出版社：International Institute for Science, Technology Education
摘要：The multiprocessor SoC designs have more than one processor and huge memory on the same chip. SoC consists of hardware cores and software cores ,multiple processors, embedded DRAM and connectors between cores .A wide range of MPSOC architectures have been developed over the past decade. This paper surveys the history of various On-Chip communication architectures present in the design of MPSoC. This acts as a primary factor of overall performance in complex SoC designs. Some of the various techniques that have driven the design of MpSoC has been discussed. Dynamically configurable communication architectures are found to improve the system performance. Currently On-chip interconnection networks are mostly implemented using shared buses which are the most common medium. The arbitration plays a crucial role in determining performance of bus-based system, as it assigns priorities, with which processor is granted the access to the shared communication resources. In the conventional arbitration algorithms there are some drawbacks such as bus starvation problem and low system performance. The bus should provide each component a flexible and utmost share of on-chip communication bandwidth and should improve the latency in access of the shared bus. The performance of SoC is improved using the probabilistic round robin algorithm with regard to the parameters, latency.Thus in this paper various issues related to bus arbitration related to design of MPSoC is analysed.
