在多層邊緣計算架構中混合定價與延遲保證的工作分配機制:斯塔克爾伯格賽局方法

Abstract

行動邊緣計算是一個在網路邊緣藉由計算能力來降低萬物聯網裝置計算負擔的極具發展潛力典範。在對於資源相對匱乏的邊緣節點有迫切資源需求的情況下，由於大量任務佔據邊緣節點造成的排隊延遲不僅對於滿足萬物聯網裝置的使用者體驗造成巨大的阻礙，也對於邊緣節點服務提供商的利益產生影響。然而，由於不同邊緣節點與雲端的服務提供商不一定一樣，由資源相對豐富的邊緣節點提供的服務必須要有經濟上的補償來彌補因為增加的計算量所造成的能源費用與初期建設的投資成本。 因此，一個在多層邊緣計算架構下具有固定定價與動態定價的混合定價工作分配機制不但是重要而且是迫切需要的。天然存在於第一層邊緣節點與第二層邊緣節點的階層關係與相互依賴性可以被斯塔克爾伯格賽局充分掌握。在此種賽局中，第二層邊緣節點決定服務價格與可以保證的計算遲延來最大化利益，而第一層邊緣節點在給定第二層邊緣節點的行為之下透過決定分配在的第二層邊緣節點、雲端與自己本身的工作量來最大化自己的利益。奠基於最佳的工作量分配，一個誠實的允入控制機制可以設計給第一層邊緣節點去決定多少與哪些萬物聯網裝置在不違反端到端延遲要求的前提下可以被服務。斯塔克爾伯格賽局均衡的存在與唯一性可以被證明。 模擬結果確認了提出方法的有效性，而且也剖析了幾個有趣的見解。

Mobile Edge Computing (MEC) is a promising paradigm to ease the computation burden of Internet-of-Things (IoT) devices by leveraging computing capabilities at the network edge. With the yearning needs for resource provision from the comparatively resource-limited edge nodes, the queueing delay owing to the massive amount of tasks at the edge node not only poses a colossal impediment to achieving satisfactory quality of experience (QoE) for the IoT device but also to the benefits of the edge nodes as a result of escalating energy expenditure. However, due to the fact that the service provider of edge nodes and the cloud may not be the same, the computing service of computationally competent entities should not be available without economic compensation for the incurred energy expenditure and the capital investment. Therefore, the mixed pricing workload allocation mechanism where fixed and dynamic pricing schemes are both inspected in the multi-layer edge computing structure is both crucial and much-needed. The inherent hierarchy and interdependence between the second-layer edge node (SLEN) and first-layer edge nodes (FLENs) are captured by Stackelberg game in which the SLEN determines the service price and the guaranteed processing delay it can provide to maximize its profit and given the action of the SLEN, FLENs prudently select the workload distribution between the SLEN, the cloud and itself to maximize its profit as well. Grounded on the optimal workload allocation, a truthful admission control mechanism is designed for FLENs to decide how many and which IoT devices are served under the requirement of meeting E2E latency constraints. The uniqueness and the existence of the Stackelberg equilibrium are proved. Simulation results confirm the effectiveness of our scheme and several insights are illustrated.