Deep Learning Approaches for Mobile Trajectory Prediction

Filippas, Yannis; Margaris, Aristotelis; Tsagkaris, Kostas

doi:10.1109/gcwkshps52748.2021.9682164

Cited by 5 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recent studies have focused on utilizing DL-based multi-output models for user mobile path prediction, such as encoder-decoder model that is built upon Seq2Seq framework [15], [34] and multimodal recurrent framework with multi-output approach [43]. A recent study utilizes encoder-decoder model architecture with Gated Recurrent Unit (GRU) to predict mobile trajectories [44]. It enhances the prediction accuracy by implementing the attention block in the encoder at the cost of elevated computation.…”

Section: Multi-output Approach For Multi-step Predictionmentioning

confidence: 99%

“…For example, AP ID 3 is converted into vector [0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0] and 5 into vector [0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]. This representation of elements is adapted for the comparison models, GRU-Attention (GRU-ATTN) model [44], and Transformer Network (TN) model [45].…”

Section: B Data Collection and Preprocessingmentioning

confidence: 99%

“…To this end, outright accuracy results of three, five, and seven steps ahead predictions for SFI model and ED model are depicted in Fig. 8 as a function of input sequence length, and compared against two stateof-the-art models GRU-ATTN [44], and TN [45]. For the comprehensive analyses, the results of CMD and OMD are separately illustrated in Figs.…”

Section: Epochmentioning

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

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: Multi-output Approach For Multi-step Predictionmentioning

confidence: 99%

Section: B Data Collection and Preprocessingmentioning

confidence: 99%

Section: Epochmentioning

confidence: 99%

See 1 more Smart Citation

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

“…The number of times that the learning algorithm will work through the entire training dataset results of three, five, and seven steps ahead predictions for SFI model and ED model are depicted in Fig. 8 as a function of input sequence length, and compared against two stateof-the-art models GRU-ATTN [44], and TN [45]. For the comprehensive analyses, the results of CMD and OMD are separately illustrated in Figs.…”

Section: Epochmentioning

confidence: 99%

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

YANG¹,

Raza²,

Choo³

et al. 2022

Preprint

View full text Add to dashboard Cite

<p>Mobile user path prediction is a prerequisite for mobility-aware efficient task offloading in Multi-access Edge Computing (MEC), ultra-low latency services, and many other resource management operations. This manuscript addresses the mobile user path prediction by proposing two Deep Learning (DL) driven models that use Long Short-Term Memory (LSTM) to process time-series network data. The training and evaluation of the proposed models are done using our collected and preprocessed dataset which has been made a publicly learnable resource. </p>

show abstract