考慮商業競爭、技術擴散與技術替代以進行電信用戶數預測

Abstract

5G技術因為雲端運算與人工智慧而開啟了行動商務的新紀元，預測用戶數可以幫助電信商事先規劃財務預算與促銷方案。然而，過去研究大都倚賴歷史資料，很少針對商業競爭、技術擴散或替代效果的影響進行深入的探討。本研究提出一個整合性的架構已達成下列目標：（1）納入主要競爭對手間的互動關係以估算市場平衡點及市占率、（2）針對同世代內的技術擴散進行3G與4G+5G的最大用戶數與出現時間點的估算、（3）將3G與4G間的技術替代延伸到5G的新市場。為了實證研究的有效性，台灣三大電信公司，包括中華電信、台灣大哥大與遠傳電信的用戶數納入商業競爭的模型建構。研究發現，規模較大的公司（受利）會捕食規模較小的公司（受害），且三家電信商用戶數在市場均衡時均呈現下滑。雖然多代模型的預測效果較佳，但給定3G與4G的迴歸參數，單代模型能協助電信商估算5G的最大用戶數與出現時間以利進行財務規劃與基地台建置。

Today, 5G technology coupled with cloud computing and artificial intelligence have boosted a new era of mobile commerce. Thus, forecasting telecom subscribers can help firms conduct planning for financial budgets and promotion plans. However, past studies rarely incorporate commercial competition, technological diffusion, and substitution into the prediction of adopters. To facilitate research gaps, this research proposes a novel framework to accomplish the following goals: (1) Interaction dynamics between the competing firms is incorporated to estimate market equilibriums and market shares. (2) Diffusion dynamics within the same generation is used to respectively estimate the maximum of 3G and 4G+5G subscribers and when to appear. (3) Substitution dynamics between 3G and 4G is extended to track revolution trends in the rising 5G market. To justify research validity, Chung Hwa (CH), Taiwan Mobile (TW), and Far Eastern (FE), are illustrated. Research findings show that the relationships, predator–prey, exist between large-scale firms (beneficial) and small-scale firms (harmful) and the subscribers for all the top three firms are expected to decrease in market equilibriums. Although multi-generation models perform much better in forecasting subscribers, single-generation models are able to estimate the maximum of the 5G subscribers and corresponding time occurrence given the fitted parameters in 3G and 4G. Therefore, it can provide a quantitative basis to help telecom firms conduct financial planning for constructing base stations.