Network on Chip: a New Approach of QoS Metric Modeling Based on Calculus Theory

According to ITRS, in 2018, ICs will be able to integrate billions of transistors, with feature sizes around 18 nm and clock frequencies near to 10 GHz. In this context, Network on Chip (NoC) appears as an attractive solution to implement future high performance networks and more QoS management. A NoC is composed by IP cores (Intellectual Propriety) and switches connected among themselves by communication channels. End-to-End Delay (EED) communication is accomplished by the exchange of data among IP cores. Often, the structure of particular messages is not adequate for the communication purposes. This leads to the concept of packet switching. In the context of NoCs, packets are composed by header, payload, and trailer. Packets are divided into small pieces called Flits. It appears of importance, to meet the required performance in NoC hardware resources. It should be specified in an earlier step of the system design. The main attention should be given to the choice of some network parameters such as the physical buffer size in the node. The EED and packet loss are some of the critical QoS metrics. Some real-time and multimedia applications bound up these parameters and require specific hardware resources and particular management approaches in the NoC switch. A traffic contract (SLA, Service Level Agreement) specifies the ability of a network or protocol to give guaranteed performance, throughput or latency bounds based on mutually agreed measures, usually by prioritizing traffic. A defined Quality of Service (QoS) may be required for some types of network real time traffic or multimedia applications. The main goal of this paper is, using the Network on Chip modeling architecture, to define a QoS metric. We focus on the network delay bound and packet losses. This approach is based on the Network Calculus theory, a mathematical model to represent the data flows behavior between IPs interconnected over NoC. We propose an approach of QoS-metric based on QoS-parameter prioritization factors for multi applications-service using calculus model.