Economics-based Resource Collaboration And Pricing For Content Delivery In Wireless Networks

The rapid development of wireless communication technology and smart terminal devices,and the sharp increase of the wireless network users,which have led to an explosive growth network traffic.Meanwhile,wireless network users have higher requirements demand for content delivery services,such as lower latency and fresher contents.Taking bandwidth resources and caching space as examples,network resources supporting the above-mentioned content delivery services in the wireless networks are extremely scarce.Thus,the explosive data traffic of users and the limited network resources have become the primary contradiction,which hinders the development of the current wireless networks.Blindly adopting hardware upgrades to achieve network resource expansion not only brings huge expenses tooperators,but also makes it difficult to catch up with the rapidly growing demand for wireless services On the other hand,content delivery business has become one of the mainstream services in the wireless networks,and its service quality is also a key factor affecting user experience.How to efficiently utilize limited network resources to provide reliable services for the users and ensure the benefits of the resource providers have been an important research direction in the wireless communication networks.On the network operation side,the rapid development of diversified content services has prompted the emergence of new business models.The emerging new roles of network operations will participate in or affect the allocation of network resources,thereby affecting the experience of users.First,individual nodes in the network often belong to different interested parties,they are selfish and form a competitive relationship.Next,it may not be possible to obtain complete network status or users' privacy information during the resource allocation process,which causes unreasonable and inefficient resource allocations.Since game theory,auction theory,and contract theory in economic theory are mathematical theories and methods for studying competitive phenomena,which can effectively solve complex resource allocation problems.This article is based on relevant economic theories and methods,considering user request-driven content delivery,to research on resource collaboration and pricing issues in different scenarios in wireless networks.Specific research work includes:1.Motivate relay collaboration between nodes via the Device-to-Device(D2D)communication.First of all,this paper designs a moderate incentive mechanism to promote D2 D relay to complete content delivery via providing dynamic incentives to maximize the utility of the system.Based on the principal-agent framework,the problem of relay node incentive design is modeled,and we propose an incentive design algorithm based on greedy thinking.The system only makes incentive decisions according to the known information and achieve optimality of the algorithm.The simulation results show that the incentive mechanism we proposed can achieve a better system utility when compared with other mechanisms.Second,we consider the problem of content cooperative caching between nodes,and cooperative nodes realize content delivery through D2 D communication.The coalitional game from game theory is adopted for modeling the problem.According to whether the transfer of utility between nodes is allowed,and whether the cost of nodes forming coalitions can be ignored,we discuss the structure and stability of the coalitions formed by nodes in four scenarios.Through analysis,the conditions for forming a stable large coalition are obtained,and a merge-and-split algorithm in the absence of a stable large coalition is proposed to obtain a stable structure with small coalitions.The simulation results analyze the influence of different scenarios on the utilities of nodes.2.Support smooth playback of video streaming on the user side through joint allocation of heterogeneous resources.We established an two-hop video transmission model under partial edge cache.The Network Service Provider(NSP)provide heterogeneous resources to the Video Streaming Service Providers(VSSPs)including caching space and bandwidth to improve the quality of experience(QoE).The video streaming queue has random arrival and departure of video packets which are all modeled as G/G/1 queues,and the numerical solution of the user's QoE is obtained by the approximate diffusion method.This paper models the joint optimization problem of heterogeneous resources as an ascending auction to obtain the optimal allocation of heterogeneous resources and maximize the social utility with the limited heterogeneous resources.Based on the structural characteristics of the VSSP valuation,the proposed ascending auction aiming at active users can ensure that VSSP reflects their real needs during the auction process.The simulation results analyzed the influences of the total amount of resources and the size of resource units and shows that the proposed heterogeneous ascending auction can achieve a better social utility than the traditional homogeneous auctions.2.Supports smooth playback of video streaming through partial edge caching.This paper aim at joint allocation of heterogeneous resources for video transmission in wireless networks.The Network Service Provider(NSP)can provide limited heterogeneous resources including caching space and bandwidth to the Video Streaming Service Providers(VSSPs)to improve their users' QoE.In this paper,the joint optimization problem of heterogeneous resources is modeled as an ascending auction to obtain the optimal allocation of heterogeneous resources.The purpose is to achieve the maximum social surplus of limited heterogeneous resources,which is the sum of QoE of all users.The ascending auction proposed in this paper can ensure that VSSP truly reveals its own demands and maximize the social surplus.The simulation results analyze the influence of system parameters such as the total amount of resources and the size of resource units on the performance of the auction algorithm.And the proposed heterogeneous ascending auction is better than the traditional homogeneous auctions.3.Support user demand for content freshness by periodically updating the contents at edge caching servers.A new kind of content delivery business model is considered,where the NSP purchases periodically updated contents from the Content Providers(CPs),and then resells the contents to users with subscription fees.We assume that the user QoE is determined by the freshness of the contents,which is reflected by the Age of Information(AoI).In order to maximize its utility from the limited caching space and protect user demands for freshness,the NSP needs to determine the appropriate CP as the content source and sets the corresponding subscription fee.We decompose the above problem into a subscribe price problem and a content purchase problem,then formulate those two sub-problems by a Stackelberg game and auctions,respectively.By using backward induction,we derive the equilibrium of the Stackelberg game.We consider the information asymmetry that each CP may have a private production cost.Such private information is not necessarily complete between the two pairs of relationships(CPs-NSP and CP-CP).Then we study the content purchase problem in four cases.Due to the characteristics of AoI,we discover that the NSP must disclose some private information during the auction to ensure the execution of the auction.We prove such disclosure will not bring the loss of NSP's utility in all cases.Finally,in the simulation,the performance of the proposed auction algorithm under different conditions is compared and analyzed.We study the influence of various parameters in the system model on the utilities of NSP and the winner CP and the results show that free competition between CPs will enhance the utility of NSP under certain conditions.4.Support user demands for content freshness and service delay through periodically content updating at edge caching servers.We consider a content platform owned by the operator,while the platform periodically updates the content in the edge cache to satisfy user concerns about content freshness and service latency.Meanwhile,the impact of service latency caused by periodical updates on the user QoE is taken into consideration.This paper propose a content subscription mechanism based on contract theory and design the optimal update cycles and subscription fees for different types of content to maximize the utility of the content platform.We first derives the necessary and sufficient conditions to ensure the feasibility of the contract,and gives the best contract structure including the renewal cycle and subscription fee.Theoretical analysis shows that users who pay more attention to content AoI can get more fresh content and they need to bear most of the update cost.From the operator's perspective,providing fresher content can achieve a better utility,reflecting the necessity of content pricing from the perspective of freshness.