences ⟨s 1 , s 2 , s 3 ⟩, ⟨s 2 , s 3 , s 4 ⟩, ⟨s 3 , s 4 , s 5 ⟩. At the end of these three steps, we shuffle the order of created sequences to prevent biased batch that limits training performance of the models due to similar trend of subsequences from same mobile [8]. Algorithm 1 describes the path sequence manipulation process.…”

Section: B Data Collection and Preprocessingmentioning

confidence: 99%

“…The proposed SFI model uses recursive approach to predict the mobile path of target length q from a given historical path of length p. This model is an extension of our previous work on single-step prediction [8] that predicts next PoA by using stacked LSTM. As shown in Fig.…”

Section: Recursive Approach Based Modelmentioning

confidence: 99%

“…Deep Learning (DL) models learn from given time-series data and have shown high prediction accuracy in real time. In particular, Long Short-Term Memory (LSTM) based DL model achieves better performance than machine learning models [7] and it has shown 91% accuracy in next PoA prediction from time-series data for proactive mobility management [8]. Single step (i.e., next PoA) prediction is not sufficient for mobility management in dense cell deployment, as it does not provide sense of the path for effective handover decisions and optimal server selection in MEC for resource allocation.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Deep Mobile Path Prediction With Shift-and-Join and Carry-Ahead

Yang

Raza

Kim

et al. 2023

IEEE Trans. Cogn. Commun. Netw.

View full text Add to dashboard Cite

Importance of user mobility has rapidly increased in 5G due to reduced cell sizes, management of Multi-access Edge Computing (MEC), and ultra-low latency services. Reactive nature of existing management systems is a bottleneck, and it can be solved by building proactive systems that exploit temporal characteristics of time-series mobility data to predict long-term user movement (i.e., path). However, user mobile path prediction with useable accuracy is a challenging task, particularly for lengthy target trajectories. This paper adopts general approaches to propose two models for predicting mobile path with high accuracy.Step Forward Iteration (SFI) model is based on recursive approach, whereas Encoder-Decoder (ED) model follows multi-output approach, and both the models use Long-Short Term Memory (LSTM) as the learning unit. Training and testing of these models is done on mobility datasets from the wireless network of Pangyo ICT Research Center, Korea and one of the Korean mobile operators. The experiment results show viability of the proposed models for leveraging mobile network management, as they outperform state-of-the-art GRU with attention (GRU-ATTN) and Transformer Network (TN) models. The highest prediction accuracies achieved for 3, 5, and 7 steps of target sequences (i.e., predicted mobile path) in the campus dataset are 96%, 90%, and 87%, respectively.

show abstract

Section: Results and Analysesmentioning

confidence: 99%

Section: B Data Collection and Preprocessingmentioning

confidence: 99%

Section: Recursive Approach Based Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Deep Mobile Path Prediction With Shift-and-Join and Carry-Ahead

Yang

Raza

Kim

et al. 2023

IEEE Trans. Cogn. Commun. Netw.

View full text Add to dashboard Cite

show abstract

“…Deep learning (DL)/Machine Learning (ML) approaches have recently achieved promising results for mobility management of gUEs [ 20 , 21 , 22 ]. In UAVs, a policy gradient based DRL utilizes the RSSI from gUEs to address the mobility management of UAV base stations for improved data rates in 3D-space [ 23 ].…”

Section: Literature Review and Backgroundmentioning

confidence: 99%

Proactive Handover Decision for UAVs with Deep Reinforcement Learning

Jang

Raza

Kim

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

The applications of Unmanned Aerial Vehicles (UAVs) are rapidly growing in domains such as surveillance, logistics, and entertainment and require continuous connectivity with cellular networks to ensure their seamless operations. However, handover policies in current cellular networks are primarily designed for ground users, and thus are not appropriate for UAVs due to frequent fluctuations of signal strength in the air. This paper presents a novel handover decision scheme deploying Deep Reinforcement Learning (DRL) to prevent unnecessary handovers while maintaining stable connectivity. The proposed DRL framework takes the UAV state as an input for a proximal policy optimization algorithm and develops a Received Signal Strength Indicator (RSSI) based on a reward function for the online learning of UAV handover decisions. The proposed scheme is evaluated in a 3D-emulated UAV mobility environment where it reduces up to 76 and 73% of unnecessary handovers compared to greedy and Q-learning-based UAV handover decision schemes, respectively. Furthermore, this scheme ensures reliable communication with the UAV by maintaining the RSSI above −75 dBm more than 80% of the time.

show abstract