Revenue Optimization in Video Caching Networks with Privacy-Preserving Demand Predictions

2025-05-09

AI Summary

In edge video caching networks, users, ISPs, and content service providers (CSPs) must collaboratively forecast demand across multiple time slots but are unwilling to share sensitive historical data due to privacy concerns. Method: This paper proposes a dynamic caching framework that jointly achieves privacy-preserving multi-slot demand forecasting and long-term revenue optimization. It innovatively integrates differentially private federated learning with a Transformer-based temporal model, explicitly modeling how prediction errors affect file popularity. Instead of optimizing cache hit rate, the framework maximizes long-term operator revenue and employs a greedy approximation algorithm to solve the NP-hard caching placement problem. Contribution/Results: Experiments demonstrate that the federated approach achieves performance close to centralized training. Under identical privacy budgets, the proposed strategy increases operator revenue by 12.7% over conventional cache-hit-rate (CHR)-based baselines.

Technology Category

• Machine Learning: Privacy
• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Security and Privacy: Privacy-enhancing technologies
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Performance of video streaming, which accounts for most of the traffic in wireless communication, can be significantly improved by caching popular videos at the wireless edge. Determining the cache content that optimizes performance (defined via a revenue function) is thus an important task, and prediction of the future demands based on past history can make this process much more efficient. However, since practical video caching networks involve various parties (e.g., users, isp, and csp) that do not wish to reveal information such as past history to each other, privacy-preserving solutions are required. Motivated by this, we propose a proactive caching method based on users' privacy-preserving multi-slot future demand predictions -- obtained from a trained Transformer -- to optimize revenue. Specifically, we first use a privacy-preserving fl algorithm to train a Transformer to predict multi-slot future demands of the users. However, prediction accuracy is not perfect and decreases the farther into the future the prediction is done. We model the impact of prediction errors invoking the file popularities, based on which we formulate a long-term system revenue optimization to make the cache placement decisions. As the formulated problem is NP-hard, we use a greedy algorithm to efficiently obtain an approximate solution. Simulation results validate that (i) the fl solution achieves results close to the centralized (non-privacy-preserving) solution and (ii) optimization of revenue may provide different solutions than the classical chr criterion.

Problem

• Optimize video caching revenue with privacy-preserving demand predictions
• Address privacy concerns in multi-party video caching networks
• Overcome NP-hard revenue optimization via efficient greedy algorithms

Innovation

• Transformer-based privacy-preserving demand prediction
• Federated learning for decentralized model training
• Greedy algorithm for NP-hard revenue optimization