A novel flow fluidity meter for BiNoC bandwidth resource allocation

Abstract

An enhanced on-chip network resource allocation method is developed for the Bi-directional Network-on-Chip (BiNoC) platform. Specifically, a novel flow Fluidity Meter (FM) is proposed to provide the real-time estimate of bandwidth utilization of the Virtual Channel (VC) buffer in a BiNoC router.The degree of fluidity of a packet transfer in each router is a reliable, yet low-cost method for measuring bandwidth utilization in the VC. We show that the overhead implementing this FM is very affordable. Finally, extensive simulation results verify that this proposed FM approach achieves superior performance compared to existing BiNoC resource allocation methods.

Network-Cognitive Traffic Control: A Fluidity-Aware On-Chip Communication

“…The major differences among this paper and our previous works of CARRS (Congestion Avoidance and Relief Routing Scheme) [11], BiNoC-FM (Bi-directional NoC Fluidity Meter) [12], and…”

“…In this section, we introduce the methodology foundations and implementation details that are based on our previous studies and researches [11,12,36] for the proposed NCogn.TC scheme, as follows.…”

“…Accordingly, in adaptive routing, a node can select to output packets to a neighboring node that is expected to be in less congestions when compared with other alternative selections, as shown in the scenario of Figure 6a. In contrast to avoiding congestions by selecting a profitable output port, Congestion Relief (CR) [11,12,37] is another congestion control scheme used for the input port selection, which provided a contention-aware input port selection algorithm for a node to receive the packets in the upstream node that is expected to contain more traffic flows than nodes connecting to other input ports, as shown in Figure 6b.…”

“…Referring to the primitive fluidity concept [11] and meter [12], in this section, we will explain how to apply both concepts of buffer fill and buffer fluidity to realize a Buffer Cognition Monitor (BCogM) in order to reflect the real-time traffic condition in multi-levels. Two six-state (L0 to L5) State Machines (SMs) as shown in Figure 10 are used to demonstrate the dynamic traffic control features.…”

“…There is yet an integrated approach to consider routing, congestion control, and flow control jointly to enhance performance of NoC communication. In this work, we leverage a novel fluidity concept proposed earlier [11,12], in order to develop a cognitive network traffic control scheme that features preferential data forwarding and corresponding integrated congestion control and flow control algorithms.…”