Cross-layer Design And Performance Analysis Of Opportunistic Wireless Content Delivery Networks

Multimedia content transmission occupies the main body of mobile Internet traffic.However,excessive traffic charges in mobile communications inhibit the user's content requirements.Based on the characteristics of the mobile communication network,this paper proposes two kinds of designs for the uplink and downlink of opportunistic wireless content distribution networks through the way of cross layer design.The main purpose is to achieve low-cost content distribution through discovering and utilizing the redundant resources in the mobile communication network.The basic principle of uplink opportunistic content distribution is utilizing the low-cost access chance which provided by the open infrastructure(such as open access micro base station or WLAN hotspot),and uploading the content after the user moves into the open cell.In order to improve the uplink transmission efficiency,we introduce a user collaboration mechanism,allowing multiple collaborative users to receive local broadcast and cache content,and increasing the probability of uplink access through the independent mobility of multiple users.We first describe the protocol process of the uplink design,and then propose an analysis framework which fuses queuing model and stochastic geometric interference model.We thoroughly analyzing the theoretical performance of the uplink content distribution,which shows the impact of coverage sensing,user and base station densities,transmit power,user mobility,and packet size on the capacity-delay trade-off.Our study found that in the scenario of interference limitation,the fundamental limit of throughput capacity per device approaches to 0.7239?o/?u bits/s/Hz,where ?b and ?u represent the densities of base stations and mobile users respectively.The basic principle of downlink opportunistic content distribution network is dynamically sensing the redundant capacity of network,and providing elastic,low-cost proactive content cache through redundant capacity in the light of user subscription and browsing habits.In the downlink,we establish a cross-layer system model which fuses the design of application layer and the physical layer,and model the content distribution problem into an expectation maximization problem in a stochastic system.We systematically analyze the impact of redundant capacity,push time slot,number of users on the cost of system distribution,and research the impact of user arrival rate on opportunistic content delivery.Through numerical analysis,we validated the relationship between the proportion of normal traffic,the push slot and the number of users on the cost of the proactive content delivery of the system.The two schemes can make full use of the characteristics of mobile communication to optimize the content distribution network,which provides a useful theoretical reference for the development of new mobile content distribution system.