Modeling and Performance Analysis of Multiserver Cloud Database over Quasi-Static Rayleigh Fading Channel
Mengying Chen,Yang Liu,Chen Dong,Wannian An,Xiaodong Xu,Ping Zhang
DOI: https://doi.org/10.1109/jiot.2024.3373783
IF: 10.6
2024-01-01
IEEE Internet of Things Journal
Abstract:With the development of communication in the post-5G era, the combination of communication and cloud computing becomes closer. In order to promote the further development of cloud-network, this article will study the performance of Multiserver cloud Database under the Communication quality of quasi-static Rayleigh fading channel with multiple antennas (MC-MD). The CLIENTS with unlimited customers, a COMMUNICATION SYSTEM subject to quasi-static Rayleigh fading, and CLOUD DATABASE with two-phase locking protocol are the three components of the MC-MD model. Transactions are 1)initiated by the CLIENTS, 2)transmitted to the CLOUD DATABASE through the COMMUNICATION SYSTEM for processing, 3)then returned to the CLIENTS. The indicators of the model is mathematically derived by using queuing theory. These include client's indicators [average concurrent quantity of the system in steady state (CQ), average transactions stay time of the system in steady state (ST), average queue length of the waiting area in steady state (QL), and average transactions wait time of the waiting area in steady state (WT)] and server's indicator [average number of service desks in the busy period at steady state (DN)]. Under the appropriate conditions, the results indicate that the theoretical value of service performance is basically consistent with the simulation value. Clearly, the high speed improves the service performance of the system and decreases the service pressure. On the basis of this, the optimization strategy is proposed and the simulation indicators Jitter of transactions sojourn time of the system in steady state (STJ) is added. The results show that the transaction scheduling optimization strategy effectively reduce the delay and its jitter.