MAD-DDS: Memory-efficient automatic discovery data distribution service for large-scale distributed control network

초록

Therampant deployment of data distribution service (DDS) as middle-ware serviceproviders for industrial network platforms has been widely investigated. AllDDS-based node discovery protocols establish communication with its intendedtarget systems by accessing matched endpoints/nodes information. These matchedendpoints/nodes information is usually embedded with the system's programmablecontrol plane network which acts as the conveyor vehicle. The introduction ofsoftware defined networking (SDN) is to characterize the control plane from embeddeddata plane. The DDS implements the simple discovery protocol (SDP) as itsinherent node discovery protocol. Deploying DDS for data packet exchange inserver-based collaborative distributed networked control (DNC) systems has gainedtraction. The current automatic discovery protocol (ADP) based on SDP is fraughtwith real-time limitations such as high memory consumption and poor packettransmission. This work presents novel memory-efficient automatic discoverydata distribution service (MAD-DDS) with enhanced threshold bloom filters(ETBF) where ETBF stores transmission packets at simulation end-nodes. The packetis further adjusted using optimized binarization and decision thresholds insideADP, hence guaranteeing memory reduction. The testbed computation recordedsignificantly improved quality of service (QoS) whereas numerical results depictsignificant decline in memory consumption with consistent packet transmissionrates that produces increased computational capacity.

키워드