Risk Optimization for Revenue-Driven Wireless Video Broadcasting Systems: A Copula-Based Framework

Abstract

The revenue of wireless service providers (WSPs) relies on their ability to efficiently satisfy the variable demands from end users (EUs). However, emerging video services create new risks owing to the diverse content requirements of heterogeneous EUs operating in uncertain markets. It is challenging as the risks created by demands and prices are highly uncertain. In this work, a risk problem of high revenues is studied in which multiple video content items with different prices are broadcast to wireless EUs. The video content prices differ in their value functions in terms of popularity, ratings, and types. The objective is to maximize the revenue of WSPs under a certain Value-at-Risk (VaR) by adjusting the content prices and allocation of bandwidth resources. Furthermore, a VaR-based optimization framework for wireless video broadcasting systems is presented. First, the content characteristics are analyzed, and a copula model is then used to build the content value structure. The copula of a multivariate distribution corresponds to the description of the price-dependent structure. Second, a risk analysis for the effects of price fluctuations on revenues caused by uncertainty is conducted. A VaR model is associated with changes in the prices and allocated rates. Copulas are used to derive a bound on the VaR for functions of dependent risks. Subsequently, detailed representations are provided to identify the distributional bounds for revenue functions of dependent risks. Lastly, a VaR-based framework that optimizes pricing and bandwidth provision is presented. For the solution, the risk regions of WSPs are modeled as polymatroidal structures to minimize the risk caused by different service demands and variable market prices. Experiments on different price markets demonstrated that the proposed method is effective, thereby verifying the feasibility of the proposed method.