Awesome

User-Preference-Learning-based-Proactive-Edge-Caching-for-D2D-Assisted-Wireless-Networks

Abstract

This work investigates proactive edge caching for device-to-device (D2D) assisted wireless networks, where user equipment (UE) can be selected as caching nodes to assist content delivery for reducing the content transmission latency. Doing so, there are two challenging problems: 1) How to precisely learn the user's preference to cache the proper contents on each UE; 2) How to replace the contents cached on UEs when there are new popular contents continuously emerging. To address these, a user preference learning-based proactive edge caching (UPL-PEC) strategy is proposed in this work. In the strategy, we first propose a novel context and social-aware user preference learning method to precisely predict user's dynamic preferences by jointly exploiting the context correlation among different contents, the influence of social relationships and the time-sequential patterns of user's content requests. Specifically, we utilize the bidirectional long short term memory networks to capture the time-sequential patterns of user's content request. And, the graph convolutional networks are adopted to capture the high-order similarity representation among different contents from the constructed content graph. To learn the social influence representation, an attention mechanism is developed to generate the social influence weights to users with different social relationship. Based on the user preference learning, a learning-based proactive edge caching architecture is proposed to continuously caching the popular contents on UEs by integrating the offline caching content placement and the online caching content replacement policy. Real-world trace-based simulation results show that the proposed UPL-PEC strategy outperforms the compared existing caching strategies at about 2.5%-45.3% in terms of the average content transmission latency.

Requirements

stellargraph == 1.2.1
tensorflow-gpu == 2.1.0
pandas = 1.3.4
numpy == 1.19.5
matplotlib == 3.5.0

Dataset

We uploaded the processed dataset to：https://pan.baidu.com/s/1i9R6PJDiEhxxhTgAr8fdLQ Extraction Code：g97g

Based on the reviewers' comments, we revised the number of users to 100 in the final version of the paper. The update dataset can be seen at:https://pan.baidu.com/s/11KETpeBprwuvOoczMMgXzQ Extraction Code：s2i8

Training

python CS-GCN-LSTM.py #To train the model in single user scenario

python CS-GCN-LSTM_AllUsers.py #To train the model in multiple user scenarios

Predict

python CS-GCN-LSTM_predict.py #To predict the results in single user scenario

python CS-GCN-LSTM_AllUsers_predict.py #To predict the results in multiple user scenarios

Results

The convergence of the proposed CS-UPL method is investigated, and comparisons are made with UPL, C-UPL and S-UPL in single user and multiple users scenarios. The results are shown in Fig.\ref{convergence}, from which it is seen that the training loss of four methods all gradually converges as the epoch increases. But, the proposed CS-UPL method outperforms UPL, C-UPL and S-UPL methods on the convergence speed and prediction accuracy for all the single user and multiple users scenarios. This proves that the user preference learning method jointly considering the context correlation among different contents and the influence of social relationships can have a better prediction performance.

Pleasae cite the work if you would like to use it

[1] Dongyang Li, Haixia Zhang, Hui Ding, Tiantian Li, Daojun Liang and Dongfeng Yuan, "User Preference Learning-based Proactive Edge Caching for D2D-Assisted Wireless Networks," in IEEE Internet of Things Journal, early access. DOI:10.1109/JIOT.2023.3244621.

[2] Dongyang Li, Haixia Zhang, Tiantian Li, Hui Ding and Dongfeng Yuan, "Community Detection and Attention-Weighted Federated Learning Based Proactive Edge Caching for D2D-Assisted Wireless Networks," in IEEE Transactions on Wireless Communications,early access. DOI:10.1109/TWC.2023.3249756.

[3] Dongyang Li, Haixia Zhang, Dongfeng Yuan and Minggao Zhang, "Learning-Based Hierarchical Edge Caching for Cloud-Aided Heterogeneous Networks," in IEEE Transactions on Wireless Communications,early access. DOI:10.1109/TWC.2022.3206236.