چكيده به لاتين
The number of mobile network users and their applications is projected to increase at an unprecedented rate in the coming years, leading to high data traffic. Many of the services offered to users in the new generation of mobile phones require a response in the shortest time and are sensitive to delays. The main solution to meet these demands is a new design for mobile networks, hence the introduction of fog radio networks. These networks are referred to as cloud networks. In the cloud network, data needed to be sent from a distributed data center (cloud) that was far from the edge of the cell, and therefore the latency was too long for some applications and did not meet the quality of user experience required. . To solve this problem, servers were created near the edge of the wireless network with the ability to store data, so-called fog. The use of fog network architecture can be considered as an effective way to meet the demands of users sensitive to latency in wireless telecommunication systems. However, the limited capacity of these servers, compared to cloud servers, necessitates the allocation of optimal resources for fog-based systems.
In this regard, in this project, we will minimize the average delay in the transfer of the user's requested file in the fog system. In general, by increasing the rate of user information received, the delay in the user's requested file is reduced. The rate of information received by users can be calculated to achieve the minimum delay and the maximum rate of information received by the user. For this purpose, we will design how to select the optimal fog server, how to perform optimal storage of files in each fog server, and the optimal structure of the beam-forming vector in the downlink channel. In this scenario, multiple F-APs are connected to multiple users, and the F-APs are equipped with massive mimo antenna array. Our goal in this issue is to minimize the average delay in providing the service to the user, and the optimal allocation of F-AP and the optimal design of the beam-forming vector will be done according to the proposed restrictions. It will also determine how to optimally cache files in F-APs. In this project, first the desired optimization problem is proposed. Due to the non-convexity of this problem, the block coordinate reduction algorithm is used for sequential optimization on different blocks of variables. The problem of optimal file storage is solved using a linear program. To select the optimal beam shaping vector, a number of newly proposed transforms are used to rewrite the problem in a convex shape. Finally, we propose the Hungarian algorithm for selecting the best F-AP. Compared to single-antenna F-APs and multi-antenna F-APs with MRT beam-forming vector, the simulations show a significant effect of the optimized design on reducing the average delay in receiving data by users.
